120
SEED UPGRADE AND GERMINATION STRATEGIES FOR ALNUS TENUIFOLIA AND BETULA OCCIDENTALIS BY CINDY LEE JONES, B.S. A Thesis submitted to the Graduate School in partial fulfillment of the requirements for the degree Master of Science Major Subject: Horticulture Minor Subject: Experimental Statistics New Mexico State University May 2000

SEED UPGRADE AND GERMINATION STRATEGIES BETULA

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Page 1: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

A Thesis submitted to the Graduate School

in partial fulfillment of the requirements

for the degree

Master of Science

Major Subject Horticulture

~ Minor Subject Experimental Statistics

New Mexico State University

May 2000

Seed Upgrade and Germination Strategies for Alnus tenuifoha and

Betula occidentahs a thesis prepared by Cindy Lee Jones in partial

fulfillment of the requirements for the degree Master of Science has

been approved and accepted by the following

Chair of the Examining Committee

Committee in charge

Dr John T Harrington Chair

Dr David R Dreesen

Dr Leigh Murray

Dr Geno A Picchioni

11

DEDICATION

This work is dedicated to my mother Eula1a Jones who

supported me unwaveringly in this eifort and who sacrificed and

worked as much as I to accomplish the end result and to my late

father Earl Jones who never doubted me even when I doubted

myself

ill

ACKNOWLEDGMENTS

I wish to thank my advisor Dr John T Harrington for his

assistance and support in every phase of this study and for allowing

me the use of the facilities at the Mora Research Center in Mora

New Mexico

I wish to thank Dr Leigh Murray for her extensive assistance in

the data analysis of this project and for her professional informative

and helpful manner

Thanks also to Dr David Dreesen for his guidance in

evaluating ideas for the study and to Dr Geno Picchioni for his

support guidance and excellent instruction over the years

My special thanks to Molycorp for the funding which made this

study possible

lowe a greatdebt ofgratitude to my cousin Eugenia Shepan

and her husband Don who opened their home to me and gave me

their love and support and to my supervisor at University Hospital

Virginia Nymeyer for her faithful friendship

IV

VITA

October 8 1957 Born at Clayton New Mexico

1975 Graduated from Belen High School Belen New Mexico

1984-1987 Medical Technologist Santa Fe Medical Labs Santa Fe New Mexico

1987-present Medical Technologist Tricore Reference Laboratories at University Hospital Albuquerque New Mexico

1997--present Research Assistant Department ofAgronomy and Horticulture New Mexico State University

PROFESSIONAL AND HONORARY SOCIETIES

American Society for Horticultural Science

American Society of Clinical Pathologists

American Society for Clinical Laboratory Science

Phi Kappa Phi

FIELD OF STUDY

Major Field Horticulture

Minor Field Experimental Statistics

v

ABSTRACT

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

Master of Science in Horticulture

New Mexico State University

Las Cruces New Mexico 2000

Dr John T Harrington Chair

Little is known about the propagation of thinleaf alder (Alnus

tenuifolia) and water birch (Betula occidentalis) These species

native to New Mexico have the potential to be useful trees for

rehabilitation of disturbed lands and possibly landscaping An

efficient and economical method for propagation is needed Birch

and alder share many common seed characteristics including small

V1

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 2: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

Seed Upgrade and Germination Strategies for Alnus tenuifoha and

Betula occidentahs a thesis prepared by Cindy Lee Jones in partial

fulfillment of the requirements for the degree Master of Science has

been approved and accepted by the following

Chair of the Examining Committee

Committee in charge

Dr John T Harrington Chair

Dr David R Dreesen

Dr Leigh Murray

Dr Geno A Picchioni

11

DEDICATION

This work is dedicated to my mother Eula1a Jones who

supported me unwaveringly in this eifort and who sacrificed and

worked as much as I to accomplish the end result and to my late

father Earl Jones who never doubted me even when I doubted

myself

ill

ACKNOWLEDGMENTS

I wish to thank my advisor Dr John T Harrington for his

assistance and support in every phase of this study and for allowing

me the use of the facilities at the Mora Research Center in Mora

New Mexico

I wish to thank Dr Leigh Murray for her extensive assistance in

the data analysis of this project and for her professional informative

and helpful manner

Thanks also to Dr David Dreesen for his guidance in

evaluating ideas for the study and to Dr Geno Picchioni for his

support guidance and excellent instruction over the years

My special thanks to Molycorp for the funding which made this

study possible

lowe a greatdebt ofgratitude to my cousin Eugenia Shepan

and her husband Don who opened their home to me and gave me

their love and support and to my supervisor at University Hospital

Virginia Nymeyer for her faithful friendship

IV

VITA

October 8 1957 Born at Clayton New Mexico

1975 Graduated from Belen High School Belen New Mexico

1984-1987 Medical Technologist Santa Fe Medical Labs Santa Fe New Mexico

1987-present Medical Technologist Tricore Reference Laboratories at University Hospital Albuquerque New Mexico

1997--present Research Assistant Department ofAgronomy and Horticulture New Mexico State University

PROFESSIONAL AND HONORARY SOCIETIES

American Society for Horticultural Science

American Society of Clinical Pathologists

American Society for Clinical Laboratory Science

Phi Kappa Phi

FIELD OF STUDY

Major Field Horticulture

Minor Field Experimental Statistics

v

ABSTRACT

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

Master of Science in Horticulture

New Mexico State University

Las Cruces New Mexico 2000

Dr John T Harrington Chair

Little is known about the propagation of thinleaf alder (Alnus

tenuifolia) and water birch (Betula occidentalis) These species

native to New Mexico have the potential to be useful trees for

rehabilitation of disturbed lands and possibly landscaping An

efficient and economical method for propagation is needed Birch

and alder share many common seed characteristics including small

V1

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 3: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

DEDICATION

This work is dedicated to my mother Eula1a Jones who

supported me unwaveringly in this eifort and who sacrificed and

worked as much as I to accomplish the end result and to my late

father Earl Jones who never doubted me even when I doubted

myself

ill

ACKNOWLEDGMENTS

I wish to thank my advisor Dr John T Harrington for his

assistance and support in every phase of this study and for allowing

me the use of the facilities at the Mora Research Center in Mora

New Mexico

I wish to thank Dr Leigh Murray for her extensive assistance in

the data analysis of this project and for her professional informative

and helpful manner

Thanks also to Dr David Dreesen for his guidance in

evaluating ideas for the study and to Dr Geno Picchioni for his

support guidance and excellent instruction over the years

My special thanks to Molycorp for the funding which made this

study possible

lowe a greatdebt ofgratitude to my cousin Eugenia Shepan

and her husband Don who opened their home to me and gave me

their love and support and to my supervisor at University Hospital

Virginia Nymeyer for her faithful friendship

IV

VITA

October 8 1957 Born at Clayton New Mexico

1975 Graduated from Belen High School Belen New Mexico

1984-1987 Medical Technologist Santa Fe Medical Labs Santa Fe New Mexico

1987-present Medical Technologist Tricore Reference Laboratories at University Hospital Albuquerque New Mexico

1997--present Research Assistant Department ofAgronomy and Horticulture New Mexico State University

PROFESSIONAL AND HONORARY SOCIETIES

American Society for Horticultural Science

American Society of Clinical Pathologists

American Society for Clinical Laboratory Science

Phi Kappa Phi

FIELD OF STUDY

Major Field Horticulture

Minor Field Experimental Statistics

v

ABSTRACT

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

Master of Science in Horticulture

New Mexico State University

Las Cruces New Mexico 2000

Dr John T Harrington Chair

Little is known about the propagation of thinleaf alder (Alnus

tenuifolia) and water birch (Betula occidentalis) These species

native to New Mexico have the potential to be useful trees for

rehabilitation of disturbed lands and possibly landscaping An

efficient and economical method for propagation is needed Birch

and alder share many common seed characteristics including small

V1

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 4: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

ACKNOWLEDGMENTS

I wish to thank my advisor Dr John T Harrington for his

assistance and support in every phase of this study and for allowing

me the use of the facilities at the Mora Research Center in Mora

New Mexico

I wish to thank Dr Leigh Murray for her extensive assistance in

the data analysis of this project and for her professional informative

and helpful manner

Thanks also to Dr David Dreesen for his guidance in

evaluating ideas for the study and to Dr Geno Picchioni for his

support guidance and excellent instruction over the years

My special thanks to Molycorp for the funding which made this

study possible

lowe a greatdebt ofgratitude to my cousin Eugenia Shepan

and her husband Don who opened their home to me and gave me

their love and support and to my supervisor at University Hospital

Virginia Nymeyer for her faithful friendship

IV

VITA

October 8 1957 Born at Clayton New Mexico

1975 Graduated from Belen High School Belen New Mexico

1984-1987 Medical Technologist Santa Fe Medical Labs Santa Fe New Mexico

1987-present Medical Technologist Tricore Reference Laboratories at University Hospital Albuquerque New Mexico

1997--present Research Assistant Department ofAgronomy and Horticulture New Mexico State University

PROFESSIONAL AND HONORARY SOCIETIES

American Society for Horticultural Science

American Society of Clinical Pathologists

American Society for Clinical Laboratory Science

Phi Kappa Phi

FIELD OF STUDY

Major Field Horticulture

Minor Field Experimental Statistics

v

ABSTRACT

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

Master of Science in Horticulture

New Mexico State University

Las Cruces New Mexico 2000

Dr John T Harrington Chair

Little is known about the propagation of thinleaf alder (Alnus

tenuifolia) and water birch (Betula occidentalis) These species

native to New Mexico have the potential to be useful trees for

rehabilitation of disturbed lands and possibly landscaping An

efficient and economical method for propagation is needed Birch

and alder share many common seed characteristics including small

V1

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 5: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

VITA

October 8 1957 Born at Clayton New Mexico

1975 Graduated from Belen High School Belen New Mexico

1984-1987 Medical Technologist Santa Fe Medical Labs Santa Fe New Mexico

1987-present Medical Technologist Tricore Reference Laboratories at University Hospital Albuquerque New Mexico

1997--present Research Assistant Department ofAgronomy and Horticulture New Mexico State University

PROFESSIONAL AND HONORARY SOCIETIES

American Society for Horticultural Science

American Society of Clinical Pathologists

American Society for Clinical Laboratory Science

Phi Kappa Phi

FIELD OF STUDY

Major Field Horticulture

Minor Field Experimental Statistics

v

ABSTRACT

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

Master of Science in Horticulture

New Mexico State University

Las Cruces New Mexico 2000

Dr John T Harrington Chair

Little is known about the propagation of thinleaf alder (Alnus

tenuifolia) and water birch (Betula occidentalis) These species

native to New Mexico have the potential to be useful trees for

rehabilitation of disturbed lands and possibly landscaping An

efficient and economical method for propagation is needed Birch

and alder share many common seed characteristics including small

V1

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 6: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

ABSTRACT

SEED UPGRADE AND GERMINATION STRATEGIES

FOR ALNUS TENUIFOLIA AND

BETULA OCCIDENTALIS

BY

CINDY LEE JONES BS

Master of Science in Horticulture

New Mexico State University

Las Cruces New Mexico 2000

Dr John T Harrington Chair

Little is known about the propagation of thinleaf alder (Alnus

tenuifolia) and water birch (Betula occidentalis) These species

native to New Mexico have the potential to be useful trees for

rehabilitation of disturbed lands and possibly landscaping An

efficient and economical method for propagation is needed Birch

and alder share many common seed characteristics including small

V1

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 7: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

size prolific seed production with low viability and seed dormancy

These characteristics make propagation by seed problematic

Stratification is used to break dormancy in many species including

alder Vegetative propagation is often difficult for alder and birch

The most likely method for propagation is seed in view of the need

for genetic diversity in plants used in restoration Problems with seed

propagation might be solved by refining or upgrading the seed and

the use of stratification to break seed dormancy

The LDS method developed by Milan Simak (1983) for

conifer seeds was evaluated for its effectiveness in refining thinleaf

alder and water birch seeds LDS involves imbibing the seeds

partially re-drying to leave a residue of moisture and separating by a

density method The viable seeds should retain moisture while the

non-viable should not thus creating a density differential between

viable and non-viable seeds

Thinleaf alder and water birch seeds were subjected to simple

density separation by the specific gravity method with and without

IDS treatment Untreated dry seeds untreated imbibed seeds and

the floating and sinking IDS treated seed fractions were subjected to

VII

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 8: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

germination tests Three levels of stratification were used in

combination with the LDS study to evaluate the presence of

physiological dormancy in thinleaf alder and water birch

Seed refinement was determined to be useful in improving

germination of thin leaf alder and water birch LDS methods were

found to be useful in the case of thinleaf alder while water birch

germination benefitted most from a simple density separation in

ethanol Twenty~eight days of stratification improved water birch

germination but the actual gain in percentage was small

Stratification was not shown conclusively to be useful in improving

thinleaf alder germination

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 9: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

TABLE OF CONTENTS

Page

LIST OF TABLES XlI

LIST OF FIGURES XVI

INTRODUCTION 1

LITERATURE REVIEW 3

Revegetation and Reconstruction 3

Species Selection 4

Planting Methods 5

Birch and Alder Suitability in Reconstruction 7

Production of Stock Plants 9

Seed Dormancy and Methods to Overcome It 10

Germination Requirements 14

Thinleaf Alder 15

Water Birch 17

Seed Quality Improvements 18

OBJECTIVES OF THIS STUDY 20

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 10: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

METHODS AND MATERIALS

Page

21

Sources 21

Separation Media 23

Seed Refinement 25

Thinleaf Alder 25

Water Birch 28

Germination Requirements 30

Thinleaf Alder 30

Water Birch 33

DATA ANALYSIS 36

RESULTS 42

Seed Refinement 42

Thinleaf Alder Fill Enhancement 42

Thinleaf Alder Recovery 47

Water Birch Fill Enhancement 49

Water Birch Recovery 54

Germination Requirements 55

Thinleaf Alder 55

x

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 11: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

Page

Water Birch 65

DISCUSSION 76

Seed Refinement 76

Germination Requirements 85

Thinleaf Alder 85

Water Birch 90

General Observations 92

LITERATURE CITED 95

Xl

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 12: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

LIST OF TABLES

Table Page

221 Seed Source Locations and Elevations

2 Alder Preparation Protocols for Seed Refinement 26

3 Birch Preparation Protocols for Seed Refinement 29

4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder 32

5 Treatment Combinations for Experimental Layout of Randomized Complete Block--Water Birch 35

6 Analysis of Variance Table for Thinleaf Alder Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 42

7 Thinleaf Alder Percentage ofFilled Seeds in Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced by Preparation Protocol 45

8 Thinleaf Alder Percentage of Filled Seeds as Influenced by Separation Fraction 46

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 13: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

Table Page

9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 47

10 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Preparation Protocol 48

11 Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source 48

12 Analysis ofV ariance Table for Water Birch Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source--Factorial Analysis 49

13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction 51

14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol and Seed Source--Factorial Analysis 54

15 Thinleaf Alder Percentage Germination Analysis ofVariance Table--Factorial Analysis 56

16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

XU1

56

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 14: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

Table Page

17 Thinleaf Alder Percentage Germination as Influenced by Separation--Factorial Analysis 57

18 Analysis of Variance Table for Thinleaf Alder Percentage Germination as Influenced By Treatment Combination and Seed Source--Augmented Factorial 61

19 Thinleaf Alder Analysis of Contrasts--Augmented Factorial 61

20 Thinleaf Alder Analysis ofV ariance Table--Factorial Analysis without Red River Canyon Seed Source 64

21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source--Factorial Analysis 64

22 Water Birch Percentage Germination Analysis of Variance Table--Factorial Analysis 66

23 Water Birch Percentage Germination as Influenced by Stratification--Factorial Analysis 66

24 Water Birch Percentage Germination as Influenced by Separation--Factorial Analysis 67

XIV

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 15: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

Table Page

25 Water Birch Percentage Gennination as Influenced by Seed Source--Factorial Analysis 67

26 Analysis ofVariance Table for Water Birch Percentage Gennination as Influenced By Treatment COInbination and Seed Source--Augmented Factorial 73

27 Water Birch Analysis of Contrasts--Augmented Factorial 73

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 16: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

LIST OF FIGURES

PageFigure

1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 44

2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction 52

3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction 53

4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source 58

5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 59

6 Alder Percentage Germination as Influenced by Imbibition and Seed Source 62

7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source 68

8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction 69

9 Birch Percentage Germination as Influenced by Stratification and Seed Source 70

10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification 71

XVI

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 17: SEED UPGRADE AND GERMINATION STRATEGIES BETULA

Figure Page

11 Birch Percentage Germination as Influenced by Imbibition and Seed Source 75

12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 81

13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol 83

INTRODUCTION

Birch (Betula) and alder (Alnus) are two genera of Betulaceae

trees found in riparian areas throughout New Mexico The presence

ofbirch and alder in riparian zones of New Mexico has been noted in

many early surveys of the region (Britton 1908 Sargent 1901 1905

Wooton and Standley 1915) Water birch (Betula ocddentalisHook

formerly B fontinalis Sarg) is found in the northern mountains of the

state (Martin and Hutchins 1980) New Mexico has two species of

alder Arizona alder (Alnus oblongifolia Torr) which is found in the

mountains of southwest New Mexico (Martin and Hutchins 1980

Vines 1960) and thinleaf alder (Alnus tenuifolia Nutt) designated by

Carter (1997) as Alnus incana ssp tenuifolia Nutt found in the

northern and western mountains (Martin and Hutchins 1980 Vines

1960) Until recently existence of these species has been of interest

mainly from a botanical standpoint However with increasing landshy

use in the western United States these trees may have a further

purpose in the revegetation of degraded riparian areas and as oasis

plants for those interested in native landscapes (Phillips 1995)

Successful revegetation of degraded areas is influenced by many

1

factors including the site conditions commonly encountered and the

chosen plant material Desirable plant material should be wellshy

adapted to the site have high survival and be economical to obtain or

produce

LITERATURE REVIEW

Revegetation and Reconstruction

Strategies for revegetation of disturbed lands are generally

divided into three categories restoration reclamation and

rehabilitation Restoration is the complete replication of the original

conditions species habitat and function of the area Reclamation

involves returning the area to a condition that is habitable by the

organisms that were originally present or organisms that approximate

the original inhabitants Rehabilitation involves returning the land to

a form and function which conforms to a prior land-use plan

including a stable ecological state that does not contribute

substantially to environmental deterioration and is consistent with

surrounding aesthetic values (Allen 1988 National Academy of

Sciences 1974) These three categories have been collectively termed

reconstruction by Allen (1988) Complete restoration is often not

practical as certain requisite intermediate conditions of varying

durations maybe necessary In the arid western United States

natural succession is slow and dependence on natural process risks

further site degradation (National Academy of Sciences 1974)

3

Reclamation and rehabilitation may be more workable concepts A

practical guiding philosophy would be the objective to create a stable

ecosystem that is compositionally and functionally similar to that

which existed prior to human disturbance with the realization that

such a goal is not completely attainable (Burton et al 1988)

Species Selection

It has long been the philosophy of those involved in

reconstruction efforts that the use ofnative and diverse species is

desirable rather than dependence on a few proven species (Daniel

et al 1979 Harker et al 1993 Nielson and Peterson 1973) The

rationale is that native species are better adapted to adverse site

conditions such as low moisture and high surface temperatures and

exposure (Nielson and Peterson 1973) Only native species survived

in a European study even though exotic species examined also

possessed characteristics which were well-adapted to the site (Herrera

et al 1993) Use of diverse native plant species can enhance

reconstruction efforts and sustain more diverse wildlife populations

(Harker et al 1993) Using plant material of local provenance (origin

of seed) to maximize survival is also important (Albers and Carpenter

4

1979 Burton et al 1988 Daniel et al 1979 Hobbs 1984) Species of

plants evolve within their habitat to site conditions including edaphic

topographic and climatic conditions such as temperature (Bewley and

Black 1994) photoperiod (Currie 1990) and growing season A plant

with origins in southern latitudes may not properly harden off for

winter in time to avoid early frost when grown in northern latitudes

with longer day1engths while a plant from northern latitudes may not

have optimal shoot growth in the shorter day1ength of southern areas

(Fowells 1965 Lane 1993)

PlantingMethods

Natural colonization processes can take anywhere from ten to

hundreds ofyears depending on site conditions (National Academy

of Sciences 1974) Planting methods used in reconstruction include

direct seeding wildling transplants and use ofbare-root or

containerized transplant material (Schubert et al 1970) Direct

seeding is often the least expensive planting method but success with

woody species is frequently limited Predation of seed germination

failure and adverse conditions for germinants can result in planting

failure (Fowells 1965 Haeussler et al 1995 Hibbs et al 1994

5

Monsen 1984 Pratt 1986) Wildling transplants may have poor

survival ifplanting is not timed properly and done carefully (Schubert

et al 1970) Use ofnursery grown seedlings bare-root or

containerized can improve survival rates relative to other

reconstruction efforts (Hobbs 1984) The ability to match stock type

(source physiological and morphological condition) to the site

known as the target seedling concept (Rose et al 1990) and greater

latitude in planting conditions (timing) can contribute to improved

transplant success of nursery stock relative to wildlings Combining

direct seeding ofnon-woody plants and nursery-grown seedlings can

be the most efficient and economical method of reconstruction when

costs ofproducing container stock can be kept low (Belcher 1982

Dunlap and Barnett 1984 Rose et al 1990) The success of

reconstruction efforts is heavily dependent on site conditions and the

quality of the plant material used (Monsen 1984) In tum quality of

plant material is dependent on well-developed germination and

culture protocols The economic feasibility of stock propagation for

reconstruction work is dependent on finding methods to efficiently

upgrade seed quality (proportion ofgerminable seeds) and optimize

6

germination capacity and seedling survival (Belcher 1982 Bonner

1984)

Birch and Alder Suitability in Reconstruction

Montane riparian vegetation zones are contained in areas where

the supply ofwater is constant (perennial) as well as areas with an

ephemeral (intermittent) water supply Riparian zones contain both

obligate and facultative riparian species Facultative riparian species

are also found in surrounding open spaces and in high cool nonshy

riparian locations (Dick-Peddie 1993) Riparian vegetation follows an

elevational gradient from the source to the mouth of the drainage

perpendicular to the zone of upland vegetation (Dick-Peddie 1993)

Other habitats where water may be caught but are not part of a true

drainage are termed pseudoriparian Pseudoriparian habitats include

gullies roadside ditches and the bottoms of talus slopes (Dick-Peddie

1993) Most of the obligate riparian species found in riparian and

pseudoriparian areas are adapted to flood conditions with the ability

to rapidly reproduce and colonize a devastated area Characteristics

ofobligate riparian species include prolific seed production efficient

7

seed dispersal fast growth short life-cycles and rapid attainment of

reproductive stage (Dick-Peddie 1993)

Birch and alder species are generally confined to montane

riparian zones (Elias 1980) Members ofboth genera have properties

indicative of obligate riparian species including fast growth prolific

seed production and short life-cycle these properties also make

members of these genera suitable candidates for use in reconstruction

efforts (Elias 1980) Birch and alder are known as pioneer species

which can successfully establish on denuded areas (Young and Young

1992) and which prefer mineral soil for germination and early growth

(Haeussler et al 1995 Schalin 1968) In addition most alder species

including thirlleaf alder and Arizona alder have the ability to fix

atmospheric nitrogen via a symbiotic relationship with root-nodule

forming species of Frankia actinomycetes (Bond 195519711976

Virtanen 1957) Many researchers believe the formation of a dynamic

rhizosphere of this type is critical to the rehabilitation of degraded

lands (Herrera et al 1993 Whitford 1988) Biological nitrogen

fixation in conjunction with the production of large amounts of litter

has been shown to help build up organic matter nitrogen and

8

improve soil structure in deficient soils such as glacial till (Bollen and

Lu 1968 Crocker and Major 1955 Tarrant and Trappe 1971)

Biological nitrogen fixation can also improve conditions for other

non-nitrogen fixing species (Tarrant 1961) and enhance species

diversity (Franklin and Pechanec 1968)

The use ofthese deciduous trees with the objective of improving

the site conditions (ie shade nutrients and organic matter) for other

species (Albers and Carpenter 1979) is a valuable strategy in the

reconstruction of disturbed areas such as mine spoils

Production ofStockP1ants

Efficient propagation ofnursery stock from seed requires

extensive knowledge of the germination requirements and cultural

methods needed for the particular species Little is known about the

propagation requirements for the two species used in this study

thinleaf alder and water birch This deficit is due in part to a lack of

demand for these species in the past Extensive work has been done

on the propagation of other species within the Alnus and Betula

genera specifically those species of commercial value to the timber

industry such as red alder (A rubra Bong) and paper birch (B

9

papyrifera Marsh) Information generated from propagation studies

on these species has elucidated some universal seed characteristics

and germination requirements for members ofBetulaceae Seeds aremiddot

characteristically very small and light and may have a winged

integument to aid in wind dispersal Average seed density for B

ocddentalis is about 2500 seeds per gram while A tenuifolia

averages about 1488 seeds per gram (Vines 1960) Seed quality and

germination capacity are often very low as it is difficult to separate

sound from empty seeds when size and weight are so low (Brinkman

1974 Schopmeyer 1974) Seed quality may vary considerably from

harvest to harvest (Bjorkbom et al 1965) Within species

germination requirements may differ with provenance (Fowler and

Dwight 1964 Wilcox 1968) or even within a provenance (Bjorkbom

et al 1965 Schopmeyer 1974) In some instances the requirements

for germination may be met but germination does not occur a

condition referred to as dormancy

Seed Dormanqr and Methods to Overcome It

Dormancy in seeds is defined as the condition where seeds will

not germinate even when environmental conditions (water

10

temperature and aeration) are permissive for germination (Bewley

and Black 1994 Hartmann et al 1997) This mechanism ensures that

germination does not take place in less than optimum conditions or at

the wrong time (Bewley and Black 1994 Thompson 1971) For

example in some species seeds of southern provenance require

longer stratifications (Fowler and Dwight 1964) probably to prevent

germination in areas where there are intermittent periods ofwarm

weather followed by frost Seed dormancy results from a

combination ofgenetic and environmental conditions and it is not

always possible to predict the dormancy of a particular species from

characteristics of other species within the genus (Schopmeyer 1974)

There are different systems for classifying dormancy but the

condition may be divided into four basic types exogenous

endogenous double or combinational and secondary (Hartmann et

al 1997) The seed dormancy exhibited by birch and alder falls under

the category of endogenous dormancy a dormancy imposed by

embryonic factors This includes morphological dormancy (an

underdeveloped embryo) and physiological dormancy ofvarying

degrees (non-deep intermediate and deep) Non-deep physiological

11

dormancy is characterized by the need for after-ripening or exposure

to red light (photodormancy) Intermediate physiological dormancy

is characterized by the need for moderate periods of cold stratification

(up to 56 days) Deep physiological dormancy requires long periods

of cold stratification more than 56 days (Hartmann et al 1997)

Seeds ofboth Alnus and Betula exhibit varying degrees of

dormancy in most cases broken by coolmoist stratification andor

germination under red light (Brinkman 1974 Dirr and Heuser 1987

Schopmeyer 1974 Young and Young 1992) In some species of these

genera chemical treatments such as potassium nitrate have been

effective to overcome dormancy (Bradbeer 1988 Hartmann et al

1997 Young et al 1984) Many birch species are known to possess a

phytochrome light detection system which prevents germination

when seeds are buried too deep to allow seedling survival after

germination (Bewley and Black 1994 Black and Wareing 1955

Bradbeer 1988) Where the phytochrome detection mechanism is

present exposure to red light during germination is required for

breaking dormancy Most species of birch and alder have seeds that

ripen in late summer or early fall fall germination would result in

12

seedling loss over the winter so an after-ripening or stratification

requirement decreases the possibility of fall germination Joseph

(1929) found non-stratified birch seeds had a higher temperature

requirement for germination The current theory is that stratification

causes phase changes in membrane fluidity and triggers membraneshy

related signal transduction pathways activating enzymes and

hormones thus allowing dormancy release (Bewley and Black 1994

Ross and Bradbeer 1971)

Leaching of certain chemical inhibitors from seeds can also

break dormancy it maybe that this is part of the mechanism by

which photo dormancy is broken by moist stratification as only small

amounts of moisture are needed (Brad beer 1988) Research indicates

that the testa and pericarp of the seeds are involved in dormancy not

because they contain the inhibitor but because they prevent leaching

of the inhibitor (Villiers and Wareing 1964 Webb and Wareing

1972) Ru40lf (1950) found that cold-soaking might in some cases be

an acceptable substitute for stratification in some conifer species this

might be due to the leaching mechanism

13

The role ofpotassium nitrate in breaking dormancy has not

been clarified but there is speculation that the nitrogen supplied or

the oxygenating properties of the nitrate are involved (Brad beer

1988) Biswas et al (1972) found that the chemical treatment

enhanced the effect of stratification but did not necessarily replace it

Hilton (1985) found the germination-stimulating properties ofnitrate

depend on the presence of light nitrate in the presence of red light is

believed to be a cofactor to the phytochrome system which is involved

in the synthesis ofgibberellins that promote germination (Hilhorst et

al 1986)

Germination Requirements

General requirements for germination include moisture

favorable temperatures adeq-qate gas exchange and for some species I)

light In the presence of these conditions the quiescent seed can

imbibe water causing the seed to swell and the seed coat to split or

break Enzymatic activity within the seed accelerates increasing

respiration and use of stored energy resulting in the commencement

of growth processes within the seed (Bewley and Black 1994

14

Pretreatment requirements for germination of alder seed are

quite variable both between and within species For many species of

alder cold stratification periods of60-180 days are recommended

(Dirr and Heuser 1987) In one study ofthinleaf alder prechilling

(stratification) did not improve germination percentage while in

European speckled alder 180 days of stratification did improve

percentage germination (Young and Young 1992) Several other

treatments including light freezing and potassium nitrate

independently and with stratification have been shown to enhance

germination ofalders In red alder stratification was not necessary

when seed was germinated in light (Kenady 1978 Radwan and

DeBell 1981) Evidence of a phytochrome-regulated dormancy was shy

subsequently found in this species (Bormann 1983) Several general

horticultural texts recommend a pretreatment with 0200 potassium

nitrate (wv) to enhance stratification effects (Hartmann et al 1997

Young and Young 1992) In one study stratification followed by

freezing of seed for 3 days at -20degC enhanced germination (Schalin

1968)

16

Water Birch

Birch species are widely distributed in the northern hemisphere

found further north than alders can grow in various habitats and are

tolerant of a wide range of soils and moisture levels but are sensitive

to drought (Ashburner 1993 deJong 1993) Birch species are thought

to be more resistant to drought than alder species (McVean 1956) B

ocddentaJis Hook occurs as a shrub or small tree along streams or in

moist canyons and occasionally in dryer sites of the mountain West

( at elevations of 1500-2700 meters (Foxx and Hoard 1995 Vines

1960) It is known in the vernacular as water birch red birch and

black birch A small tree it is not used for lumber but can be used as

firewood posts browse by livestock or wildlife and sometimes as a

landscape tree (BrenzeI1995 Elias 1980 Preston 1968 Vines 1960)

Germination requirements for species of Betula generally

include stratification or red light treatment (Brinkman 1974)

indicating the presence ofphytochrome far-red inhibition (Bevington

1986 Bevington and Hoyle 1981 Schopmeyer 1974) Occasionally

both red light and stratification are recommended to improve

germination rate (Dirr and Heuser 1987) Potassium nitrate 02

17

pretreatment is recommended for birch species by Hartmann et al

(1997) Seeds of this species are considered to have a fairly shallow

dormancy (Lane 1993)

Seed Quality Improvements

Methods to upgrade seed quality (separate viable from nonshy

viable seeds) have been developed for different species Conventional

seed separation techniques are based on density such as air column or

liquidseparation or by size and shape such as with screens

Separation ofviable and non-viable seeds is extremely problematic

with very light winged seeds like those of alder and birch Air

separation techniques may not be practical for winged light-weight

seed Flotation techniques often employ lighter-than-water solvents

but some of these substances may have adverse effects on seed

viability (Barnett 1971 McLemore 1965) Widescale use of some

solvents is not considered desirable because of health and safety

concerns

A method of seed refinementupgrade originally developed in

Sweden by Milan Simak called the LDS method (Incubation

Drying Separation) shows promise for separating live and dead seeds

18

(cited in Bonner 1984 Downie and Wang 1992 Simak 1983

Sweeney et al 1991) Seeds are imbibed for several hours then

incubated at cool temperatures (15~or several hours in 100

relative humidity Seeds are then dried for several hours at 35

relative humidity at cool temperatures (timing and relative humidity

must be adjusted for the particular species) During the drying

dead seeds will lose most of the water previously imbibed while live

seeds should retain most of their imbibed water This differential

moisture content would make separation by flotation and other

density separation methods potentially feasible Similar methods of

conditioning have been shown to improve seed quality in lettuce

tomato and onion (Hill et al 1989) It has also been shown that

drying of stratified seeds for storage or for separation from

stratification medium need not result in loss of viability (Danielson

and Tanaka 1978 Schopmeyer 1974)

19

OBJECTIVES OF THIS STUDY

The purpose of this study is to determine the effectiveness of the

LDS seed refinement technique and othi separation procedures in

increasing the percentage of live seeds in a seed lot and to develop

germination strategies for water birch and thinleaf alder investigating

the use of stratification Secondly this study will examine the within-

species variability of different seed lots in their response to LDS and

stratification treatments

METHODS AND MATERIALS

Sources

Alder strobiles were collected in October and November of

1998 in Catron County New Mexico near the towns of Luna and

Reserve in the Cottonwood Canyon Campground and in the Head of

the Ditch Campground and in Taos County New Mexico in the

Red River Canyon near the Molycorp molybdenum mine Table 1

shows the seed source elevations and locations Strobiles were kept

cool and allowed to dry for several weeks Seeds were separated from

the opening strobiles by rubbing on a coarse screen

Birch strobiles were collected in October and N overrtber of 1998

in Taos County in the Red River Canyon near the Mo1ycorp

molybdenum mine (Table 1) Strobiles were kept cool and allowed to

dry for several weeks allowing the release of seeds from the bracts

In addition commercial seed sources ofbirch and alder were

purchased in the summer of 1999 (collected in the fall of 1998) The

seed lots collected in 1998 (Table 1) were used in the seed refinement

study providing four seed lots for that study For the final seed

refinement-germination study the two Red River Canyon seed lots of

21

Table 1 Seed Source Locations and Elevations

Species Source Lot Baseline Description Elevation Latitude Notes No Fill (meters) Longitude

Thinleaf Alder Luna NA 234 Head ofthe Ditch CG 2134 N 33deg49 W 108deg59

t+

Reserve NA 268 Cottonwood Canyon 1829 N 33deg37 W 108deg55

t+

RRC-l 98108 08 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

RRC-2 98109 09 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 537 W of Poncha Springs CO 2438 N 38deg31 W 106deg05

I

Water Birch RRC-3 98104 69 Red River Canyon 2469 N 36deg41 W 105deg29

t+sect

Moly-1 98106 39 Molycorp Tailings Rd 2469 N 36deg41 W 105deg29

t+

Moly-2 98107 52 Molycorp Low Dump 2469 N 36deg41 W 105deg29

t+

Mo1y-3 98105 56 Molycorp Front Dump 2469 N 36deg41 W 105deg29

t+sect

Chaffee NA 297 W ofPoncha Springs CO

t used in Experiment 1--Seed Refinement I used in Experiment 2--Germination Requirements sectpooled for use in Experiment 2-Germination Requirements

2377 N 38deg31 W 106deg05

I

alder and the Red River Canyon and Moly 3 seed lots ofbirch were )

pooled due to limited amounts of seed The seed lots collected in

1998 and the 1998 purchased seed lots from Chaffee County

Colorado provided four seed lots of each species for that study

All seed sources were evaluated for percentage of filled seeds by

means ofdissection performed under a dissecting microscope at 30X

magnification (Berry and Torrey 1985) Alder species baseline

percentage of filled seeds was estimated using 25 samples of 100 seeds

pooled into one percentage response for each seed source Birch

species baseline percentage of filled seeds was determined using 15

samples of 50 seeds pooled into one percentage response for each seed

source Baseline percentage fill (Table 1) is the estimate of the

percentage of filled seed in the entire seed collection for each source

Separation Media

Ethanol and water were not particularly effective in separation

ofthinleaf alder seeds either using IDS methods or when separating

dry seed It was necessary to choose a fluid with a lower specific

gravity than ethanol (SG=O 79) in order to separate filled and empty

seeds with very low densities Falleri and Pacella (1997) found that

23

low-density London plane tree (Platanus x acerifolia [Aid Willd)

seeds could not be separated using water as the separation medium

due to the very small density differences between sound and empty

seeds and chose petroleum ether as a separation medium Petroleum

ether was chosen for the separation of thinleaf alder seeds because of

its low specific gravity (SG middot060) its relative stability low

reactivity and rating as a slight health risk Contact with skin may

cause dryness and irritation but no chronic systematic effects have

been reported with industrial use (Mallinckrodt Baker Inc 1997a)

As observed previously for thinleaf alder seeds the simple

specific gravity method using water was not effective for separating

water birch seeds In preliminary studies ethanol and petroleum

ether were found to be effective in separation of dry water birch seeds

and petroleum ether ethanol and water were somewhat effective in

separation of water birch seeds treated by the LDS method but

ethanol was chosen as the separation medium because of its lower

cost greater effectiveness and availability

Denatured ethanol is actually rated a greater health risk than

petroleum ether because ingestion is more likely to result in death or

24

permanent damage and prolonged skin contact may affect the

nervous system and other organ systems of the body Ethanol also

has a higher reactivity rating Gloves goggles and lab coat (personal

protective equipment) proper ventilation avoidance of ingestion and

proper fire safety measures should prevent problems with use of either

solvent (Mallinckrodt Baker Inc 1997a 1997b)

Seed Refinement

Thinleaf Alder

Separation treatments examined includeddensity separation of

dry seed samples in petroleum ettter (the control) and imbibed seed ~

samples treated with the IDS method at 0 1 18 and 24 hour drying

times followed by density separation in petroleum ether (Table 2)

Seeds were imbibed for 24 hours by submersion in a 10-gallon glass

aquarium filled with distilled water and equipped with an aeration

pump and filter Seeds were packaged in filter paper then the

packages were enclosed in wire cages (purchased tea balls were used

for this purpose) weighted with marbles to keep them submerged At

the end of the imbibition period seeds were removed from the cages

thoroughly blotted and placed on clean filter paper The drying

25

incubation was performed in a closed chamber with a constant

humidity obtained by the use ofCaC12middot6H20 salt in a saturated

solution prepared by adding SOOOg CaClzmiddot6HzO to 30 liters of

distilled water (Slavik 1974 Young 1967) Imbibed seeds were placed

on filter paper and suspended on a screen above the calcium chloride

solution Humidity was monitored using an hygrometer and held

steady at 50 in the presence of the wet seeds and filter paper

Table 2 Alder Preparation Protocols for S~d Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) 0 0

2

3

24

24

0

1 )

4 24 18

5 24 24

At the end of the appropriate drying incubation the seeds were

placed in petroleum ether and briefly and vigorously stirred to

separate seeds adhering to one another Floating seeds were removed

from the surface of the petroleum ether by means of a small net

andor a spatula placed on clean moistened filter paper and placed in

26

a labeled plastic bag to await counting The sinking seeds were

strained through the net and packaged in a similar manner Five

repetitions were performed for each of the five treatments using 100

seeds per repetition Percentage of filled seeds contained in each

fraction was determined by means of dissection tests performed on the

floating and sinking fractions using a scalpel and a dissecting

microscope with 30X magnification

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that particular repetition

of filled seeds in the sinking fraction X 100=percentage recovery

of filled seeds in the sinking fraction + of filled seeds in the floating fraction

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product of percentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

27

Water Birch

Separation treatments included density separation of dry seed in

95 ethanol (the control) and seed samples treated with the IDS

method at 005 1 and 2 hour drying times followed by density

separation in 95 ethanol (Table 3) Seeds were imbibed for 12 hours

by submersion in a 10-gallon glass aquarium filled with distilled water

and equipped with an aeration pump and filter Seeds were packaged

in filter paper then the packages were enclosed in wire cages

(purchased tea balls were used for this purpose) weighted with

marbles to keep them submerged At the end of the imbibition

period seeds were removed from the cages thoroughly blotted and

placed on clean filter paper The drying incubation was performed in

a closed chamber with a constant humidity obtained by the use of

CaCI2middot6H20 salt in a saturated solution prepared as described in the

previous section (Slavik 1974 Young 1967) Imbibed seeds were

placed on filter paper and suspended on a screen above the calcium

chloride solution Humidity was monitored using an hygrometer and

held steady at 50 in the presence of the wet seeds and filter paper

28

Table 3 Birch Preparation Protocols for Seed Refinement

Preparation Protocol Imbibition Time (Hours) Drying Time (Hours)

1- (Control) o o

2 12 o

3 12 05

4 12 1

5 12 2

At the end of the appropriate drying incubation the seeds were

placed in 9500 ethanol and briefly and vigorously stirred to separate

seeds adhering to one another Floating seeds were removed from the

surface of the ethanol by means of a small net andor a spatula

placed on clean moistened filter paper and placed in a labeled plastic

bag to await counting The sinking seeds were strained through the

net and packaged in a similar manner Three repetitions were

performed for each of the five treatments using 50 seeds per

repetition Percentage of filled seeds contained in each fraction was

determined by means of dissection tests performed on the floating and

sinking fractions using a scalpel and a dissecting microscope with

30X magnification

29

In addition the percentage recovery of filled seeds from the

sinking fraction was calculated based on the total number of filled

seeds present in that repetition (as given in the previous equation)

The most effective drying-incubation time combination was chosen

for use in the germination testing portion of the study on the basis of

the largest product ofpercentage of filled seeds in the sinking fraction

multiplied by percentage recovery of filled seed from the sinking

fraction

Germination Requirements

Thinleaf Alder

Separations were performed using the separation method

chosen from the seed refinement study alder preparation protocol 4

24-hour imbibition followed by I8-hour drying time and separation in

petroleum ether as described in the seed refinement experiment

(Table 2) Unseparated imbibed seeds and seeds from both the

floating and sinking fractions were subsequently treated with 028

and 56 days of stratification In addition an unseparated nonshy

stratified control of dry seeds was tested for germination Seeds for

stratification treatments were placed in layers ofpaper towel

30

moistened with 25 m1 of distilled water and placed in sealed zip-lock

plastic bags The bags were placed in a cooler at I-5degC (average

temperature 50degC) for periods of 28 or 56 days Initiation of

stratification treatments was staggered so that all treatments came out

ofstratification at the same time

Following stratification the seeds were sown in Ray-Leach

Super Cells (Steuwe amp Sons Inc Corvalis OR) containing a 2 1 1

ratio ofpeatmossperlitevermiculite (vvv) with OsmocoteR 14-14-10

slow release fertilizer at a rate of 4007 gm3bull Five seeds were sown

per tube Treatments were distributed in a randomized complete

block design consisting of4 blocks (locations on the greenhouse

bench) with each block containing the 10 treatment combinations for

each of four seed lots (Table 4) Each repetition contained 20 tubes

repetitions were placed in random order four repetitions to a rack ten

racks to each block Each repetition for each treatment contained 100

seeds therefore 100 seeds were used for each seed source by

treatment by block combination pooled to one measurement for the

response variable germination percentage Racks were placed in a

greenhouse for germination Germination conditions included

31

ambient light and 70 relative humidity with average daily

temperature 243degC (daytime temperature range 200-272degC) and

average night temperature 216degC (nighttime temperature range 206shy

239degC) Tubes were watered at 2 hour intervals six times a day

Germination was recorded at weekly intervals 7 1421 and 28 days

after planting

Table 4 Treatment Combinations for Experimental Layout of Randomized Complete Block--Thinleaf Alder

Treatment Number Stratification (Days) Separation Fraction

1 56 None

2 56 Floating

3 56 Sinking

4 28 None

5 28 Floating

6 28 Sinking

7 0 None

8 0 Floating

9 0 Sinking

blocks Qocations on the greenhouse bench) with each block

containing the 10 treatment combinations for each of four seed lots

(Table 5) Each repetition contained 20 tubes repetitions were placed

in random order four repetitions to a rack ten racks to each block

Each repetition for each treatment contained 100 seeds therefore 100

seeds were used for each seed source by treatment by block

combination pooled to one measurement for the response variable

germination percentage Racks were placed in a greenhouse for

germination Germination conditions included ambient light and

7000 relative humidity with average daily temperature 243 degC

(daytime temperature range 200-272degC) and average night

temperature 216degC (nighttime temperature range 206-239degC)

Tubes were watered at 2 hour intervals six times a day Germination

was recorded at weekly intervals 7 1421 and 28 days after planting

34

Table 5 Treatment Combinations for Experimental Layout of Randomized Complete Block-Water Birch

Treatment Number Stratification (Days) Separation Fraction

56 None

2 56 Floating

3 56 Sinking

4 21 None

5 21 Floating

6 21 Sinking

7 0 None

8 0 Floating

9 0 Sinking

10 0 None

DATA ANALYSIS

The seed refinement experiment was performed to determine

the mostadvantageous separation technique for use in the

germination studies with the percentage of filled seeds present in the

sinking fractions (percentage fill) and proportion of filled seeds

recovered from the total filled seeds available in the sample

(percentage recovery) as response variables and the preparation

protocols and seed sources as independent variables

The second experiment utilized the chosen seed refinement

method with levels of stratification seed separation fraction and seed

source as independent variables (or in the augmented factorial

treatment combination as the independent variable) with germination

percentage measured as the response variable Germination rate was

also recorded however the rapid germination between the time of

sowing and the first sampling (at 7 days) prevented meaningful

analysis of this da~

Data was analyzed by using categorical data modeling analysis

as found in the SAScopy statistical program The PROC CATMOD

procedure can perform analysis and giveanalysis of variance in the

36

general sense that it analyzes the response functions fits linear models

to functions of response frequencies and partitions the variation

among those functions into various sources (SAS Institute 1989)

CATMOD analyzes data that can be represented in a two-

dimensional contingency table with the rows corresponding to

populations or samples defined by one or more independent variables

and the columns corresponding to one or more dependent (response)

variables The frequencies in the table are assumed to follow a

product multinomial distribution with a simple random sample taken

for each population The probability for the response ofeach cell is

estimated and the vector (P) of these proportions is transformed into a

vector of functions F =F(P) If It denotes the vector of true

probabilities for the table then the functions of the true probabilities

F(It) are assumed to follow a linear model

I

where EA denotes asymptotic expectation X is the design matrix

containing fixed constants and Pis a vector ofparameters to be

37

estimated CA TMOD provides two estimation methods the

maximum-likelihood method and the weighted-least-squares method

which was used in this analysis (SAS Institute 1989)

Hypotheses about linear combinations of the parameters can be

tested these statistics are approximately distributed as chi-square for

sufficiently large sample sizes (SAS Institute 1989)

All of the response variables considered had a binomial type of

probability distribution (seed filled or not filled seed germinated or

not germinated) All treatments ofboth experiments were analyzed

using the PROC CATMOD procedure to examine the general model

as well as planned comparisons using contrast statements where ~

appropriate The PROC MEANS procedure was used to calculate

marginal percentages (main effect and interaction combinations)

along with standard errors Pairwise Z-tests were used to separate

percentages in those effects which were determined to be significant

by categorical modeling at the observed significance level laquo=005 this

method of percentage separation is analogous to Fishers LSD for

separating means

38

Categorical modeling of the seed refinement experiments used

two models one for the percentage of filled seeds attained in the

sinking and floating fractions and one for the percentage of filled

seeds recovered from those available in the baseline sample The

treatment structure for both of these seed refinement studies was a 5

X 2 X 4 factorial (preparation protocol by separation fraction by seed

source)

The model in CATMOD for the percentage of filled seeds

attained in the fractions is as follows

fill = source prep fraction sourceprep sourcefraction prepfraction sourceprepfraction

where fill is the response variable the number of filled seeds

source is the seed source prep is the LDS treatment protocol

(imbibition plus one of four drying times or no treatment) and

fraction is the separation fraction (floating or sinking)

The CATMOD model for the percentage of filled seed

recovered in the floating or sinking fractions is as follows

39

rec = source prep sourceprep

where rec is the response variable (number of filled seeds floating or

sinking) source is the seed source and prep is the IDS

treatment protocol

The treatment structure for the germination studies was a 3 X 3

factorial (stratification by separation) with an additional control (no

treatment) for 4 seed sources termed an augmented factorial design

by Lentner and Bishop (1986) The PROC CATMOD procedure

was used to analyze the data without the no treatment control as a

simple 3 X 3 X 4 factorial (stratification by separation by seed source)

with the model

germ = strat sep stratsep source stratsource sepsource stratsepsource

where germ is the response variable strat is the level of

stratification sep is the IDS fraction and source is the seed

source

40

The augmented factorial design (3 X 3 factorial plus a noshy

treatment control for 4 seed sources) was also analyzed by PROe

CATMOD in order to evaluate the effect of imbibition The

treatment structure for the germination studies in this analysis was a

lOX 4 factorial (treatment by seed source) with the model in

CATMOD as follows

germ = treat source treatsource

where germ is the response variable (germinated or not) treat is

the treatment combination (LDS fraction plus level ofstratification)

and source is the seed source

RESULTS

Seed Refinement

Thinleaf Alder Fill Enhancement

Preparation protocol seed source and the separation fraction

had significant (alpha=O05) effect on the percentage fill (Table 6)

The effect of separation fraction was influenced by both source and

preparation protocol

Table 6 Analysis ofVariance Table for Thinleaf Alder Percentage of Filled Seeds as Influenced by Preparation Protocol Separation Fraction and Seed Source-Factorial

Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 11663 00000

Seed Source 3 17367 00000

Preparation Protocol 4 4490 00000

Separation Fraction 1 8829 00000

SourcePrep 12 541 09427

SourceFraction 3 971 00212

PrepFraction 4 986 00429

SourcePrepFraction 11 714 07878

Protocols 1 4 and 5 the control and 24 hour imbibition

followed by either 18 or 24 hours drying respectively all had greater

than 8000 filled seed in the sinking fraction (Table 2 Figure 1)

Twenty-four hour imbibition alone or in conjunction with 1 hour of

drying both had lower percentages of filled seeds in the sinking

fraction (less than 3500) Protocol 4 the 24-hour imbibition followed

by 18 hours of drying and density separation in petroleum ether was

chosen as the separation method for the germination requirements

study

The proportion of filled seed in the sinking and floating

fractions was also influenced by seed source Percentage of filled

seeds in the sinking fraction ranged from 444 for the Red River

Canyon 1 source to over 86 for the Luna source (Table 7)

Percentage of filled seeds in the floating fraction ranged from less than

1 to just over 1200 while the baseline percentage of filled seeds in

the seed sources ranged from less than 100 to over 26 The

separation process improved percentage fill in the sinking fraction

compared to the percentage fill in the floating fraction by about sevenshy

43

bull bull bull bull bull

100~------------------------------------------~

80

60El ~

~ ~ 40s ~ 1-4 0 ~

20

0

used to represent the percentage (protocols are described in Table 2)

44

_ Floating Fraction -0- Sinking Fraction

1- (010) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 1 Alder Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent +- one Standard Error Bars which are not visible are smaller than the symbol

Table 7 Thlnleaf Alder Percentage ofFilled Seeds in the Fractions as Influenced by Source and Compared to Baseline Fill Uninfluenced

by Preparation Protocol

Seed Baseline Fill- SE Fill- SE n Source Fill Sinking Fraction Floating Fraction

Luna 234 8634c 180 1265d 072 4000

Reserve 268 4644b 139 631c 070 4000

RRC-1 08 444a 131 O44a 014 4000

RRC-2 09 909a 328 062a 0)5

Percentages followed by the same letter are not significandy different at laquo=0 5

fold for the Luna and Reserve seed sources ten-fold for the Red River

Canyon 1 source and almost fifteen-fold for the Red River Canyon

2 source Separation improved the percentage of filled seeds in the

sinking fraction compared to the unseparated seed source by almost

four-fold for the Luna source almost two-fold for the Reserve source

almost six-fold for the Red River Canyon 1 source and ten-fold for

the Red River Canyon 2 source

Floating separation fractions had a much lower percentage of

filled seeds (464) than sinking fractions (471100) (Table 8)

Percentage of filled seeds was consistently low in the floating fraction

but varied with the preparation protocol in the sinking fraction

(Figure I)

45

Table 8 Thinleaf Alder Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SE 11

Floating Fraction 463a 023 8030 (Sinking Fraction 47l1b 112 1970

Percentages followed by the same letter are not significantly different at laquo=005

Thinleaf Alder Recovery

Seed source and preparation protocol both influenced the

percentage of filled seeds recovered (Table 9) In contrast to the

percentage of filled seeds in the sinking fraction (Figure 1) the

percentage of seeds recovered was improved by 24 hours imbibition

alone or with one hour drying at 5000 humidity (Table 10) These two

treatments had in excess of 80 recovery whereas the other three

separation treatments all averaged less than 6700 recovery

Table 9 Analysis ofVariance Table for Thinleaf Alder Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 694 00084

Seed Source 3 11055 00000

Preparation Protocol 4 2371 00001

SourcePrep 12 1715 01439

Percentage recovery as influenced by source varied from

approximately 3200 for the Red River Canyon 2 collection to over

88 for the Reserve seed source (Table 11) The Red River Canyon

47

1and Luna sources both had percentage recoveries slightly greater

than 50

Table 10 Thlnleaf Alder Percentage ofFilled Seeds Recovered in the Sinking Fraction as Influenced by

Preparation Protocol

Protocol (SoakDry) Recovery SE n

1- 010 6447a 290 273

2 - 240 8094b 236 278

3 - 241 8225b 230 276

4-2418 6667a 312 228

5 -2424 6041 a 312 245

Percentages followed by the same letter are not significantly different at cx=005

Table 11 Thlnleaf Alder Percentage of Filled Seeds Recovered in the Sinking Fraction as Influenced by Seed Source

Seed Source Recovery SE n

Luna 5392b 206 586

Reserve 8852c 123 671

RRC-l 5238ab 1090 21

RRC-2 3182a 993 22

Percentages followed by the same letter are not significantly differerit at a=005

48

Water Birch Fill Enhancement

The preparation protocol and the separation fraction influenced

the percentage of filled seed in the fractions Seed source did not

impact the percentage of filled seed in the fractions The effect of

separation fraction on percentage of filled seed in the fractions was

influenced by both seed source and preparation protocol

independently (Table 12)

Table 12 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds as Influenced by Preparation Protocol Separation

Fraction and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 47402 00000

Seed Source 3 603 01103

Preparation Protocol 4 956 00486

Separation Fraction 1 8190 00000

SourcePrep 12 1204 04423

SourceFraction 3 2223 00001

PrepFraction 4 1304 00111

SourcePrepFraction 12 1453 02680

Overall the sinking fraction had higher percentage of filled seed

than the floating fraction (Table 13) The influence ofpreparation

protocol on the percentage of filled seed in the two fractions is

illustrated in Figure 2 All four of the LDS treatments reduced the

percentage of filled seeds in the sinking fraction relative to the nonshy

imbibed control treatment (Table 3 Figure 2) The percentage of

filled seeds in the sinking fraction ranged from near 900 to 12 for the

imbibed treatments whereas the percentage in the control treatment

was over three times these amounts The floating fractions had

consistently low percentage of filled seeds while the sinking fraction

treated by protocol 1 (separation without imbibition) had a higher

percentage of filled seeds than the sinking fractions treated by the

other protocols (all with imbibition) The control treatment (protocol

1) was chosen as the separation protocol for the germination

requirements study

The influence ofseed source on the percentage of filled seed in

the two fractions is illustrated in Figure 3 The Moly 2 and Moly 3

seed sources had much higher percentage of filled seeds in the sinking

50

Table 13 Water Birch Percentage ofFilled Seeds as Influenced by Separation Fraction

Separation Fraction Fill SB n

Floating Fraction 177a 030 1867

Sinking Fraction 1165b 095 1133

Percentages followed by the same letter are not significantly different at ct=O05

fraction than the Red River Canyon 3 and the Moly 1 sources All

of the floating fractions had a low percentage of filled seeds

bull bull bull

50--------------------------------------------

40

30S It ltU

$ 20I ltU ~ ltU

tl-i 10

0

-e- Floating Fraction -0- Sinking Fraction

t-----

1- (00) 2- (120) 3- (1205) 4- (1211) 5- (1212)

Preparation Protocol

Figure 2 Birch Percentage Fill as Influenced by Preparation Protocol and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage (protocols are described in Table 3)

52

bull bull

60-------------------------------------------~

50

~ ~

~ 5 t

p

40

30

20

10

o

__ Floating Fraction -0- Sinking Fraction

RRC3 Molyl Moly2 Moly3

Seed Source

Figure 3 Birch Percentage Fill as Influenced by Seed Source and Separation Fraction Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

53

Water Birch Recovery

Preparation protocol and seed source did not influence the

percentage of sound seeds recovered in the sinking fraction (Table 14)

Percentage of filled seeds recovered for the various protocols ranged

from 64-91 Percentage of filled seeds recovered in the sinking

fractions for the 4 seed sources ranged from 70-89

Table 14 Analysis ofVariance Table for Water Birch Percentage ofFilled Seeds Recovered in the Sinking and Floating Fractions as Influenced by Preparation Protocol

and Seed Source--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 3378 00000

Seed Source 3 487 01816

Preparation Protocol 4 217 07040

SourcePrep 12 1831 01066

Table 15 Thinleaf Alder Percentage Germination Analysis ofVariance--Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 108035 00000

Stratification 2 008 09616

Separation Fraction 2 22684 00000

Strat Sep 4 362 04595

Seed Source 3 36042 00000

Strat Source 6 1004 01230

Sep Source 6 10961 00000

Strat Sep Source 12 2604 00106

Table 16 Thinleaf Alder Percentage Germination as Influenced by Source--Factorial Analysis

Source Percentage Germination SB n

Luna 2011b 067 3600

Reserve 1914b 066 3600

RRCpoo1 075a 014 3600

Chaffee 3283c 078 3600

Percentages followed by the same letter are not significantly different at =005

56

Table 17 Thlnleaf Alder Percentage Gennination as Influenced by Separation-Factorial Analysis

Separation Fraction Percentage Germination SE n

No Separation 1960b 057 4800

Floating Fraction 292a 024 4800

Sinking Fraction 32l0c 067 4800

Percentages followed by the same letter are not significantly different at a==005

unseparated controls however regardless of separation fraction or

stratification treatment overall germination was quite low ( lt 100) in

this seed source (Table 16)

The influence of stratification was variable across all four seed

sources and the separation fractions evaluated (Figure 5) For

example only in the sinking fraction of the Chaffee source did

germination continue to increase with increasing stratification

duration (Figure 5d) In several combinations of separation protocol

and seed source the mid-level (28 days) of stratification had the

highest germination In the case of the sinking fraction of the Reserve

source this level actually reduced the germination response (Figure )

5b) Overall the low germinating fractions (Red River Canyon and

57

60

___ Luna 50

t 400-a

~

5 E 30 d v 00 ~ 20 t v U M V

10p

0

-0- Reserve -T shy RRC

~ -v- Chaffee

Q

~

Nosep Floating Sinking

Separation Fraction

Figure 4 Alder Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

58

M-------------______~ ~-------------------~

21 21

Stratification Period (Days) Stratification Period (Days)

Figure 5a Luna Source Figure Sb Reserve Source

M-------------------~ ~-------------------~

212S

Stratification Period (Days) Stratification Period (Days)

Figure 5c RRC Source Figure 5d Chaffee Source I

-- No Separation --0- Floating Fraction -- Sinking Fraction

Figure 5 Alder Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

59

the floating fractions of the other three seed sources) were slightly

impacted by stratification duration

In the analysis of the augmented factorial stratification and

separation fraction treatments were grouped to consider the effect of

imbibition (Table 18) Treatment 10 (the non-imbibed control) was

compared to the imbibed O-day stratification treatments (789) using

planned comparisons by contrast to determine ifpre-soaking of the

seeds had an effect Seed sources were also compared using planned

comparisons by contrast (Table 19) Overall pre-soaking (imbibing)

did not intluence germination however individual sources varied in

- response to this procedure (Tables 18 and 19 Figure 6) Pre-soaking

did not influence the percentage germination of the Red River

Canyon pool or the Chaffee seed sources (Figure 6) However preshy

soaking increased percentage germination in the Luna seed source

and decreased the percentage germination in the Reserve seed source

In order to detect differences in seed source in response to

treatment the Luna source and the Reserve source from the same

general geographical area were compared to each other and to the

Red River Canyon source and the Chaffee source The Luna and

60

Table 18 Analysis ofVariance Table for Thinleaf Alder Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source OF Chi-Square Observed Significance

Level

Intercept 1 123226 00000

Treatment 9 28267 00000

Seed Source 3 41066 00000

TreatSource 27 19476 00000

Table 19 Thinleaf Alder Analysis of Contrasts-Augmented Factorial

Contrast OF Chi-Square Observed Significance

Level

Treatment 10 vs 7+8+9t 1 230 01292

Treatment 10 vs 3 1549 00014 7+8+9Seed Sourcet

Luna + Reserve vs RRCsect 1 14117 00000

Luna + Reserve vs Chaffeesect 1 20161 00000

Luna vs Reservesect 022 06368

tOegrees of freedom by Treatment tDegrees of freedom from Source Main Effect sectOegrees offreedom by Source

61

40

35

30Q00tl CIS

25o~ d 20 Q) t)I)

S Q 15Q)

~ Q)

~ 10

5

0

_ Non-Imbibed _Imbibed

Luna Reserve RRC Chaffee

Seed Source

Figure 6 Alder Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent + - one Standard Error

62

Reserve sources were not different from each other but these sources

differed from the Red River Canyon source and the Chaffee source

(Table 19)

It was noted that the Red River Canyon seed source had

uniformly poor germination which might affect the interpretation of

the analysis (Figures 4 Sc) The analysis was repeated deleting this

source (Table 20) With the elimination of the Red River Canyon

seed source stratification effect becomes significant but the

differences are not dramatic (Table 21) There was no interaction

between stratification and separation or stratification and source the

other effects and interactions are similar to those detected in the

analysis using all sources

r

Table 20 Thinleaf Alder Analysis ofVariance--Factorial Analysis without Red River Canyon Seed Source

Source DF Chi-Square Observed Significance

Level

Intercept 1 150244 00000

Stratification 2 936 00093

Separation Fraction 2 81026 00000

Strat Sep 4 918 00568

Seed Source 2 18498 00000

Strat Source 4 752 01108

Sep Source 4 10657 00000

Strat Sep Source 8 2225 00045

Table 21 Thinleaf Alder Percentage Germination as Influenced by Stratification without Red River Canyon Seed Source-Factorial

Analysis

Stratification Period Percentage Germination SE n

oDays 2167a 069 3600

28 Days 260Ob 073 3600

56 Days 2442b 072 3600

Percentages followed by the same letter are not significantly different at laquo=005

64

Water Birch

Total germination of imbibed water birch seed was influenced

by stratification separation fraction and seed source and by all

interactions of these three treatments (Table 22) Increasing

stratification length improved germination (Table 23) Seed in the

sinking fraction regardless of seed source or stratification had the

greatest total germination (Table 24 Figure 7) While stratification

regardless of duration improved germination overall seeds in the

sinking fraction responded best to the 56-day stratification treatment

(Table 23 Figure 8) The Chaffee seed source had the greatest total

germination followed by the Moly 2 and the Red River Canyon

seed sources while the Moly 1 source had a low germination

percentage (Table 25) Improvement in germination of the sinking

fraction relative to the unseparated control ranged from two-fold for

the Chaffee source to over ten-fold for the Moly 2 source (Figure 7)

The response of individual seed sources to stratification varied

considerably (Figure 9) The response to stratification was also not

consistent across seed sources and separation fractions (Figure 10)

The lack ofconsistent stratification effects is most apparent in the

65

Table 22 Water Birch Percentage Germination Analysis of Variance Table-Factorial Analysis

Source DF Chi-Square Observed Significance

Level

Intercept 1 74900 00000

Stratification 2 4503 00000

Separation Fraction 2 7719 00000

Strat Sep 4 2266 00001

Seed Source 2 2671 00000

Strat Source 4 9154 00000

Sep Source 4 7030 00000

Strat Source 8 13885 00000

Table 23 Water Birch Percentage Germination as Influenced by Stratification-Factorial Analysis

Stratification Period Percentage Germination SE n

oDays 1108a 045 4800

21 Days 1363b 050 4800

56 Days 1623c 053 4800

Percentages followed by the same letter are not significantly different at laquo=005

66

1

Table 24 Water Birch Percentage Germination as Influenced by Separation-Factorial Analysis

----------------~-------

Separation Fraction Percentage Gennination SE ----------------shy

n

No Separation 669b 057 4800

Floating Fraction 123a 024 4800

Sinking Fraction 3302c 067 4800

Percentages followed by the same letter are not significantly different at laquo=005

Table 25 Water Birch Percentage Gennination as Influenced by Seed Source-Factorial Analysis

Source Percentage Germination SE n

Moly-1 475a 034 3600

Moly-2 1503c 056 3600

RRCpool 1295b 053 3600

Chaffee 1855d 061 3600

Percentages followed by the same letter are not significantly different at laquo=005

67

---

50

40

I= 0

0 d 30

~ Q)

d 20

s ~ I=

~ 10 Q)

~

0

-e-- Moly 1 -0- Moly2 -T RRC -ry- Chaffee Atj

Itt ~

Iffjl

-shy-- J-~I

Nosep Floating

Separation Fraction

Sinking

Figure 7 Birch Percentage Germination as Influenced by Separation Fraction and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

68

40

50~--------------------------------------------~

-- shy _--i ____ Y- --- -shy

y---- ---shy___ No Separation

-0- Floating Fraction - - Sinking Fraction

bullbullbull _-0 0------()---shyo

o 21 56

Stratification Period (Days)

Figure 8 Birch Percentage Germination as Influenced by Stratification and Separation Fraction Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

69

r

35

30

25 0=

Q

5 ~

20 e ti Cl 15 ~ ~ = 10 ~ v ~

5

0

-+- Moly 1 -0- Moly2 -- RRC -V- Chaffee

o 21

Stratification Period (Days)

Figure 9 Birch Percentage Germination as Influenced by Stratification and Seed Source Error bars represent one + - Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

70

56

bullbull bull

80

Q 0 60 c

5 ~ D 40

t 20

~ p

J

8O-----------------~

0-- -0--0---shy

2

Stratification Period (Days)

Figure lOc RRC Source

t=

21 56

Stratification Period (Days)

Figure lOa Moly 1 Source

100--------------- ~

I

2

Stratification Period (Days)

Figure lOb Moly 2 Source

100--------------

Stratification Period (Days)

Figure lOd Chaffee Source

- No Separation -0- Floating Fraction -I- Sinking Fraction

Figure 10 Birch Percentage Germination as Influenced by Separation Fraction Seed Source and Stratification Error bars represent + - one Standard Error Bars which are not visible are smaller than the symbol used to represent the percentage

71

variable responses in the sinking fraction among seed sources In the

Moly 1 source 21 days ofstratification had no impact while 56 days

had a 7-fold improvement in germination (Figure lOa) However in

the Moly 2 source 21 days of stratification yielded the highest

germination and continued stratification reduced total germination

(Figure lOb) A similar but less pronounced trend was seen in the

germination ofseeds in the sinking fraction of the Red River Canyon

source (Figure IDe) where 21 days of stratification increased

germination of the sinking fraction but 56 days ofstratification

reduced germination below the level ofno stratification

Stratification regardless of duration reduced germination in seeds

from the sinking fraction of the Chaffee source however the decrease

was not linear with stratification length (Figure lOd) The floating

fraction and unseparated seed were relatively unaffected by

stratification treatment

Treatment (combination of stratification and separation

fraction) and source were considered in the analysis of the augmented

factorial in order to evaluate the effect ofpre-soaking by the use of

planned contrasts (Tables 26 27) Pre-soaking (the contrast of

72

Table 26 Analysis ofVariance Table for Water Birch Percentage Germination as Influenced by Treatment

Combination and Seed Source-Augmented Factorial

Source DF Chi-Square Observed Significance

Level

Intercept 1 88622 00000

Treatment 9 101947 00000

Seed Source 3 2994 00000

TreatSource 27 34727 00000

Table 27 Water Birch Analysis of Contrasts--Augmented Factorial

Contrast DF Chi-Square Observed Significance

Level

Treatment 10 vs 7+ 8+9t 1 1777 00000

Treatment 10 3 3280 00000 vs 7+8+9 Seed Sourcet

tDegrees of freedom by Treatment tDegrees offreedom from Seed Source Main Effect

73

Treatment 10 with Treatments 7 8 and 9) influenced germination

regardless of source but the seed sources varied in their response to

irnbibition (the contrast of Treatment 10 with the average of

Treatments 7 8 and 9seed source) Imbibition improved

germination for the Red River Canyon and Chaffee seed sources

Germination of the Moly 1 source was reduced by imbibition while

Moly 2 imbibed seeds had no germination (Figure 11)

35

30

c 250-Q

Cd

-~ 20 Q)

d ~ 15Cd c ~ Q) 10p

5

0

I Non-Imbibed _Imbibed

Moly 1 Moly 2 RRC Chaffee

Seed Source

Figure 11 Birch Percentage Germination as Influenced by Imbibition and Seed Source Error bars represent one + -Standard Error

75

DISCUSSION

Seed Refinement

Traditionally seed refinement has been thought of as enhancing

the number ofpotentiaily viable seeds (filled seeds) in a seed lot

Previously published studies have used total germination as the

measure of seed refinement efficacy In this study the number of

filled seeds in the sinking fraction was used The LDS treatments

imposed did not improve the number of filled seeds in the sinking

fraction in comparison with ordinary gravity separation for either of

the species evaluated in this study In two of the alder LDS

treatment levels 24-hour soak with either no drying time or one hour

of drying time actually reduced the percentage of filled seeds in the

sinking fraction The two remaining alder ID S treatments had

considerably longer drying times and resulted in percentages of filled

seeds in the sinking fraction similar to those of the non-imbibed

control treated by gravity separation The influence of drying time on

the efficacy of the LDS treatment has been seen in other species

(Faileri and Pacella 1997 Sweeney et al 1991) In a study ofLondon

plane tree researchers found that as drying time increased from 75

76

I

hours to 24 hours observed germination percentage was greater than

control (Falleri and Pacella 1997) At drying times less than 75

hours observed germination was comparable to unseparated controls

In the same study only seed receiving 24 hours of drying as part of an

IDS treatment had greater germination than non-treated seed

separated in petroleum ether

The response of the alder seed to IDS indicates there may be

potential for IDS as a seed refinement tool using longer imbibition

and drying times The difference in times from the I-hour to the 18shy

hour drying is considerable and corresponds to a significant difference

in the percentage of filled seeds in the sinking fraction The shorter

drying times may have been of insufficient duration to allow the

unfilled seed to lose sufficient moisture and hence these seeds ended

up in the sinking fraction In contrast the 18- and 24-hour drying

times may have allowed the imbibed unfilled seeds to lose the

majority of the water imbibed and resulted in percentages of filled

seeds in the sinking fraction similar to those seen in the non-imbibed

controls

77

The percentage of filled seeds in the sinking fraction in response

to the ID S treatments used in the water birch experiment indicates

that drying times may have been too short to allow the empty seeds to

lose sufficient moisture This would result in an increase in the

percentage ofempty seeds in the sinking fraction The seeds ofwater

birch are similar to those of thinleaf alder both are borne in strobiles

and have winged integuments almost entirely surrounding the seed

The alder integument is rather leathery while the birch integument

appears thinner Water birch seeds are smaller and rounder as

opposed to the flattened shape of alder seeds (alder seeds averaged

1469gram birch seeds 12S4gram) The seeds may also differ in

their ability to lose water following imbibition The assumption was

made that the birch seeds being smaller with larger integuments

relative to the size of the seed would lose imbibed water at a faster

rate This may not have been the case as indicated by the higher

percentage of empty seeds in the sinking fraction

While all thinleaf alder sources had improved percentages of

filled seeds in the sinking fractions there appear to be differences

between sources in response to seed refinement This difference was

78

detailed studies examining source differences in the rate of moisture

loss would be beneficial

The above discussion focuses primarily on reducing the number

of empty or non-viable seeds in a seed lot During seed refinement

some viable seed is also lost in the floating fraction (Downie and

Wang 1992 Falleri and Pacella 1997 Sweeney et al 1991) In cases

where there is more than adequate seed supply the loss of viable seed

in the floating fraction is not a problem In those cases where the

amount of available viable seed is limited and losses of viable seeds

needs to be minimized other criteria can be used to determine the

most effective seed refinement technique Such was the case in this

study

The percentage of filled seeds recovered in the sinking fraction

provides a measure ofhow efficient the refinement technique is at

reducing the number of filled (potentially viable) seeds lost in the

floating fraction In the current study involving alder those protocols

with low percentages of filled seeds in the sinking fraction had a high

percentage of filled seeds recovered (Figure 12) In the case of alder

the high recovery of filled seeds was inversely related to the LDS

80

r

0

~ S Il Q) u M Q)

~

50

40

30

20

-e- Fill -0- Recovery

1- (00) 2- (240) 3- (241) 4- (2418) 5- (24124)

Preparation Protocol

Figure 12 Alder Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (Protocols described in Table 2)

81

~ 0 u ~ M 0

S ~ I+-lt

90

80

70

60

treatments ability to remove non-viable seed A similar trend was

observed in another study in an attempt to upgrade germinated

cabbage seeds using density gradients As percentage recovery

increased the percentage of germinated seeds decreased because of

the increased recovery of non-germinated seeds (Taylor and Kenny

1985) In the case ofwater birch separation technique did not

impact the percentage of filled seeds recovered but there were

differencesin the percentage of filled seeds in the sinking fraction

(Figure 13) The technique employed to determine which seed

refinement protocol to use in the germination studies was to multiply

the percentage of filled seeds in the sinking fraction by the percentage

of filled seeds recovered This value addresses both the protocols

ability to remove non-viable seeds as well as its ability to reduce the J

loss of potentially viable seeds

Depending on a growers constraints either greenhouse space

or seed supply the evaluation of a seed refinement technique could be

based on one of three criteria discussed above percentage of filled

seeds in the sinking fraction percentage of filled seeds recovered or

the product generated by multiplying these two values as was done in

82

-

120~------------------------------------------~

___ Fill

-0- Recovery

20

J O~------~----~------~------~------~----~

1- (010) 2- (120) 3- (1205) 4- (121) 5- (1212)

Preparation Protocol

Figure 13 Birch Percentage Fill and Percentage Recovery of the Sinking Fraction as Influenced by Preparation Protocol Error bars represent + - one Standard Error (protocols described in Table 3)

83

this case In cases where seed supply is a greater constraint selection

of seed refinement technique may be based solely on the percentage

of filled seeds recovered This seed refinement technique may not be

as efficient in removing unfilled seeds but loss of filled seeds would

be minimized In the case where growing space is the greater

constraint the percentage of filled seeds in the sinking fraction would

be the criteria used for seed refinement technique selection Ifboth

greenhouse space and seed supply are limited then the product of the

two may be used to determine the appropriate protocol The use of

this information in conjunction with spreadsheet-based seed sowing

programs allows nursery managers to select the best seed refinement

technique for their nursery (Harrington and Glass 1997 Wenny

1993)

The particular separation medium found to be most effective

will vary with species Large and dense seeds may often be effectively

separated using water as the medium (Simak 1983) This is known as

the specific gravity method of separation when used on untreated

seeds In very small seeds where the density gradient between

empty dead and filled live seeds is not great water may not be

84

effective and it is more advantageous to adjust the specific gravity of

the separation medium rather than trying to make fine adjustments in

the density gradient of the seeds to be separated (Downie and Wang

1992)

Germination Requirements

Thinleaf Alder

The IDS separation process significantly improved alder

percentage germination with greater improvement in the better

quality seedlots Simak (1983) achieved an enhancement of lodgepole

pine seeds in which the sinking fraction had almost 7000 germination

at 7 days following sowing compared to 13 for the control and over

90 germination at 21 days post-sowing compared to 6800 for the

control Falleri and Pacella (1997) improved the germination of

London plane tree to 86) with LDS compared to 4800 for the

control

As was the case for total numbers of filled seeds seed sources

varied considerably in the observed germination There was no

distinct latitudinal gradient observed in the data The extremely low

observed germination of the Red River Canyon seed source has been

85

observed in previous studies (Dreesen and Harrington 1998)

Provenance variation in seed properties and germination is not

uncommon and has been reported for a wide range of other woody

species (Young and Young 1992 Baskin and Baskin 1998)

Germination percentage differences in source were noted in a study of

seaside alder (A maritima [Marsh] Nutt) (Schrader and Graves

2000) In the latter study non-stratified seeds varied in germination

percentage from less than 20 for seeds from the Georgia and

Delmarva peninsula sources to more than 40 for the Oklahoma

source Stratified seeds of the Oklahoma source had a 55

germination while the Georgia source had just over 31 germination

and the Delmarva source had nearly 15 germination Schrader and

Graves also noted that germination varied among half-sibling groups

within each source popUlation Stratified groups within the

Oklahoma seed source had germination percentages ranging from 38shy

82 within the Georgia seed source ranging from 12-58 and

within the Delmarva source ranging from 4-29 Non-stratified seeds

from one group in the Delmarva source had germination percentage

less than 1 (Schrader and Graves 2000)

86

There was an interaction between separation and source

Sources with a moderately low percentage ofviable seeds responded

more effectively to the IDS separation method than the source with

a very low percentage fill or the sources with a moderate percentage

of filled seeds Germination for the Luna source (1400 germination in

the non-separated fraction) and the Red River Canyon source (0500

germination in the non-separated fraction) can both be improved

three-fold by separation However a three-fold improvement of a low

germination percentage still results in a low germination percentage

Donald (1985) found that the ID S technique could improve the

viability of a seed source of Pinus eDiottiwhich had a reasonably

high germination capacity but that the technique might not be

valuable for seed stock of very low viability because it cannot

separate normal live seeds from live seeds which have abnormal

germination

Stratification appears to be advantageous for many species of

alder but the influence of stratification was not detectable when all

seed sources were analyzed When analyzed without the very low

viability seed source (Red River Canyon) stratification does affect

87

germination but the actual difference in percentage germination is

not impressive Longer stratification period (56 days) does not appear

to confer any advantage In seaside alder response to stratification is

also source-specific One Oklahoma seaside alder source had

optimum germination with six weeks of stratification and longer

periods tended to reduce germination percentage (Schrader and

Graves 2000) Germination in the other two sources of seaside alder

was not improved by six weeks of stratification This suggests that

stratification and source interactions may be a feature of Alnus

species

Interaction between stratification and separation fraction

among sources can best be explained by the fact that the floating

fractions of all sources and all fractions of the Red River Canyon

source had very little viable seed and therefore very little response to

stratification In contrast the non-separated and sinking fractions of

the other three sources which contained greater amounts ofviable

seed responded to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

in comparison to the sinking fractions

88

There is a three-way interaction between stratification

separation and source The floating fractions of all sources tended to

have a low germination percentage which was not significantly

affected by stratification as was the case with the non-separated and

sinking fraction of the very low viability Red River Canyon source

While the sinking fractions and non-separated samples of the Luna

source and the non-separated Chaffee seed sources responded best to

28 days of stratification the non-separated Reserve seed source

showed no response to stratification The sinking fraction of the

Reserve source responded negatively to 28 days of stratification and

germination at 56 days of stratification was similar to the germination

of the non-stratified sinking fraction Only the sinking fraction of the

Chaffee source responded in linear fashion to increasing stratification

With the elimination of the Red River Canyon source these

interactions are still present (Figure 6)

Pre-soaking of alder seeds had variable effects on germination

depending on seed source Imbibition improved germination for the

Luna source decreased it for the Reserve source and had no effect on

the Red River Canyon or Chaffee sources Pre-soaking with aeration

89

was found to improve ge~nation of red alder and speckled alder

(Berry and Torrey 1985)

Water Birch

Water birch percentage germination was significantly improved

by the density separation in ethanol Falleri and Pacella (1997) found

that density separation of London plane tree seeds in petroleum ether

improved germination to 6000 compared to 4800 for the unseparated

control

Source influenced the effectiveness of separation Seed sources

with moderately rather than extremely low germination percentage

(Moly 2 and Red River Canyon) had an encouraging ten-fold

improvement The Moly 1 source which had an even lower nonshy

separated germination percentage had a five-fold improvement while

the Chaffee source with a 17 germination in the non-separated

portion had a two-fold improvement We can again refer to the study

by Donald (1985) and see that in the case ofwater birch seeds the

moderately low germination percentages of the non-separated seeds

can be effectively improved by separation

90

Stratification significantly improved birch germination

percentage with the longest stratification period showing the most

improvement but the actual gain in percentage was rather low at the

cost ofseveral weeks of time The four seed sources responded

differently to stratification The Moly 2 and Red River Canyon

sources had the greatest germination at 21 days of stratification while

the Moly 1 source had an increase in germination only at 56 days

The Chaffee source responded negatively to stratification especially

at 21 days Paper birch also responds to stratification (Bevington and

Hoyle 1981) with an optimum chilling period of 2 to 3 weeks This

response varied between the New Hampshire and Alaska seed sources

studied with the germination of the New Hampshire source

decreasing somewhat with longer periods of stratification but the

Alaska source maintaining a high percentage with longer stratification

periods

Interaction between stratification and separation fraction can

best be explained by the fact that the floating fractions had very little

viable seed and therefore very little response to stratification while

the non-separated and sinking fractions which contained viable seed

91

had a response to stratification This would also explain the

somewhat muted response of the non-separated seeds to stratification

Source was a significant factor in birch germination percentage

with the Moly 1 source showing very poor germination and the

other sources ranging from 13-1900 Germination of distinct paper

birch sources differed in response to different treatments (Bevington

1986) and germination of separate populations of seaside alder varied

by source (Schrader and Graves 2000)

Pre-soaking was beneficial for the Red River Canyon and

Chaffee seed sources but not the Moly 1 or Moly 2 sources This

difference may be due more to the poor quality of the seed sources

than to different responses to pre-soaking Overall pre-soaking

germination was more than twice that of the non-treated control

General Observations

Stratification increases the germination percentage ofwater

birch and thinleaf alder but does not appear to afford great

improvements in view of the time required Thinleaf alder benefitted

most from a 28-day stratification while water birch responded best to

92

the longer 56-day stratification There are considerable differences in

response to stratification among seed sources for both species

The seed refinement process significantly improves germination

percentage for both thinleaf alder and water birch The LDS method

appears to be an effective tool for seed refinement in thinleaf alder

seed collections where there is a percentage of filled seeds high

enough to make the process practical and the same could be said for

the specific gravity methodmiddotof separation for water birch seeds using

ethanol as a separation medium The techniques could be further

refined for each species In the case ofthinleaf alder an optimum

drying period between 1 hour and 18 hours could be identified using

differences in moisture content to pinpoint the best drying time An

LDS treatment for water birch using drying periods longer than 2

hours might be more effective for seed refinement than the specific

gravity method

Seed source and quality of the seed lot would also determine

how effectively the method works Fine adjustments in method could

make the process for the respective species and seed lot more

effective Preliminary evaluations of a particular seed collection for

93

percentage of filled seeds moisture content and general response to

separation medium might point to the most effective manner of

dealing with that seed source Whether the process is worthwhile in

particular cases would depend on the value of the seed collection in

relation to its potential for improvement and the previously

mentioned constraints of seed supply and growing space

LITERATURE CITED

Albers Daniel J and Stanley B Carpenter 1979 Influence of site environmental conditions mulching and herbaceous ground cover on survival growth and water relations ofEuropean alder seedlings planted on surface mine spoil In Proceedings of the Symposium on Surface Mining Hydrology Sedimentology andReclamation ed Stanley B Carpenter 23-32 Lexington Univ ofKentucky College ofEngineering Press

Allen Edith B 1988 Ecological approaches in theory and practice To what degree is reconstruction possible The Reconstruction ofDisturbedAridLands -An Ecological Approach Academy for the Advancement of Science Sel Symposium 109 ed Edith B Allen 1-4 257-261 Boulder Westview Press

Ashburner K B 1993 Birches in the wild their habitats and ecology In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 19-28 Surrey International Dendrology Society

Barnett James P 1971 Flotation in ethanol reduces storability of southern pine seeds Forest Science 17(1)50-51

Baskin C C and J M Baskin 1998 Within-species variations in seed dormancy Chapter 8 of Seeds Ecology Biogeography andEvolution ofDonnancyand Gennination San Diego CA Academic Press

Belcher E W 1982 You reap what you sow In Proceedings ofthe Southem Containerized Forest Tree Seedling Conference USDA FS GTR-SO-37 eds R W Guilan and James P Barnett 25-28 New Orleans Southern Forest Experiment Station

95

Berry Alison Mand John G Torrey 1985 Seed Germination seedling inoculation and establishment of Alnus spp in containers in greenhouse trials Plant and Soi187161-173

Bevington John M and Merrill C Hoyle 1981 Phytochrome action during prechilling induced germination of Betula papyrifera Marsh Plant Physiol 67705-710

Bevington John M 1986 Geographic differences in the seed germination ofpaper birch (Betula papyrifera) American Journal ofBotany 73(4)564-573

Bewley J Derek and Michael Black 1994 Seeds Physiology of Development and Germination 2nd ed New York Plenum Pub Corp pp 213-287

Biswas P K P A Bonamy and K B Paul 1972 Germination promotion of loblolly pine and baldcypress seeds by stratification and chemical treatments PhysiolPlant 2771-76

Bjorkbom John C D A Marquis and F E Cunningham 1965 The variability ofpaper birch seed production dispersal and germination USFS Research Paper NE-41 Washington GPO

Black M and P F Wareing 1955 Growth studies in woody species VII Photoperiodic control of germination in Betula pubescens Ehrh Physiol Planta 8300-316

Bollen W B and K C Lu 1968 Nitrogen Transformation in soils beneath red alder and conifers In Biology ofAlder eds J M Trappe et aI 141-148 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

96

Bond G 1955 An isotopic study of the fixation of nitrogen associated with nodulated plants of Alnus Myrica and Hippophae Joum ofExper Botany6303-311

---------- 1971 Root nodule formation in non-leguminous angiosperms In Biological Nitrogen Fixation in Natural and Agricultural Habitats Proceedings ofthe Technical Meetings on Biological Nitrogen Fixation ofthe Intemational Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and EG Mulder 317-324 The Hague Martinus Nijhoff Plant and SoilSpecial Volume 1971

---------- 1976 Evidence for fixation of nitrogen by root nodules of alder(Alnus) under field conditions New Phytologist 55147-153

Bonner FT 1984 New forests from better seeds The role of seed physiology In Seedling Physiology and Reforestation Success eds Mary L Duryea and Gregory N Brown 37-60 Dordrecht Martinus NijhoffDrW Junk Pub

Bormann Bernard T 1983 Ecological implications of phytochrome-mediated seed germination in red alder Forest Science 29734-738

Bradbeer JW 1988 Seed Dormancy and Germination London Chapman amp Hall pp 39-131

Brenzel Kathleen N (ed) 1995 Sunset Westem Garden Book Menlo Park Sunset Publishing Corporation pp 183-184

Brinkman Kenneth A 1974 Betula L Birch In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 252-257 Washington GPO

Britton Nathaniel Lord 1908 North American Trees New York Henry Holt amp Co pp246-264

97

Burton Philip J Kenneth R Robertson Louis R Iverson and Paul G Risser 1988 Use of resource partitioning and disturbance regimes in the design and management of restored prairies In The Reconstruction ofDisturbedAnaLands - An Ecological Approach ed E B Allen 46-88 BoulderWestview Press

Carter Jack L 1997 Trees and Shrubs ofNew Mexico Boulder Mimbres Publishing pp 400-402

Crocker Robert L and Jack Major 1955 Soil development in relation to vegetation and surface age at Glacier Bay Alaska Journal ofEcology43427-448

Currie Robin (moderator) 1990 Betula propagation Seminar CombProcInter Plant Prop SocAO315-317

Daniel Theodore John A Helms and Frederick S Baker 1979 Principles ofSilviculture 2nd ed New York McGraw Hill Book Co pp373-376

Danielson H Rodger and Yasuomi Tanaka 1978 Drying and storing stratified ponderosa pine and douglas-fir seeds Forest Science 24(1)11-16

deJong PC 1993 An introduction to Betula its morphology evolution classification and distribution with a survey of recent work In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 7-18 Surrey International Dendrology Society

Dick-Peddie William A 1993 Riparian vegetation Chapter 9 of New Mexico Vegetation Pas Present andFuture Albuquerque UNM Press

Dirr Michael A and Charles W Heuser 1987 The Reference Manual ofWoody Plant Propagation From Seed to Tissue Culture Athens GA Varsity Press ppII-2290-9196-97

98

Donald D G M 1985 The separation of full dead seed from live seed in Pinus elliottii In Proceedings ofthe Intemational Symposium on Nursery Management Practices for the Southem Pines Montgomery AL August 4-9 1985 ed David B South 83-88 Auburn AL Auburn University

Downie Bruce and Ben S P Wang 1992 Upgrading germinability and vigour ofjack pine lodgepole pine and white spruce by the IDS technique Can J For Res 22(8)1124-1131

Dreesen D R and J T Harrington 1998 Propagation of native plants for restoration projects in the southwestern U S shyPreliminary investigations In Proceedings ofthe Westem Forest and Conservation Nursery Association Meeting Boise ID August 19-21 1997 ed T D Landis pp 77-88

Dunlap J R and J P Barnett 1984 Manipulating loblolly pine (Pinus taeda L) seed germination with simulated moisture and temperature stress In Seedling Physiology andReforestation Success eds Duryea Mary L and Gregory N Brown 61-74 Dordrecht Martinus NijhoffDrWJunk Pub

Elias Thomas S 1980 The Complete Trees ofNorth Amenca-Field Guide andNatural History New York Outdoor LifeNature Books Van Nostrand Reinhold Co pp 385-412

Falled Elisabetta and Rosetta Pacella 1997 Applying the IDS method to remove empty seeds in Platanus x acemolia Can J For Res 271311-1315

Fowler D P and T W DWight 1964 Provenance differences in the stratification requirements of white pine Can Joum ofBotany 42669-675

Fowells H A 1965 Silvics ofForest Trees ofthe United States USDA-Forest Service Agricultural Handbook 271 compo H A Fowells 1-4 82-88 92-109 Washington GPO

99

Foxx Teralene S and Dorothy Hoard 1995 Flowering Plants ofthe Southwestern Woodlands Los Alamos Otowi Crossing Press pp26-27

Franklin JerryF andAnnaA Pechanec 1968 Comparison of vegetation in adjacentalder conifer and mixed alder-conifer communities In Biology ofAlder eds J M Trappe et al 37shy44 PortlandUSDA FS Pacific Northwest Forest arid Range Experiment Station

Haeussler Sybille J C Tappeiner II and B J Greber 1995 Germination survival and early growth of red alder seedlings in the ~entral Coast Range of Oregon Canadian Journal of Forest Research 25(10) 1639-1651

Harker Donald Sherri Evans Marc Evans and Kay Harker 1993 Landscape Restoration Handbook Boca Raton Lewis Publications pp 19-2465-71

Harrington J T and P A Glass 1997 Determining the number of seeds to sow per cell An application of the geometric distributions Tree Planters Notes 828-34

Hartmann Hudson T Dale E Kester Fred T Davies Jr and Robert L Geneve 1997 Plant Propagation Principles and Practices 6th ed Upper Saddle River NJ Simon amp Schuster pp194-215 671673-674

Herrera M A C P Salamanca and J M Barea 1993 Inoculation ofwoody legumes with selected arbuscular mycorrhizal fungi and Rhizobia to recover desertified Mediterranean ecosystems AppliedandEnv Micro 59129-133

Hibbs David E Dean S DeBell and Robert F Tarrant eds 1994 The Biology andManagement ofRedAlder Corvallis Oregon State University Press

100

Hilhorst H WM A Smitt and C M Karssen~ 1986 Gibberellin-biosynthesis and sensitivity mediated stimulation of seed germination of Sisymbrium oflicinale by red light and nitrate Physiol Plant 67285-290

Hill HI A G Taylor and T -G Min 1989 Density separation of imbibed and primed vegetable seeds J Amer Soc Hort Sci 114(4)661-665

Hilton Janet R 1985 The influence of light and potassium nitrate on the dormancy and germination of A vena fatua L (wild oat) seed stored buried under natural conditions JExp Botany 36974-979

Hobbs S D 1984 The influence of species and stocktype selection on stand establishment An ecophysiological perspective In Seedling Physiology andReforestation Success eds Mary L Duryea and Gregory N Brown 179-224 Dordrecht Martinus NijhoffDrWJunk Pub

Johnson Frederic D 1968 Taxonomy and distribution of Northwestern alders In Biology ofAlder eds J M Trappe et aI 9-22 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Joseph Hilda C 1929 Germination and vitality ofbirch seeds Bot Gazette 87 127-151

Kenady Reid M 1978 Regeneration of red alder In Utilization and Management ofAlder compiled by D G Briggs D S DeBell and W A Atkinson 183-191 USDA Forest Service GTR-PNW-70 Portland Pacific Northwest Forest and Range Experiment Station

101

-

Lane C G 1993 Propagation of the genus Betula In Betula Proceedings ofthe IDS Betula Symposium Sussex England October 1992 ed David Hunt 51-60 Surrey International Dendrology Society

Lentner M and T Bishop 1986 Experimental Design and Analysis Blacksburg VA Valley Book Company p 173

Mallinckrodt Baker Inc 1997a Material Safety Data Sheet-Petroleum Ether MSDS Number P1696 Phillipsburg NJ Mallinckrodt Baker Inc

Mallinckrodt Baker Inc 1997b Material Safety Data Sheet-Proprietary Solvent III-I Anhydrous (Denatured Ethanol) MSDSNumberP6735 Phillipsburg NJ Mallinckrodt Baker Inc

Martin W C and C R Hutchins 1980 A Flora ofNew Mexico Vaduz J Cramer AR Gantner Verlag K G pp510-514

Mayer A M and A Poljakoff-Mayber 1989 The Germination of Seeds 4th ed Oxford Pergamon Press pp 71-99

McLemore B F 1965 Pentane flotation for separating full and empty longleaf pine seeds Forest Science 11(2)242-243

McVean D N 1956 Ecology ofAlnusglutinosa (L) Gaertn III Seedling establishment JEcol44195-218

Monsen Stephen B 1984 Use of shrubs on mine spoils In The Challenge ofProducing Native Plants for the Intermountain Area Proc Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT-168 26-31 Ogden Intermountain Forest and Range Experimental Station

102

National Academy ofSciences (NAS) 1974 Rehabilitation Potential ofWestem Coal Lands eds Thadis W Box Richard F Hadley andM Gordon Wolman 11-1549-7273-94 Cambridge Ballinger Pub Co for the N A S (for the Ford Foundation)

Nielson Rex F and HB Peterson 1973 Establishing vegetation on mine tailings waste In Ecology andReclamation ofDevastated Land vol II eds Russell J Hutnik and Grant Davis 103-115 New York Gordon amp Breach

Phillips Judith 1995 Plants for Natural Gardens Santa Fe Museum ofNew Mexico Press ppl08-109

Pratt Carl R 1986 Environmental factors affecting seed germination ofgray birch (Betula popuJifoJia) collected from abandoned anthracite coal mine spoils in northeast Pennsylvania Ann Appl BioI 108649-658

Preston Richard J Jr 1968 RockyMountain Trees A Handbook of the Native Species New York Dover Publications Inc pp 120shy123124-129134-135

Radwan M A andD S DeBell 1981 Germination of red alder seed USDA Forest Service Res Note PNW-370 Portland Pacific Northwest Forest and Range Experiment Station

Rose R W C Carlson and P Morgan 1990 The target seedling concept In Target Seedling Symposium Proceedings Combined Meeting ofthe Westem Forest Nursery Associations August 13-17 1990 Roseburg Oregon eds Robin Rose et al General Technical Report RM-200 Ft Collins CO USDA Forest Service Rocky Mountain Forest and Range Experiment Station

103

Ross JD and J W Bradbeer 1971 Studies in seed dormancy V The content of endogenous gibberellins in seeds of Corylus avellana L Planta 100288-302

Rudolf PaulO 1950 Cold soaking-a short cut substitution for stratification Joum ForeSl1y4831-32

SAS Institute Inc 1989 SASSTATUsers Guide Version 6 Fourth Edition Volume 1 Cary NC SAS Institute Inc

Sargent Charles S 1901 New or little known North American trees m Bot Gazette 31(4)217-240

---------- 1905 ManuaJ ofthe Trees ofNorth America New York Houghton Mifflin (reprint 1965 New York Dover Pub) pp 205-206218-220223-226

Schalin TImari 1968 Germination analysis of grey alder (Alnus incana) and black alder (Alnus glutinosa) seeds In Biology of Alder eds J M Trappe et aI 107-114 Portland USDA FS Pacific Northwest Forest and Range Experiment Station

Schopmeyer C S 1974 AlnusB Ehrh In Seeds ofWoody Plants in the United States USDA Agricultural Handbook 450 ed C S Schopmeyer 19-40 126-135 140-152206-211 Washington GPO

Schrader James A and William R Graves 2000 Seed germination and seedling growth ofAlnus maritima from its three disjunct populations J Amer Soc Hort Sd 125(1)128-134

Schubert Gilbert H L J Heidman and M M Larson 1970 ArtificiaJ Reforestation Practices for the Southwest USDA Agricultural Handbook 370 Washington GPO

104

Simak Milan 1983 A new method for improvement of the quality of Pinus contorta seeds In Lodgepole pine regeneration and management ed Mayo Murray 39-41 USDA For Servo Gen shyTech Rep PNW-157

Slavik Bohdan 1974 Water exchange between plant and atmosphere Chapter 5 of Methods ofStudying Plant Water Relations Prague Academia Publishing House of the Czechoslovak Academy ofSciences

Sweeney J D Y A EI-Kassaby D W Taylor D G W Edwards and G E Miller 1991 Applying the IDS method to remove seeds infested with the seed chalcid Megastigmus spermotrophus Wachtl in douglas-fir Pseudotsuga menziesii (Mirb) Franco New Forests 5327-334

Taylor A G and T J Kenny 1985 Improvement of germinated seed quality by density separation J Amer Soc Hort Sci 110(3)347-349

Tarrant Robert F 1961 Stand development and soil fertility in a douglas-fir - red alder plantation Forest Science 7238-246

Tarrant Robert F and James M Trappe 1971 The role of Alnus in improving the forest environment In Biological Nitrogen Fixation in Natural andAgricultural Habitats Proceedings of the technical meetings on biological nitrogen fixation ofthe Ind Biological Programme (Section PP-N) Prague and Wageningen 1970 eds T A Lie and E G Mulder 335-348 The Hague Martinus Nijhoff Plant and Soil Special Volume 1971

Thompson P A 1971 Research into seed dormancy and germination Comb Proc Inter Plant Prop Soc 21211-228

Villiers T A and P F Wareing 1964 Dormancy in fruits of Fraxinus excelsior L Joum Exp Botany 15(44)359-367

105

Vines Robert A 1960 Trees Shrubs and Woody Vines ofthe Southwest Austin University of Texas Press pp139-142

Virtanen Artturi 1 1957 Investigations on nitrogen fixation by the alder II Associated culture of spruce and inoculated alder without combined nitrogen Physio1 P1anta 10164-169

Webb DP and PF Wareing 1972 Seed dormancy in Acer Endogenous germination inhibitors and dormancy in Acer pseudop1atanusL P1anta 104115-125

WennyDL 1993 Calculating filled and empty cells based on number of seeds sown per cell A microcomputer application Tree Planters Notes 4449-52

Whitford W 1988 Decomposition and nutrient cycling in disturbed arid ecosystems In The Reconstruction ofDisturbedArid Lands -An Ecological Approach ed E B Allen 136-16l Boulder Westview Press

Wilcox James R 1968 Sweetgum seed stratification requirements related to winter climate at seed source Forest Science 1416-19

Wooton E O and Paul C Standley 1915 Flora ofNew Mexico Washington GPO pp 163-164

Young James A Jerry D Bundy and Raymond A Evans 1984 Germination of seeds of wildland plants In The Challenge of Producing Native Plants for the Intermountain Area Proceedings ofthe Intermountain Nurserymans Assoc 1983 Conference Las Vegas Nevada USDA Forest Service GTR-INT 168 1-5 Ogden Intermountain Forest and Range Experimental Station

106

Young James A and Cheryl G Young 1986 Collecting Processing and Germinating Seeds ofWildland Plants Portland Timber Press pp 59-65 84-85

Young James A and Cheryl G Young 1992 Seeds ofWoody Plants in North America Portland Dioscorides Press pp 26-28 55-59

Young J F 1967 Humidity control in the laboratory using salt solutions - a review J Applied Chem 17241-245

Page 18: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 19: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 20: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 21: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 22: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 23: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 24: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 25: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 26: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 27: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 28: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 29: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 30: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 31: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 32: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 33: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 34: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 35: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 36: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 37: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 38: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 39: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 40: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 41: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 42: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 43: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 44: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 45: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 46: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 47: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 48: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 49: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 50: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 51: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 52: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 53: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 54: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 55: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 56: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 57: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 58: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 59: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 60: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 61: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 62: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 63: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 64: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 65: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 66: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 67: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 68: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 69: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 70: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 71: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 72: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 73: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 74: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 75: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 76: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 77: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 78: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 79: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 80: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 81: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 82: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 83: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 84: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 85: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 86: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 87: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 88: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 89: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 90: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 91: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 92: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 93: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 94: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 95: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 96: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 97: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 98: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 99: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 100: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 101: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 102: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 103: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 104: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 105: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 106: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 107: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 108: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 109: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 110: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 111: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 112: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 113: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 114: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 115: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 116: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 117: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 118: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 119: SEED UPGRADE AND GERMINATION STRATEGIES BETULA
Page 120: SEED UPGRADE AND GERMINATION STRATEGIES BETULA