Assessment of quantitative and genetic molecular variation of Acacia karroo

in two extreme populations

By

Georges Bayonne Mboumba

Thesis presented in partial fulfilment of the requirements for the degree of

Master of Science in Conservation Ecology

at the

Faculty of Agricultural and Forestry Sciences

University of Stellenbosch

Supervisor: Professor David Ward

April 2006

Declaration

I, the undersigned, hereby declare that the work contained in this thesis is my own

original work and that I have not previously in its entirety or part submitted it at any

university for a degree.

Signature:

Georges Bayonne Mboumba

Date:

Abstract

Abstract

Acacia karroo is widespread in southern Africa and displays remarkable

phenotypic plasticity over its geographical range. However, it is currently unknown

whether this phenomenon is merely phenotypic variation due to environmental variance

or whether such plasticity represents adaptation to different habitats (known as adaptive

phenotypic plasticity). Adaptive phenotypic plasticity implies that genotypes differ and

that there is local adaptation to the local environment. To shed light on this phenomenon,

we used a common-garden experiment to investigate among-population variation in

plastic responses to simulated rainfall and browsing in two populations originating from

contrasting environments, namely arid Karoo (Leeu Gamka) and subtropical coastal

forest (Richards Bay). We also studied genetic variation among populations by means of

allozyme markers. The results suggest that the populations investigated are both

genetically distinct and phenotypically plastic. In addition, there were high levels of

polymorphism within populations and great differences in their range of plastic responses

to treatments. Of the two populations investigated, the slow-growing one (Leeu Gamka)

was phenotypically more plastic with regard to defence-related traits (longer spines, more

tannin) while the fast-growing one (Richards Bay) was phenotypically more plastic

regarding growth-related traits (taller, with longer leaves). Patterns of performance

revealed that the populations have pure strategies of either growth (forest) or defence

(arid). The interactions between populations and environments in some traits indicated

genetic differentiation in plastic responses between populations and, consequently, that

phenotypic plasticity is locally adaptive and not merely due to environmental differences.

The two populations appear to have pure strategies; when environmental conditions were

i

Abstract

improved by addition of water, the forest population increased investment in growth but

not defence, while the arid populations increased defence production but not growth.

ii

Opsomming

Opsomming

Acacia karroo is wydverspreid in suidelike Afrika and vertoon merkwaardige

fenotipiese plastisiteit dwarsoor die spesie se geografiese verspreidingsgebied. Nietemin

is dit tans nie bekend of hierdie verskynsel maar net fenotipiese variasie as gevolg van

omgewingsverskille is, of dat hierdie plastisiteit aanpassings aan verskillende habitatte

(bekend as fenotipiese plastisiteit) verteenwordig nie. Aanpaslike fenotipiese plastisiteit

dui daarop dat genotipes van mekaar verskil en dat plaaslike aanpassings aan die

plaaslike omgewing voorkom. Om meer ingligting te verkry oor dié verskynsel het ons 'n

gemeenskaplike-tuin eksperiment gebruik om tussen-die-populasie variasie in plastiese

reaksies op nagebootste reënval en weiding te bestudeer in twee populasies afkomstig van

kontrasterende omgewings, naamlik droë Karoo (Leeu Gamka) en subtropiese kusbos

(Richards bay). Ons het ook genetiese variasie binne populasies bestudeer deur die

gebruik van allosiemmerkers. Die uitslae dui daarop dat die populasies wat ondersoek is

terselfdertyd geneties verskillend en fenotipies plasties is. Daar was boonop hoë vlakke

van polimorfie binne verskillende populasies en groot verskille in die omvang van die

plastiese reaksies op behandelings. Tussen die twee studie populasies was die stadig-

groeiende een (Leeu Gamka) meer fenotipies plasties ten opsigte van verdedigings-

verwante kenmerke (langer dorings, meer tannien) terwyl die vinnig-groeiende een

(Richards bay) fenotipies meer plasties was ten opsigte van groei-verwante kenmerke

(langer groei vorm, met langer blare). Die manier waarop populasies vertoon het, het

duidelike strategiëe van óf groei (bos) óf verdediging (droë omgewing) tussen die

populasies aangedui. Die wisselwerking tussen populasies en die omgewing het, ten

opsigte van sekere kenmerke, daarop gewys dat genetiese verskille in die plastiese

iii

Opsomming

reaksies tussen populasies bestaan en, gevolglik, dat fenotipiese plastisiteit plaaslik

aanpasbaar en nie net 'n gevolg van omgewingsverskille is nie. Dit blyk dat die twee

populasies suiwer strategiëe volg; die verbetering van omgewings-omstandighede deur

die toevoeging van water, het in die bos populasie gelei na verhoogde belegging in

groeivermoë maar nie in verdediging nie, terwyl die droë omgewing populasies ’n

verbetering in verdedigingsvermoë getoon het, maar nie in groei nie.

iv

Acknowledgements

Acknowledgements

I would like to thank all those people who helped me all along the tortuous route

towards the realization of this work. Nir Peleg and Gavin Gouws for teaching me the gel

electrophoresis of allozymes, Dr Theron for technical advice on greenhouse experiment,

Michiel Meets from Omnia Specialities for providing me with guidance in the use of

Vitagro (fertilizers), Heidi Thunemann, Cara Nieuwoudt, Khanysile Mbatha for technical

assistance in the laboratory and, Jack Kambatuku, Christina Potgieter and Dr Megan

Griffiths for the proofreading of manuscripts. By the time I was losing hope about my

work some people gave me extremely good advice that is the reason for which I would

like to single them out and thank them sincerely: Graziella Ntolo, my dad Georges

Mboumba Snr., my sisters and brothers. To you mom, Issope Jacqueline, may your soul

rest in peace. To Robert and Morean Aling I extend my heartfelt gratitude for the

wonderful environment in which I wrote this manuscript. This would not have been done

without the funding of the National Research Foundation to Prof David Ward and the

Gabonese Government towards my studies. A special thank to my Supervisor Prof David

Ward for believing in me knowing that I was in the process of learning this new language

(English). Many thanks to you and Dr Megan Griffiths for making me feel at home in

your house.

v

Foreword

Foreword

This study aims to investigate the apparent phenotypic plasticity displayed by

Acacia karroo growing in different habitats in order to contribute towards a more

comprehensive understanding of this phenomenon. The main objective is to determine

whether such a phenomenon is due to environmental variance or whether it represents

adaptation to different habitats (i.e. adaptive phenotypic plasticity). To achieve this aim,

we therefore compared the performances of two populations experiencing a variety of

selection pressures in controlled conditions and measured the amount of genetic

population differentiation.

This thesis encompasses the procedures, processes and outcomes of our

investigations and is presented in four chapters. Chapter 1 is an introduction to the

themes related to the study and explains the theoretical concepts of phenotypic plasticity

and adaptive phenotypic plasticity. It also introduces standard methodological approaches

for experimental evaluations of adaptive phenotypic plasticity and measurements of the

amount of genetic population differentiation. The chapter further provides insights into

stable environment vs. variable environment models, phenotypic plasticity of Acacia

karroo, phenology, breeding system, morphological and genetic variation in the species

and the objectives of this study. Chapters 2 and 3 have both been submitted for

publication to and hence follow the format of the Journal of Forest Ecology and

Management. Chapter 2 presents the findings on the genetic variation in both

populations. Chapter 3 deals with phenotypic plasticity and local adaptation in both

populations. The general conclusions are given in Chapter 4.

vi

Contents

Contents

Abstract............................................................................................................................... i Opsomming....................................................................................................................... iii Acknowledgements ........................................................................................................... v Foreword........................................................................................................................... vi Contents ........................................................................................................................... vii

Chapter 1

Introduction........................................................................................................... 1 Phenotypic plasticity........................................................................................... 1 Adaptive phenotypic plasticity ............................................................................ 3 Experimental evaluations of adaptive phenotypic plasticity .............................. 5 Measuring the amount of genetic population differentiation ............................. 7 Stable environment vs. variable environment................................................... 10 Phenotypic plasticity of Acacia karroo ............................................................. 11 Phenology and breeding system of Acacia karroo ............................................ 13 Morphological and genetic variation in Acacia karroo.................................... 15

General aims and methodology ......................................................................... 17 Objectives.......................................................................................................... 17 Study sites.......................................................................................................... 17

References............................................................................................................ 21 Chapter 2

Genetic variation in two extreme populations of phenotypically-plastic Acacia karroo ....................................................................................................................... 33

Introduction......................................................................................................... 35 Phenotypic and genotypic variation in Acacia karroo...................................... 37

Materials and methods ....................................................................................... 40 Population sampling ......................................................................................... 40 Allozyme electrophoresis .................................................................................. 41 Data analysis .................................................................................................... 42

Results .................................................................................................................. 43 Within-population polymorphisms.................................................................... 43 Among- population polymorphisms .................................................................. 45

Discussion............................................................................................................. 47 Population genetic variation............................................................................. 47

Acknowledgements ............................................................................................. 50 References............................................................................................................ 51

Chapter 3

Phenotypic plasticity and local adaptation in two extreme populations of Acacia karroo ....................................................................................................................... 57

Abstract................................................................................................................ 58 Introduction......................................................................................................... 60

vii

Contents

Strategies of growth and defence...................................................................... 61 Materials and methods ....................................................................................... 63

Study site and populations ................................................................................ 63 Data analysis .................................................................................................... 66

Results .................................................................................................................. 66 Phenotypic plasticity......................................................................................... 66 Trade-offs between growth and defence ........................................................... 72 Genotypic by environment interactions among populations ............................ 73

Discussion............................................................................................................. 78 Acknowledgements ............................................................................................. 81 References............................................................................................................ 82

Chapter 4

Conclusions.......................................................................................................... 88 References............................................................................................................ 92

viii

Chapter 1

Introduction

Phenotypic plasticity

A species can display marked variation and differences in attributes and

appearances as a result of adaptive responses to dissimilar environmental conditions

(Briggs and Walters, 1997). Environmentally imposed variations in a species may

manifest themselves in the forms, sizes or patterns of growth. Such an ability by a single

genotype to produce an array of phenotypes depending on the environmental scope is

termed phenotypic plasticity (Bradshaw, 1965; Schlichting & Levin, 1984; Via et al.,

1995; Macdonald and Chinnappa, 1989). This definition of phenotypic plasticity

encompasses physically discernible morphological growth or development of organs, as

well as physiological changes at the cellular level. Woltereck (1909) experimentally

documented the nonlinear relationship between environmental variation in food

availability and phenotypic variation in traits for different varieties of Hyalodaphnia and

referred to it as a reaction norm. This term has subsequently been revived by

Schmalhausen (1949) as a synonym for phenotypic plasticity; however, reaction norm is

not an entirely appropriate alternate term to phenotypic plasticity given that not all

reaction norms are necessarily plastic. A genotype exhibiting a constant phenotype across

different environments is not plastic but is a reaction norm (Fig. 1). To maintain clarity of

meaning and avoid confusions, phenotypic plasticity and reaction norms will not be used

interchangeably in this work.

Although the phenomenon of phenotypic plasticity occurs in both animals and

plants, it is more frequent and prominent in plant species because of (i) their relative

1

Chapter 1

immobility once established, which confines them to a given environment, as well as (ii)

the continuity of organ development during their life history (e.g. Bradshaw, 1965;

Agrawal et al., 1999; Novoplansky, 2002). Continual growth of plants means that new

organs produced during a particular growth season will constantly be tailored to the

environmental conditions and pressures prevailing at the time. As a result, a plant may

exhibit phenotypic characteristics of its spatial locality as well as the cumulative effect of

temporal environmental changes. The literature is replete with evidence of phenotypic

plasticity observed as variation in the expression of phenotypic plasticity within

(Jasienski et al., 1996; Schlichting and Pigliucci, 1998) and among populations of the

same plant species (Turesson, 1961; Briggs and Walters, 1984; Bradshaw, 1965; Counts,

1993; Jasienski et al., 1996; Schlichting and Pigliucci, 1998) and in related taxa

(Bradshaw, 1965; Cook, 1974). The capacity for phenotypic plasticity varies such that a

marked differentiation in ability of plasticity is found within plants. Widespread species

such as weeds and invasive plants are known to be highly plastic relative to plants with a

narrow distribution range (e.g. Baker, 1974; Williams et al., 1995; Linhart and Grant,

1996).

Our understanding of the ecological and evolutionary significance of phenotypic

plasticity is incomplete. Some studies have shown phenotypic plasticity to fulfil an

adaptive role (e.g. Jefferies, 1984; Schlichting, 1986; Sultan, 1987; Macdonald and

Chinnappa, 1989) by the plants to the environment. The role of plasticity in plants is best

elucidated in studies by Caldwell et al. (1991) who suggested that plasticity of organs

increases a plant’s opportunistic ability in variable environments. Thus, the plasticity of a

plant’s organs augments the plant’s performance under varying conditions of resource

2

Chapter 1

abundance (Bilbrough and Caldwell, 1997). By extension, the flexibility of plants to

adjust the development and growth of organs in accordance with temporal and spatial

changes in resource availability enables them to maximise their acquisition of available

resources in their environments. Sultan and Bazzaz (1993) suggested that plasticity of

organs helps plants to remain functional despite limiting and stressful environmental

conditions. These viewpoints are not necessarily mutually exclusive because varying

resource availability on a spatial and temporal scale is among the factors contributing to

environmental limitation and stress. As such, adaptive opportunism based on phenotypic

plasticity would allow a plant to retain its vitality in a variably limiting environment.

Phenotypically plastic responses involve an amount, which refers to the magnitude of

response to the environmental change (small or large) and a pattern, referring to the

shape of response (monotonic increase/decrease, or curves) (Bradshaw, 1965; Schlichting

and Levin, 1984), a speed of response (Kuiper and Kuiper, 1988), as well as a

reversibility, referring to the ability of switching between alternative states (Piersma and

Lindstrom, 1997) that are under genetic control and, consequently, subject to influence

by evolutionary forces.

Adaptive phenotypic plasticity

Alpert and Simms (2002) argued that phenotypic plasticity could be adaptive “…

plasticity in a trait that results from direct selection”, detrimental “… plasticity in a trait

due to inability to maintain a constant phenotype despite fitness reduction due to

variation” or neutral “… plasticity in a trait resulting from a lack of selection either for

or against variation accumulated through processes such as mutation or selection on

3

Chapter 1

other traits that are functionally related”. Whatever the outcome or consequence of

adaptive phenotypic plasticity, it is a phenomenon of considerable evolutionary and

ecological interest. Indeed, adaptive phenotypic plasticity implies that genotypes differ in

their genetic make-up and that there is local adaptation (also known as genotype-by-

environment (G×E) interaction) (e.g. Schlichting and Levin, 1984; Winn, 1996).

Adaptation can take the form of location-specific phenotypic expression through selective

pressures that produce the morphology and structure best suited to exploit local

conditions. Simply put, local adaptation can be defined as the evolution, through natural

selection, of traits that have high fitness in the environmental conditions specific to a

population (Freeman and Herron, 2001). The question of perennial interest is why this

occurs. Ridley (1996) and Freeman and Herron (2001) agreed that the degree to which

local adaptation can occur in a species depends on the potential for populations to evolve

differences from each other and the potential for natural selection to occur within each

population. The basis for adaptive flexibility of populations is the presence of variant

alleles in a population’s genome on which environmental factors will act and interact to

bring about morphological adjustment. This would mean that populations with a high

level of genetic drift (i.e. random change in allele frequencies), a phenomenon related to

small population size, or with high rates of gene flow preventing genetic differentiation

of populations would have lower potential for local adaptation than larger populations or

those with a low rate of gene flow. Examples of local adaptation include morphological

differences between aerial and aquatic leaves in plants (Bradshaw, 1965; Cook, 1974),

differences in growth and defence between plants growing in very poor environmental

conditions and those growing in rich environments (e.g. Coley et al., 1985). Evidence of

4

Chapter 1

local adaptation has also been demonstrated within populations of clonal species

(Turkington and Harper, 1979; Fisher et al., 2000; Alpert et al., 2003; Knight and Miller,

2004) as well as within subpopulations in non-clonal species that are separated by large

distances (Sork et al., 1993) or that have other barriers preventing gene flow (Shrestha et

al., 2002). The study conducted by Shrestha et al. (2002), on isolated populations of

Acacia raddiana in the Negev desert suffering from high mortality and limited

recruitment, showed that there was local adaptation in some morphological traits, with

water availability acting as the main selection pressure with regard to genetic

differentiation among the populations studied.

Experimental evaluations of adaptive phenotypic plasticity

Reaction norms are measures of the extent to which genotypes can be plastic and,

consequently, are indices of the nature of adaptive phenotypic plasticity. If one measures

the same trait in two or more different environments, the following results can be

obtained (Stearns 1992 - see Figure 1):

(a) parallel reaction norms indicate differences in the mean response (e.g.

genotypes 3 and 4).

(b) non-parallel or crossing reaction norms indicate genotype by environment

interaction (that is genetic variation in the plastic response in the trait of interest (e.g.

genotypes 2 and 3, 2 and 4)). Note that, while crossing reaction norms are typically used

to represent genotype by environment interactions, these reaction norms need only be

non-parallel (i.e. in an ANOVA, there must be a significant interaction effect) (Stearns

1992, Lynch and Walsh 1998).

5

Chapter 1

0

2

4

6

8

10

env 1 env 2

Environment

Phen

otyp

egenotype 1genotype 2genotype 3genotype 4

Figure 1: Reaction norms of imaginary genotypes of plants in two distinct environments.

Two types of experiments are efficient and informative for detecting genetic basis

of phenotypic plasticity and adaptive phenotypic plasticity (see e.g. Platenkamp and

Shaw 1992). Common-garden and reciprocal transplant experiments have been successful

in early studies (e.g. Kerner, 1895; Johannsen, 1909; Bonnier, 1920; see Briggs and

Walters, 1984) and more recently (e.g. Shaver et al., 1986; Schmid, 1992; Schmidtling,

1994; Shrestha et al., 2002; Volis et al, 2003; Knight and Miller, 2004). In general, they

include replicates of genetically-related individuals (e.g. full or half sib families, clones)

and focus on different traits such as growth (Sultan, 1992; Sultan and Bazzaz, 1993) or

physiological responses (Chapin and Shaver, 1996). Indeed, reciprocal transplant

experiments consist of growing these replicates from different populations of different

provenances in switched environments. Results from such experiments give an idea about

how critical evolutionary differences are to the persistence of a population at a site

6

Chapter 1

(MacGraw and Antonovics, 1983; Schwaegerle, 1996). Common garden experiments

can also be used to assess genetic differences in traits relevant to environmental change

(e.g. Corn and Hiesey, 1973). There is sufficient evidence for genetic variation in plastic

responses both within (van Kleunen et al., 2000; Smekens and van Tienderen, 2001;

Schlichting and Smith, 2002; Shrestha et al., 2002) and between conspecific plant

populations (Leiss and Muller-Scharer, 2001; Wilson, 2001; Botto and Smith, 2002;

Schlichting and Smith, 2002). Although these quantitative genetic experiments detect

genetic variation in plastic responses of traits among populations they do not directly

indicate the genetic variation existing within populations, which can be measured

indirectly with genetic markers that show the extent to which populations differentiate

genetically. Thus far, most studies carried out on the topic have considered quantitative

genetics experiments and genetic marker-based assessments separately. However, in

order to show that phenotypic plasticity is locally adaptive there is a need to combine

both types of assessment.

Measuring the amount of genetic population differentiation Molecular studies of plants provide insights into patterns of genetic diversity in natural

populations, variation in the mating system and genes controlling quantitative traits that

can be utilized for genetic improvement (e.g. Butcher, 2002, Parker et al., 1998). The

earliest study to use molecular markers reported that protein electrophoresis was able to

measure molecular variations in animals (e.g. Lewontin and Hubby, 1966). Since then, an

enormous number of studies have employed protein electrophoresis and other molecular

markers to examine population differentiation in plants and animals (e.g. Cheliak and

7

Chapter 1

Pitel, 1984; Soltis and Soltis, 1989; Godt and Hamrick, 1998; Joly et al., 1992; Oballa,

1993; Paschke et al., 2002; Shrestha et al., 2002 and so forth). A variety of molecular

tools capable of efficiently detecting differences among individuals exist to date

(reviewed by Parker et al. 1998). DNA-based techniques such as DNA sequencing,

random amplified polymorphic DNA (RAPDs), restriction fragment length

polymorphisms (RFLPs), microsatellite DNA (SSRs), and amplified fragment length

polymorphisms (AFLPs) are now widely employed. Although new methods of probing

the molecular basis of genetic variation have been developed, protein (enzyme)

electrophoresis remains a useful tool for population genetics and systematics because of

its reliability, modest cost and ease of application (Grant, 1989). An allozyme is an

enzyme that is the product of a particular allelic form of a gene (Hedrick, 1985) and an

isozyme is any of the distinct forms of enzyme that have identical or nearly identical

chemical properties but are encoded by different loci (Li and Graur, 1991). A number of

studies using allozymes as markers have been successful in demonstrating how

phenotypically-plastic conspecific populations differ. For example, Macdonald and

Chinnappa (1989) investigated population differentiation for phenotypic plasticity in the

Stellaria longipes complex and found that populations differed genetically. However, the

degree of plasticity was not related to the degree of isozyme variability in the five

populations. Strong similarities between patterns of variation for isozymes and

morphology among populations of Norway spruce (Picea abies) were found in a study by

Lagercrantz and Ryman (1990). Similar results were also found among populations of

Petrorbagia prolifera (Lonn and Prentice, 1990). Morphometric and allozymic studies in

populations of a montane herb, Ipomopsis aggregate, from three localities showed

8

Chapter 1

differentiation of floral morphology and allozymes among populations (Wolf and

Campbell, 1995). The amount of differentiation was different for allozymes and for floral

morphology; the variance component of allozymes decreased monotonically on an

increasing spatial scale, whereas the variance component of morphological characters

among localities was higher than among populations within localities. Interesting results

have been obtained by studies that show both genetic differentiation (that is, molecular

genetic variation) and genetic variation in plastic responses of traits among populations,

as they provide different approaches to exploring adaptive phenotypic plasticity.

However, it is unfortunate that few studies of this nature have been conducted on plants

(e.g. Ritland and Jain (1984) and Shrestha et al. (2002)). The discrepancy between neutral

and adaptive variation resides in the fact that neutral variations have effect when

evolutionary processes that impact individuals, populations and species, that is drift and

migration, occur; while adaptive variations involve when selection operate on molecular

variation by means of phenotype. As a consequence, there is a dilemma when using

neutral markers and then make conclusions with reference to adaptation. Forms of neutral

or nearly neutral markers like allozymes and macrosatellites are unlikely to precisely

foretell patterns of variation in quantitative traits when selection, rather than drift, is the

most important force acting, for illustration local adaptation and speciation (Reed and

Frankham, 2001). Quantitative trait loci (QTLs) are advanced markers that are

satisfactory tools for detecting adaptation and fitness.

9

Chapter 1

Stable environment vs. variable environment

The basic features of a variable environment are the instability and

unpredictability of environmental factors and conditions in space and time. In such an

environment, abrupt and inconsistent changes in physical conditions, resource

availability, climatic and biological factors are found within a short space of time and

location. The heterogeneity in a variable environment is often associated with the level of

stress experienced by organisms inhabiting such an environment. As an example, arid

regions are both highly stressed and highly variable in terms of availability of resources

in space and time (e.g. Noy-Meir, 1973). In contrast, stable environments such as forests

are relatively less stressed and less variable in terms of resource availability (e.g. Grime,

1979). Differences between variable and stable environments result in and are reflected in

a variety of acquired features and characteristics among the species inhabiting the

different environments. This is much more prominent in plant species and may be

manifested in greater plasticity as well as greater genetic variation (e.g. Hedrick et al.,

1976; Hedrick, 1986; Via and Lande, 1987). These studies serve to show that genetic

variation is related to the spatial heterogeneity of environments and that phenotypic

plasticity should also be more distinct in environments that are highly variable. Clearly, if

a comparison has to be made, plants from variable environments should be more

phenotypically variable (i.e. have higher phenotypic plasticity) and have relatively high

genetic variability. In a variable environment, each individual in a genotype has to adapt

to a mosaic of diverse biophysical conditions that are dissimilar in both time and space.

There would likely be more than one selective pressure per variable factor in a variable

10

Chapter 1

environment giving rise to the selection for a variety of different traits in a population.

This would, in turn, increase phenotypic plasticity (Via and Lande, 1987).

Phenotypic plasticity of Acacia karroo

Acacia karroo is the most widespread Acacia species in southern Africa and

displays remarkable phenotypic plasticity over its geographical range (Ross, 1975). It

grows variously as a multi-stemmed shrub, a slim and thinly branched shrub, or a tree up

to 40 m in height (e.g. Ross, 1975) and occurs from sea level to an altitudinal limit of

1524 m in the Drakensberg Mountains (Gordon-Gray and Ward, 1975) with the frost-line

as its upper limit. Dry thornveld, river valley scrub, bushveld woodland, grassland, the

banks of dry water courses, river banks of perennial courses, coastal dunes and coastal

scrub are all within the extensive range of different habitat types that A. karroo occupies

(Ross, 1979; Goldsmith and Carter, 1981, Acocks, 1988). The species can grow on

shallow soils of very low water-holding capacity and on weathering bedrock within 400

mm of surface (Hansley and Laker, 1979; Oballa, 1993). Additionally, it has high

tolerance to arsenic soils, which has attracted huge interest in its use for revegetating

abandoned old mining sites (Wild, 1974; Oballa, 1993). Because of its wide geographical

range, the species is subject to diverse climatic conditions and influences. On the

southern and south-western coast of South Africa, it is found within a Mediterranean

climate with winter rainfall. However, in other parts of South Africa, A. karroo grows in

a sub-tropical climate with summer rainfall (Barnes et al., 1996). According to Ross

(1979), there are eight major phenotypic varieties of A. karroo, which can be classified as

entities. More recently, Swartz (1982) considered these entities to perhaps be at the level

11

Chapter 1

of sub-species and even species (see Coates Palgrave, 2002). However, there is still

confusion in distinguishing A. karroo from its variants in the field. The eight variations of

the species are listed below. Note should be taken of the fact that the description here is

summarized following Oballa (1993). For detailed descriptions, see Ross (1971; 1979) or

Acocks (1988):

(1) The typical form of A. karroo grows in the arid Karoo region and drier

parts of the Cape Province as a shrub or tree with dark bark, with usually

(1) 2-3 (5) pinnae pairs per leaf.

(2) A form referred to as A. natalitia (Meyer, 1836) is a small-to-medium

sized tree with white bark, monoliform pods, and 4-7 (13) pinnae pairs per

leaf (von Breitenbach, 1989).

(3) Small slender shrubs up to 1 m high found in the Kei River.

(4) A fire-resistant shrubby form grows widely in the Nongoma District of

South Africa (A. inconflagrabilis).

(5) Slim, scarcely-branched trees up to 6 m occurring in the Hluhluwe and

Umfolozi Game Reserve (“spindle A. karroo”). These trees possess bright

reddish-brown minutely-flaking bark, glaucous foliage, with or without

large flattened petiolar glands.

(6) Large trees of A. karroo with greyish-white bark, with spines up to

25 cm long and long moniliform pods as described by Gordon-Gray and

Ward (1975), and distributed along the KwaZulu-Natal Coast, from the

Thukela River Mouth to Mozambique.

12

Chapter 1

(7) The A. karroo from Pretoria (Tswane) eastwards, characterized by sparse

indumentum on the young branchlets, leaves, peduncles and pods (A.

karroo var. transvaalensis).

(8) The shrubby form which resembles A. tenuispina, but lack spinulose-

mucronate leaflet apices and glandular pods.

Phenology and breeding system of Acacia karroo

The biology of the species is well documented. Phenologically, A. karroo is

deciduous apart from coastal and other communities in frost-free areas where it has

leaves throughout the year according to Teague (1988). He also reported that the optimal

day and night temperatures for growth of the species are between 25oC and 35oC and,

10oC and 15oC, respectively. The leaf growth is initiated at the beginning of the season

(September-October) by the emergence of two to three leaves at each node, depending on

the availability of subsoil moisture. Shoots start growing once leaves have reached their

full size. Flowering starts in October and can continue until January-February, depending

on the conditions of the environment (Poynton, 1984; Teague, 1988). The seeds of A.

karroo are contained in pods, which develop once they are filled with seeds because it is

believed that this delay protects against animals feeding on them (Coe and Coe, 1987).

Investigations on the breeding system have revealed that the species is zoomophilous; the

main pollinators are insects (Gordon-Gray and Ward, 1975; Coe and Coe, 1987). In

addition, there is evidence that the species exhibits sexual reproduction, and is mainly an

outcrosser as it is self-incompatible and is tetraploid with 2n = 4x = 52 chromosomes

(Oballa, 1993; Oballa and Olng’otie, 1994). In its natural range, A. karroo regenerates by

13

Chapter 1

coppicing and seed; however, the most frequent mode is by seed (Teague, 1988). Pods

dry when they reach the maturity, and then release seeds that are dispersed either by wind

or animals (Milton, 1987). Seeds of acacias possess hard coats, which need to be broken

to allow their germination (Doran et al., 1983). Dormancy is broken by micro-organisms,

soil acids, temperature fluctuations and herbivores that feed on the pods (e.g. Miller,

1993). It has also been found that bushfires sometimes help in scarification of seeds of A.

karroo (Coates-Palgrave, 1977); after fire, seeds germinate in 2 to 8 days. This gives the

species the opportunity to develop as pioneer or secondary species (e.g. Weisser and

Muller, 1983).

14

Chapter 1

Morphological and genetic variation in Acacia karroo

The remarkable phenotypic plasticity of A. karroo throughout the contrasting

environments in which it occurs has attracted the attention of researchers such as

Archibald and Bond (2003) who studied its architecture and allometry in forest, savanna,

and arid environments. Their investigation of the effects on the phenotype of

environmental factors other than light, has shown that adult trees in open environments

evolved different forms to those in forests. Furthermore, trees exposed to mammal

browsing differ in form from those exposed to frequent burning.

Swartz (1982) examined six purported variants of A. karroo using 19

morphological characters. In this study, she claimed that the results of her Principal

Component Analysis (PCA) of these characters supported the differentiation of these

taxa, perhaps at the subspecies level. She later (in Coates Palgrave, 2002) raised these

taxa to the level of species, without any indication that new research had been done in the

interim. Examination of the original work (Swartz, 1982) indicates that there is much

overlap between these variants and that cluster analyses revealed inconsistent groupings

of individuals among the groups she recognised. Unfortunately, no overall analysis (PCA

or cluster analysis) was performed. While we recognize that her study reveals interesting

variation among geographic groupings of A. karroo populations, it is difficult to assess

whether such variation represents anything more than broad phenotypic variability with

some local geographic variation.

Brain (1985; 1989; 1996) found variation in a leaf peroxidase among populations

of different environments, as did Oballa (1993) in a study using 12 enzyme systems

(allozymes). Brain suggested the existence of distinct geographic races and the

15

Chapter 1

correlation of isozyme phenotypes with environmental factors such as low temperature

and rainfall. Oballa’s study revealed that all populations surveyed expressed a high level

of genetic diversity. Unweighted pair-group and rooted dendrogram analyses clustered

the populations into three phylogenetic groups, which he characterized as the northern,

eastern and south-central-eastern groups. He also found that the distributions of some

common alleles at the shikimate dehydrogenase and alcohol dehydrogenase loci were

significantly correlated with some geographical factors, viz. latitude, longitude and

rainfall.

Despite the attempts by Swartz (1982), Archibald and Bond (2003), Brain (1996)

and others to shed light on striking phenotypic variations in A. karroo and associated

population differentiation, no conclusive experiments have been conducted to establish

whether this phenomenon is merely phenotypic variation due to environmental variance

or whether such apparent plasticity represents adaptation to different habitats (i.e.

adaptive phenotypic plasticity). No experiments where proper quantitative genetic

approaches involving the growing of full sib or half sib families within populations from

various environments in controlled conditions (such as common-garden experiments)

have been carried out for the species. Additionally, these studies should be combined

with population genetic studies, which would focus on a larger number of loci (see e.g.

Hamrick and Godt, 1989) than was the case in Brain’s (1985; 1989; 1996) studies.

16

Chapter 1

General aims and methodology

Objectives

The main objectives of the study were to establish whether the outstanding

phenotypic variations displayed in A. karroo between populations at two extremes of its

distribution are merely due to environmental variance or whether such plasticity

represents adaptation to different habitats (known as adaptive phenotypic plasticity).

To achieve these objectives, the study sought answers to two questions:

1. Are there genetic differences between populations?

To elucidate these differences, nine enzyme systems were assayed using starch gel

electrophoresis. The specific objective was to determine allozyme variation within and

between sampled populations (genetic population differentiation).

2. Are the observed phenotypic differences adaptive?

A common-garden experiment, in which seeds from different environments were

grown at two different levels of water availability and browsing by mammals, assessed

whether these genetic differences affect phenotypic plasticity and whether this plasticity

is potentially adaptive in their original environments

Study sites

Two populations of A. karroo likely to be under different selective pressures (that

is, climatic and environmentally different conditions consequently causing variability in

their phenotypes) for local adaptation were chosen for this study: one population in a

17

Chapter 1

semi-arid environment (Karoo) and the other one in subtropical coastal forest (KwaZulu-

Natal) (Figure 2).

In semi-arid environments such as in the Karoo, where plants encounter difficult

conditions such as limited resources and heavy herbivory, plants have slow growth and

high amounts of physical and chemical defences (e.g. Coley et al., 1985; Cooper and

Owen-Smith, 1985; Milewski et al., 1991; Milton, 1991). Attention was focused on the

shrubby Leeu Gamka population of A. karroo (32.3oS, 22.3oE), which faces heavy

grazing by livestock (sheep, cattle and goats) and wild herbivores such as kudu

Tragelaphus strepsiceros. This population also receives a low mean annual rainfall of

150 mm and experiences minimum monthly temperatures of 5˚C and mean maximum of

32˚C (Archibald and Bond, 2003).

In tropical or subtropical forest environments, where there are unlimited resources

and nonexistent herbivory by large mammals, plants have rapid growth and generally

display low amounts of physical and chemical defences (e.g. Coley et al., 1985;

Archibald and Bond, 2003). The dune coastal forest population of Richards Bay in the

east coast of KwaZulu-Natal (28.46˚S, 32.06˚E), which experiences a mean annual

rainfall of 1200 mm and respective mean minimum and maximum monthly temperatures

of 10˚C and 28˚C (Archibald and Bond, 2003) formed the second component of the

comparative study. The dominant forms at this location are trees reaching heights over 30

m and which face low or nonexistent grazing (Brain et al., 1996; Boyes, 2004), which

makes them very good candidates for such a study.

18

Chapter 1

Expected results and outcomes from this study were as follows:

(1) Populations should be genetically divergent because, according to Volis et al.

(2001), “under natural selection, populations or groups of populations that are

distant in space but environmentally alike should be genetically more similar than

those that are distant both spatially and environmentally”. This implies that the

likelihood of having similar genotypes should be higher in environmentally

similar populations or groups of populations as opposed to environmentally

dissimilar populations.

(2) The Leeu Gamka population should have a higher genetic variation because this

environment (Leeu Gamka) is highly variable (in the sense of form of selection as

mentioned in page 17) in comparison with Richards Bay, which is in a more

stable environment (see e.g. Hedrick et al., 1976; Hedrick, 1986). Furthermore,

the population with the higher genetic variation should be more variable in their

plasticity. This can be assessed by comparing the coefficients of variation of each

trait within populations.

(3) Both populations should have distinct strategies of investment; the Leeu Gamka

population is expected to invest more in defence and Richards Bay population in

growth. Therefore, there should be trade-offs between growth- and defence-

related traits (e.g. Coley et al., 1985).

19

Chapter 1

Figure 1: Map showing location of study sites (Richards Bay and Leeu Gamka) and other sites listed in the text.

20

Chapter 1

References Acocks, J.P.H., 1988. Veld types of South Africa. 3rd Edition. Mem. Bot. Surv. S. Afr.

57 Botanical Research Institute, South Africa. pp. 146.

Agrawal, A.A., Laforsch, C., Tollrian, R., 1999. Transgenerational induction of defences

in animals and plants. Nature 401, 60-63.

Archibald S., Bond, W.J., 2003. Growing tall vs. growing wide: tree architecture and

allometry of A. karroo in forest, savanna, and arid environments. Oikos 102, 3-14.

Alpert, P., Holzapfel, C., Slominski, C., 2003. Differences in performance between

genotypes of Fragaria chiloensis with different degrees of resource sharing. J.

Ecol. 91, 27-35.

Alpert, P., Simms, E.L., 2002. The relative advantages of plasticity and fixity in different

environments: when is it good for a plant to adjust? Evol. Ecol. 16, 285-297.

Baker, H.G., 1974. The evolution of weeds. Ann. Rev. Ecol. Syst. 5, 1-24.

Barnes, R.D., Filer, D.L., Milton S.J., 1996. Acacia karroo: monograph and annotated

bibliography. Trop. For. Pap. 32. Oxford Forestry Institute, UK.

Bilbrough, C.J., Caldwell, M.M., 1997. Exploitation of springtime ephemeral N pulses by

six Great Basin species. Ecology 78, 231-243.

Bonnier, G., 1920. New observations on experimental cultures at various altitudes and

cultures by sowing. Rev. Gen. Bot. 32, 305-326. [Nouvelles observations sur les

cultures expérimentales à diverses altitudes et cultures par semis.]

Botto, J.F., Smith, H., 2002. Differential genetic variation in adaptive strategies to a

common environmental signal in Arabidopsis accessions: phytochrome-mediated

shade avoidance. Plant Cell Env. 25, 53-63.

21

Chapter 1

Boyes, L.J., 2004. Regeneration failure and the Acacia karroo successional pathway in

coastal dune forest in KwaZulu-Natal. Unpubl. BSc (Hons) thesis, University of

KwaZulu-Natal, Pietermaritzburg.

Brain, P., 1985. Leaf peroxidase types in Acacia karroo. Geographical distribution and

influence of the environment. S. Afr. J. Bot. 52, 47-52.

Brain, P., 1989. Genetic races in a ring species of Acacia karroo. S. Afr. Sci. 85, 181-

185.

Brain, P., Harris, S.A., Barnes, R.D., 1996. Leaf peroxidase types in Acacia karroo

Hayne (Acacieae, Leguminosae): a range-wide study. Silvae Genet. 46, 88-94.

Bradshaw, A.D., 1965. Evolutionary significance of phenotypic plasticity in plants. Adv.

Genet. 13, 115-155.

Briggs, D., Walters, S.M., 1984. Plant variation and evolution. 2nd Edition. Cambridge

University Press, Cambridge, UK.

Briggs, D., Walters, S.M., 1997. Plant variation and evolution. 3rd Edition. Cambridge

University Press, Cambridge, UK.

Butcher, P.A., 2002. Molecular breeding of tropical trees. In: Rimbawanto, A., Susanto,

M. (Eds.), Proceedings of the international conference on advances in genetic

improvement of tropical trees species, 1-3 October 2002. Yogyakarta, Indonesia.

pp, 59-66.

Caldwell, M.M., Manwaring, J.H., Jackson, R.B., 1991. Exploitation of phosphate from

fertile soil microsites by three Great Basin perennials when in competition. Funct.

Ecol. 5, 757-764.

22

Chapter 1

Chapin, F.S. III, Shaver, G.R., 1996. Physiological and growth responses of arctic plants

to a field experiment simulating climatic change. Ecology 77, 822-840.

Cheliak, W.M., Pitel, J.A., 1984. Techniques for starch gel electrophoresis of enzymes

from forest tree species. Petawawa National Forestry Institute, Canada.

Coates-Palgrave, K., 1977. Trees of southern Africa. Struik, Cape Town. pp. 59.

Coates-Palgrave, K., 2002. Trees of southern Africa. 3rd Edition, Struik, Cape Town. pp.

19, 286.

Coe, M., Coe, C., 1987. Large herbivores, acacia trees and bruchid beetles. S. Afr. J. Sci.

83, 624-635.

Coley, P.D., Bryant, J.P. and Chapin, F.S. III, 1985. Resource availability hypothesis and

plant antiherbivore defence. Science 230, 895-899.

Cook, C.D.K., 1974. Waterplants of the world. The Hague: Junk.

Cooper, S.M., Owen-Smith, N., 1985. Condensed tannins deter feeding by browsing

ungulates in a South African savanna. Oecologia 67, 142-146.

Corn, C.A., Hiesey, W.M., 1973. Altitudinal variation in Hawaiian Metrosideros. Am. J.

Bot. 60, 991-1002.

Counts, R.L., 1993. Phenotypic plasticity and genetic variability in Zizania spp. along a

latitudinal gradient. Can. J. Bot. 71, 145-154.

Doran, J.C., Turnbull, J.W., Boland, D.J., Gunn, B.V., 1983. Hand book of seeds of dry-

zone acacias. FAO, Rome. pp. 92.

Fisher, M., Husi, R., Prati, D., Peintinger, M, van Kleunen, M., Schmid, B., 2000. RAPD

variation among and within small and large populations of the rare clonal plant

Ranunculus reptans (Ranunculaceae). Am. J. Bot. 87, 1128-1137.

23

Chapter 1

Freeman, S., Herron, J.C., 2001. Evolutionary analysis, 2nd Edition. Prentice Hall, Upper

Saddle River, NJ.

Godt, M.J.W., Hamrick, J.L., 1998. Allozyme diversity in the endangered pitcher plant

Sarracenia ssp. alabamensis (Sarraceniaceae) and its close relative S. rubra ssp.

rubra. Am. J. Bot. 85, 802-810.

Goldsmith, B., Carter, D.T., 1981. The indigenous timbers of Zimbabwe. Forest

Commission, Zimbabwe. Res. Bull. 9, 9.

Gordon-Gray, K.D., Ward, C.J., 1975. A contribution to the knowledge of floral variation

in Acacia karroo in eastern South Africa. Boissiera 24, 279-284.

Grant, W.S., 1989. Animal protein electrophoresis: Laboratory methods and statistics.

University of Witwatersrand Press, Johannesburg.

Grime, P., 1979. Plant strategies and vegetation process. Wiley and Sons, Chichester.

Hamrick, J.L., Godt, M.J.W., 1989. Allozyme diversity in plant species. In: Brown,

A.H.D., Clegg, M.T., Kahler, A.L., Weir, B.S. (Eds), Plant population, breeding,

and genetic resources. Sinauer Associates, Sunderland, Massachusetts, pp. 43-63.

Hedrick, P.W., Ginevan, M.E., Ewing, E.P., 1976. Genetic polymorphism in

heterogeneous environments. Annu. Rev. Ecol. Syst. 7, 1-32.

Hedrick, P. W., 1985. Genetics of populations. Jones and Bartlett, Inc., Boston. Partolla

Valley. pp. 629.

Hedrick, P.W., 1986. Genetic polymorphism in heterogeneous environments: a decade

later. Annu. Rev. Ecol. Syst. 17, 535-566.

Hansley, M., Laker, M.C., 1979. Land resources of Ciskei. In: Laker, M.C. (Ed), The

agricultural potential of Ciskei-Amended Report. University of Fort Hare, Alice.

24

Chapter 1

Jasienski, M., Francisco, J.A., Bazzaz, F.A., 1996. Phenotypic plasticity and similarity of

DNA among genotypes of an annual plant. Heredity 78, 176-181.

Jefferies, R.L., 1984. The phenotype: its development, physiological constraints, and

environmental signal. In: Dirzo, R., Sarukhan, J., (Eds), Plant Population

Ecology. Sinauer Associates, Sunderland, Massachusetts. pp. 347-358.

Johannsen, W., 1909. Elemente der exakten Erblichkeitslehre. Fisher, Jena.

Joly, H.I., Zeh-Nlo, M., Danthu, P., Aygalent, C., 1992. Population genetics of an

African acacia, Acacia albida. I. Genetic diversity of populations from West

Africa. Aust. J. Bot. 40, 59-73.

Kerner, A., 1895. The natural history of plants, their forms, growth, reproduction and

distribution. Blackie, London.

Knight, T.M., Miller, T.E., 2004. Local adaptation within a population of Hydrocotyle

bonariensis. Evol. Ecol. Res. 6, 103-114.

Kuiper, D., Kuiper, P.J.C., 1988. Phenotypic plasticity in a physiological perspective.

Oecol. Plant 9, 43-59.

Lagercrantz, U., Ryman, N., 1990. Genetic structure of Norway spruce (Picea abies):

concordance of morphological and allozymic variation. Evolution, 44, 38-53.

Leiss, K.A., Muller-Scharer, H., 2001. Performance of reciprocally sown populations of

Senecio vulgaris from ruderal and agricultural habitats. Oecologia 128, 210-216.

Lewontin, R.C., Hubby, J.L., 1966. A molecular approach to the study of genetic

heterozygosity in natural populations. II. Amount of variation and degree of

heterozygosity in natural populations of Drosophila pseudoobscura. Genetics 54,

595-609.

25

Chapter 1

Li, W. H., Graur, D., 1991. Fundamentals of molecular evolution. Sinauer Associates,

Sunderland, Massachussetts. pp. 284.

Linhart, Y.B., Grant, M.C., 1996. Evolutionary significance of local genetic

differentiation in plants. Annu. Rev. Ecol. Syst. 27, 237-277.

Lonn, M., Prentice, H.C., 1990. Mosaic variation in Swedish Petrorbagia prolifera: the

partitioning of morphometric and electrophoretic diversity. Biol. J. Linn. Soc. 41,

353-373.

Lynch, M., Walsh, B., 1997. Genetics and analysis of quantitative traits. Sinauer

Associates, Sunderland, Massachusetts.

MacGraw, J.B., Antonovics, J., 1983. Experimental ecology of Dryas octopetala

ecotypes. 1. Ecotypic differentiation and life-cycle stages of selection. J. Ecol. 71,

879-897.

Macdonald, E.S., Chinnappa, C.C., 1989. Population differentiation for phenotypic

plasticity in the Stellaria longipes complex. Am. J. Bot. 76, 1627-1637.

Meyer, E.H.F., 1836. Acacia natalitia. Comm. Pl. Afr. Austr. 1, 167.

Milewski, A.V., Young, T.P., Madden, D., 1991. Thorns as induced defences:

Experimental evidence. Oecologia 86, 70-75.

Miller, M.F., 1993. Is it advantageous for Acacia seeds to be eaten by ungulates? Oikos

66, 365-368.

Milton, S.J., 1987. Phenology of seven Acacia species in South Africa. S. Afr. J. Wildl.

Res. 17, 1-6.

Noy-Meir, I., 1973. Desert ecosystems: environment and producers. Annu. Rev. Ecol.

Syst. 4, 25-41.

26

Chapter 1

Novoplansky, A., 2002. Developmental plasticity in plants: implications of non-cognitive

behavior. Evol. Ecol. 16, 177-188.

Oballa, P.O., 1993. Genetic variation within Acacia karroo Hayne. Unpubl. D. Phil.

Oxford University, Oxford England.

Oballa, P.O., Olng’otie, P.A.S., 1994. Chromosome numbers in two African Acacia

species. Kew Bulletin 49, 107-113.

Parker, P.G., Snow, A.A., Schug, M.D., Booton, G.C., Fuerst, P.A., 1998. What

molecules can tell us about populations: choosing and using a molecular marker.

Ecology 79, 361-382.

Paschke, M., Abs, C., Schmidt, B., 2002. Relationaship between population size,

allozyme variation, and plant performance in the narrow endemic Cochlearia

bavarica. Conserv. Genet. 3, 131-144.

Piersma, T., Lindstrom, A., 1997. Rapid reversible changes in organ size as a component

of adaptive behaviour. Trends Ecol. Evol. 12, 134-138.

Platenkamp, G.A., Shaw, R.G., 1992. Environmental and genetic constraints on adaptive

population differentiation in Anthoxanthum odoratum. Evolution 46, 341-352.

Poynton, R.J., 1984. Characteristics and uses of selected trees and shrubs cultivated in

South Africa. Government Printer, Pretoria. pp. 201.

Quattro, J.M., Vrijenhoek, R.C., 1989. Fitness differences in remnant populations of the

endangered Sonoran topminnow. Science 245, 976-978.

Reed, D.H., Frankham, R., 2001. How closely correlated are molecular and quantitative

measures of genetic variation? A meta- analysis. Evolution 55, 1095-1103.

27

Chapter 1

Ridley, M., 1996. Evolution: Adaptive explanation. 2nd Edition. Blackwell Science, Inc,

Atlanta, Georgia. pp 338-368.

Ritland, K., Jain, S., 1984. A comparative study of floral and electrophoretic variation

with life history variation in Limnanthes alba (Limnanthaceae). Oecologia 63,

243-251.

Ross, J.H., 1971. Acacia karroo in Southern Africa. Bothalia 10, 385-401.

Ross, J.H., 1975. Notes on African Acacia species. Bothalia 11, 443-447.

Ross, J.H., 1979. A conspectus of the African Acacia species. Mem. Bot. Surv. S. Afr.

44, 93-95.

Schlichting, C.D., Levin, D.A., 1984. Phenotypic plasticity of annual Phlox: tests of some

hypotheses. Am. J. Bot. 71, 252-260.

Schlichting, C.D., 1986. The evolution of phenotypic plasticity in plants. Annu. Rev.

Ecol. Syst. 17, 667-693.

Schlichting, C.D., Pigliucci, M., 1998. Phenotypic evolution, a reaction norm perspective.

Sinauer Associates, Sunderland, Massachusetts.

Schlichting, C.D., Smith, H., 2002. Phenotypic plasticity: linking molecular mechanisms

with evolutionary outcomes. Evol. Ecol. 16, 189-211.

Schmalhausen, I.I., 1949. Factors of evolution: the theory of stabilizing selection.

University of Chicago Press, Chicago.

Schmid, B., 1992. Phenotypic variation in plants. Evol. Trends Plants 6, 55-60.

Schmidtling, R.C., 1994. Use of provenance tests to predict response to climate change:

loblolly pine and Norway spruce. Tree Physiol. 14, 805-817.

28

Chapter 1

Schwaegerle, K.E., 1996. Evolutionary response. 2nd Edition. In: Molau, U., Molgaard,

P. (Eds), International Tundra Experiment (ITEX) manual. Danish Polar Center,

Copenhagen. pp 49-50.

Shaver, G.R., Fetcher, N., Chapin, F.S. III, 1986. Growth and flowering in Eriophorum

vaginatum: annual and altitudinal variation. Ecology 67, 1524-1535.

Shrestha, M.K., Golan-Goldhirsh, A., Ward, D., 2002. Population genetic structure in

isolated populations of Acacia raddiana investigated by random amplified

polymorphic DNA (RAPD) markers. Biol. Conserv. 108, 119-127.

Smekens, M.J., van Tienderen, P.H., 2001. Genetic variation and plasticity of Plantago

coronopus under saline conditions. Acta Oecol. 22, 187-200.

Soltis, D.E., Soltis, P.S., 1989. Isozymes in plant biology. Department of Botany,

Washington State University.

Sork, V.L., Stowe, K.A., Hochwender, C., 1993. Evidence for local adaptation in closely

adjacent subpopulations of northern red oak (Quercus rubra L.) expressed as

resistance to leaf herbivores. Am. Nat. 142, 928-936.

Stearns, S.C., 1992. The evolution of life histories. Cambridge University Press,

Cambridge, UK.

Sultan, S.E., 1987. Evolutionary implications of phenotypic plasticity in plants. Evol.

Biol. 21, 127-178.

Sultan, S.E., 1992. What has survived of Darwin’s theory? Evol. Trends Plants 6, 61-71.

Sultan, S.E., Bazzaz, F.A., 1993. Phenotypic plasticity in Polygonium persicaria. III. The

evolution of ecological breadth for nutrient environment. Evolution 47, 1050-

1071.

29

Chapter 1

Swartz, P.P., 1982. A numerical-taxonomical evaluation of the taxon Acacia karroo

Hayne in southern Africa. Unpl. MSc thesis, University of Pretoria South Africa.

[‘n Numeries-taksonomiese evaluering van die takson Acacia karroo Hayne in

suidelike Afrika.]

Teague, W.R., 1988. The response of Acacia karroo plants to defoliation by hand

compared to defoliation by goats. J. Grassl. Soc. Sth. Afr. 5, 122-124.

Turesson, G., 1961. Habitat modifications in some widespread plant species. Bot. Not.

114, 435-452.

Turkington, R., Harper, J.L., 1979. The growth, distribution and neighbour relationships

of Trifolium repens in a permanent pasture. J. Ecol. 67, 245-254.

van Kleunen, M., Fisher, M., Schmid, B., 2000. Costs of plasticity in foraging

characteristics of the clonal plant Ranunculus reptans. Evolution 54, 1947-1955.

Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S.M., Schlichting, C.D., Van

Tienderen, P., 1995. Adaptive phenotypic plasticity: consensus and controversy.

Trends Ecol. Evol. 10, 212-217.

Via, S., Lande, R., 1987. Evolution of genetic variability in a spatially heterogeneous

environment: effects of genotype-environment interaction. Genet. Res. 49, 147-

156.

Volis, S., Shulgina, I., Ward, D., Mendlinger, S., 2003. Regional subdivision in wild

barley allozyme variation: adaptive or neutral? J. Hered. 94, 341-351

von Breitenbach, F., 1989. The lost Natal thorn for a reinstatement of Acacia natalia E.

Mey. J. Dendrol. 12, 1-10.

30

Chapter 1

Walters, M., Milton, S.J., 2003. The production, storage and viability of seeds of Acacia

karroo and A. nilotica in a grassy savanna in KwaZulu-Natal, South Africa. Afr.

J. Ecol. 41, 211-217.

Weisser, P.J., Muller, R., 1983. Dune vegetation dynamics from 1937-76 in Mlalazi-

Richards Bay of Natal, South Africa. Bothalia 14, 661-667.

Wild, H., 1974. Arsenic tolerant plant species established on arsenical mine dumps in

Rhodesia. Geobotanical anomalies in Rhodesia. The vegetation of arsenical soils.

Kirkia 9, 265-278.

Williams, D.G., Mack, R.N., Black, R.N., 1995. Ecophysiology of introduced

Pennisetum setaceum on Hawaii: the role of phenotypic plasticity. Ecology 76,

1569-1580.

Wilson, R.S., 2001. Geographic variation in thermal sensitivity of jumping performance

in the frog Lymnodynastes peronii. J. Exp. Biol. 204, 4227-4236.

Winn, A.A., 1996. Adaptation to fine-grained environmental variation: an analysis of

within-individual leaf variation in an annual plant. Evolution, 50, 1111-1118.

Wolf, P.G., Campbell, D.R., 1995. Hierarchical analysis of allozymic and morphometric

variation in a montane herb Ipomopsis aggregata (Polemoniaceae). J. Hered. 86,

386-394.

Woltereck, R., 1909. Further experimental investigations in relation to specie changes,

particularly over the quantitative traits differences in Daphnia. Attempt. German

zool. Soc. 19, 110-172. [Weitere experimentelle Untersuchungen über

31

Chapter 1

Artveränderung, speziell über das Wesen quantitativer Artunderscheide bei

Daphniden Versuch. Deutsch. Zool. Ges.]

32

Chapter 2

Genetic variation in two extreme populations of phenotypically-plastic

Acacia karroo

Submitted as:

Mboumba, G.B., Ward, D. Genetic variation in two extreme populations of

phenotypically-plastic Acacia karroo. Journal of Forest Ecology and Management.

33

Chapter 2

Abstract

Genetic variation of Acacia karroo was compared by allozyme electrophoresis for

two populations from extreme environments in South Africa: arid Karoo (Leeu Gamka)

and wet coastal forest (Richards Bay). These populations display remarkable phenotypic

differentiation: in the arid Karoo, adult trees are <4 m tall and have large thorns and in

coastal forests they may be as much as 40 m tall and invest little in defence. Of the nine

enzymes analyzed by horizontal starch gel electrophoresis, 10 enzyme-coding loci were

scored. Both populations exhibited higher genetic diversity (Richards Bay, H = 0.45 and

Leeu Gamka, H = 0.27) than in many other species studied to date. Principal Coordinate

analysis (PCoA) revealed little genetic overlap among populations, which is consistent

with the high Nei’s genetic distance value obtained (D = 0.26), as well as the significant

population differentiation (Φst = 0.422, p < 0.001) revealed by AMOVA. The high level

of genetic diversity may be due to the predominantly-outcrossing breeding system of the

species, and the degree of differentiation probably to their geographic isolation as well as

local adaptation to their environments.

Keywords: Acacia karroo, genetic variation, allozyme, starch gel electrophoresis.

34

Chapter 2

Introduction

Phenotypic diversity in widespread plant species is considered as an adaptation to

the variable environmental conditions in which they live (e.g. Bradshaw, 1965; Via et al.,

1995). The need to understand whether this diversity is the result of local adaptation is of

particular importance as it will provide insights into the ability of populations to adapt to

particular environments as well as predict their fate in the long-term. The ability of

species to occupy a variety of habitats has also been linked to genetic diversity as it is the

raw material of evolutionary changes (e.g. Pigliucci and Schlichting, 1997); there may be

much hidden genetic diversity in phenotypically-plastic species. From a conservation

viewpoint, it is judicious to differentiate adaptive phenotypic plasticity, which implies

that individuals differ genetically and that there is local adaptation, and phenotypic

plasticity, which is the ability of genotypes to modify their morphologies and

physiologies to changing environmental conditions (Schlichting, 1986). In terms of

conservation of genotypes, those populations that are genetically differentiated and that

exhibit significant genotype-by-environment interactions (i.e. are locally adapted), should

be conserved separately. It might be said that the objective of conservation is to preserve

genotype-by-environment interactions rather than populations or species per se.

There is abundant evidence that genetic differentiation is low in widely-

distributed plant species (Hamrick et al., 1981; Hamrick and Godt, 1989) while in small

isolated populations it is most likely to be high as it is enhanced by genetic drift (Moran

and Hopper, 1987), and selection in local environments. Contrastingly, genetic diversity

should be higher in widely-distributed plant species than in small isolated populations

(see e.g. Hamrick and Godt, 1989; 1998; Hamrick et al., 1992; Kang and Shang, 2000).

35

Chapter 2

Allozymes

The assessment of genetic diversity at the molecular level has rapidly improved

over the last 20 years, allowing us to answer questions relating to breeding system,

speciation, taxonomic boundaries, dispersal ability and so forth (Jeffrey and Gavin,

2000). There are three major kinds of genetic markers: (1) morphological markers that

display Mendelian inheritance (e.g. chlorophyll deficiency), (2) biochemical markers at

the protein level (e.g. isozymes, also known as allozymes) and, more recently (3) DNA-

based markers (Jeffrey and Gavin, 2000). The choice of a technique depends on the aims

of the study, the sensitivity and convenience of the technique and, of course, the

availability of resources (Oballa, 1993).

Allozymes are variants of the same enzyme encoded by different alleles at the

same locus. Because of amino acid charge differences, allozymes can be differentiated by

their relative migration speed during gel electrophoresis. Many enzymes are invariant

within populations (or even between species), and most polymorphic enzymes have only

a few variants. Although this limits the power of allozyme analysis to resolve genetic

diversity, they are time- and cost-efficient for research (Mueller and Wolfenbarger,

1999). Although less informative than DNA sequencing, randomly amplified

polymorphic DNA (RAPDs), and restriction fragment length polymorphisms (RFLPs),

allozymes were chosen for the ease with which one can address questions dealing with

genetic variation and for the low cost. Despite their lower efficiency in detecting

variation than some DNA markers (ca. 30 % - Van Straalen and Timmermans, 2002),

they have been found to be successful (since 1966) in quantifying genetic variation in

36

Chapter 2

many plant populations (see e.g. Brain, 1985 and 1989; Oballa, 1993; Yu et al., 2001;

Mateu-Andres and Segarra-Moragues, 2002).

Phenotypic and genotypic variation in Acacia karroo

Acacia karroo is the most widespread and most phenotypically-plastic Acacia tree

species in southern Africa and occupies a diverse range of habitats including dry

thornveld, river valley scrub, bushveld, woodland, grassland, the banks of dry

watercourses, riverbanks, coastal dunes and coastal scrub (Ross, 1979). There is evidence

that the large phenotypically-plastic variation is regional; plants in various parts of the

species’ geographical range often have different phenotypes (Ross, 1971). The typical

form of A. karroo grows in the arid Karoo and in the drier parts of the Western Cape

Province as a shrub or tree with dark rough bark (Figure 1). Others forms are

recognizable within the species; in the eastern Cape, KwaZulu-Natal, and Mpumulanga

provinces of South Africa, and in adjacent Swaziland, Zimbabwe and Mozambique, they

grow as white-barked trees or shrubs with short spines. Small slender shrubs up to 1 m

high are found in some parts of the Eastern Cape Province. “Fire resistant” shrubs are

found in some regions of northern KwaZulu-Natal province. Slender trees up to 6 m high

with few branches are found in KwaZulu-Natal province north of the Tugela River. Large

trees with greyish-white bark and spines up to 25 cm long are found along the KwaZulu-

Natal coast from the mouth of the Tugela River northwards into Mozambique (Ross,

1979). On the east coast of KwaZulu-Natal, particularly in the Richards Bay area, tall

trees up to 40 m with short thorns dominate the coastal dune forest community. On the

Gauteng highveld from Pretoria (Tshwane) eastwards, there is a local tendency for the

production of a sparse indumentum on the young branchlets, leaves, peduncles and pods.

37

Chapter 2

Finally, on the Springbok Flats (North West Province), they grow as small shrubby plants

(Ross, 1979).

Figure 1: Map showing location of study sites (Richards Bay and Leeu Gamka) and other sites listed in the text.

38

Chapter 2

The variability observed in phenotypes within the range of A. karroo may have a

genetic basis. Brain (1985; 1989), in a study of leaf peroxidase variation in South African

populations of A. karroo, revealed interesting patterns of variation among them,

suggesting the existence of distinct geographic races and the correlation of isozyme

phenotypes with environmental factors such as low temperature and rainfall.

Unfortunately, only one enzyme system was used by Brain (1989), which cannot be

considered a sufficient basis for differentiation among populations (Hamrick and Godt,

1989). A similar correlation of allele distribution and environmental factors in A. karroo

has also been shown for shikimate dehydrogenase and alcohol dehydrogenase loci

(Oballa, 1993). Hedrick et al. (1976) pointed out that when populations are sampled over

space, there is always a clinal pattern in one or more loci that correlates with some

environmental parameter, e.g. rainfall, temperature, or salinity (see also Volis et al.

2001). It is generally considered than 10-20 loci is the minimum for detecting population

differentiation (e.g. Hamrick and Godt, 1989). Provenance trials on A. karroo have also

shown genetic differences in growth form among populations (Barnes et al., 1996).

Archibald and Bond (2003) have shown large phenotypic differences in growth and

defence among A. karroo populations in the wild but, unfortunately, did not differentiate

between phenotypic and genotypic variability.

We attempted to remedy some of the shortcomings of earlier studies of genetic

differentiation in A. karroo by studying a large number of allozyme loci in two

populations from extreme environments. Leeu Gamka (arid Karoo of central South

Africa) and Richards Bay (coastal forest) populations differ considerably in phenotype

39

Chapter 2

(Figure 1). In the arid Karoo, adult trees are <4 m tall and have large thorns and in coastal

forests they may be as much as 40 m tall and invest little in defence.

Leeu Gamka is characterized by a very dry environment with a mean annual

rainfall of 150 mm with winter peak in rainfall, minimum monthly temperature of 5˚C

and maximum 32˚C (Archibald and Bond, 2003). At this location, the dominant forms of

A. karroo are shrubs and face heavy browsing by livestock and kudu Tragelaphus

strepsiceros. At Richards Bay, the environment is subtropical coastal forest (Richards

Bay) with a mean annual rainfall of 1200 mm. Monthly, the minimum and maximum

temperatures are respectively 10˚C and 28˚C (Archibald and Bond, 2003). The dominant

forms are trees reaching heights of over 30 m and face low or nonexistent browsing

(Brain et al., 1996; Boyes, 2004). The aim of the present study was to investigate genetic

variation within and between these two populations using allozymes as markers.

Materials and methods

Population sampling

Seeds were collected from 15 trees in each population. Trees were at least 10 m

apart to ensure that each tree represented a unique individual. Seeds were put in paper

bags and then brought to the laboratory. They were dried at 25˚C in order to remove all

unwanted organisms and kept in the refrigerator at 2-8˚C.

40

Chapter 2

Allozyme electrophoresis

Extraction of enzymes was done using a homogenizing buffer; placed in 400 μl of

0.01 M Tris buffer (pH 8). Seeds were ground, homogenized and centrifuged for 30 s at

12000 rpm. Thereafter, samples were kept in the refrigerator at -80˚C. However, because

of the poor state of most seeds collected, leaves were used for the rest of the study. A

number of enzyme systems were tried, with a focus on enzymes that showed inter-

individual variability between populations. Enzymes were extracted from one seed per

tree for nine trees randomly chosen out of 15 per population. Starch gel (13%)

electrophoresis was used and run with fresh samples either with TC buffer pH 7.5

(following Soltis et al., 1983) or HC pH 6.5 (Cardy et al., 1981) or RW pH 8.1 (Ridgway

et al., 1970) and MF pH 8.1 buffer (Markert and Faulhaber 1965)(enzymes and their

buffer systems are listed in Table 1).

Table 1: Enzymes and buffers used for staining.

Codes Stain EC. NO Gel Buffer Adh Alcohol dehydrogenase 1.1.1.1 Tris-citrate (TC) Dia Diaphorase 1.6.4.3 Histidine-citrate (HC) Est Esterase (alpha) 3.1.1.1 Tris-citrate-borate-lithium hydroxide

(RW) Idh Isocitrate dehydrogenase 1.1.1.42 Tris-borate-EDTA (MF) Mdh Malate dehydrogenase 1.1.1.37 Histidine-citrate (HC) Me Malic enzyme 1.1.1.40 Tris-citrate-borate-lithium hydroxide

(RW) Pgd Phosphogluconate dehydrogenase 1.1.1.44 Tris-citrate (TC) Pgi Phosphoglucose Isomerase 6.3.1.9 Histidine-citrate (HC) Sdh Shikimate dehydrogenase 1.1.1.25 Tris-borate-EDTA (MF)

Proteins were transferred to the gel in two ways: (1) via pipette into combs that

were filled with 20 μl of the sample, and (2) via filter paper wicks that were impregnated

41

Chapter 2

with the supernatant and inserted into the starch gel. The gels were covered with plastic

and run for 4-5 h at 50 mA in the refrigerator (2-8˚C). After electrophoresis, the gels were

cut into thin horizontal slices and each slice was stained using specific protein-staining

solutions. The chemical composition of each staining solution depended on the locus to

be stained (Cheliak and Pitel, 1984; Soltis and Soltis, 1989). After each electrophoresis

run, the allozymes were alphabetically scored (Soltis and Soltis, 1989).

Data analysis

The interpretation of allozyme patterns in Acacia karroo is extremely difficult

because of its ploidy level (tetraploid with 2n = 4x = 52 chromosomes), as indicated by

Oballa (1993) and, Oballa and Olng’otie (1994). Data were treated as phenotypes,

scoring the presence or absence of bands representing alleles (e.g. Brain, 1985, 1989;

Brain et al., 1996; Shrestha et al., 2002). The following parameters indicating genetic

diversity were described:

(1) the proportion of polymorphic loci (% P), obtained by counting the number of

loci with ≥ 2 bands (alleles) for total number of loci within each population;

(2) mean number of alleles per polymorphic locus (AP), and

(3) mean number of alleles per locus (A).

χ2 tests were used to test for significant differences in allele frequencies among

populations. We used POPGENE 1.31 (Yeh et al., 1998) to calculate the Shannon index

of gene diversity (H) and Nei’s genetic distance (D) (1972). Partitioning of genetic

diversity within and among populations was also investigated by an analysis of molecular

variance (AMOVA; WINAMOVA 1.55 program-Excoffier et al., 1992). The multivariate

42

Chapter 2

relationships among individuals and between populations was analyzed using a Principal

Coordinate Analysis (PCoA) using the Multi-Variate Statistical Package (MSVP-

Kovach, 1999).

Results

Within-population polymorphisms

Ten enzyme coding loci were consistently scored. In the two populations

analyzed, we found a total of 32 alleles. Four of these loci were monomorphic in the Leeu

Gamka population (Sdh, Idh, Me and Pgd); the remaining six loci were polymorphic

(Mdh, Dia, Adh, Pgi, α Est-1, and α Est-2). Two loci were monomorphic in the Richards

Bay population (Idh and Pgd); the remaining eight loci were polymorphic (Sdh, Mdh, Me,

Dia, Adh, Pgi, α Est-1, and α Est-2). All loci segregated at most four (rarely five) alleles

in a population (Table 2).

43

Chapter 2

Table 2: Allele frequencies at 10 loci in two populations of A. karroo.

Locus Allele Leeu Gamka Richards Bay Est-1 a 0.300 -

b 0.450 0.462 c 0.100 0.385 d 0.150 0.154

Est-2 a 0.409 0.556 b 0.091 -

c 0.273 0.444 d 0.227 -

Adh a 0.167 0.250 b 0.167 0.250 c 0.333 0.250 d 0.333 0.250

Pgi a - 0.310 b 0.053 0.241

c 0.474 0.276 d 0.474 0.172

Dia a - 0.192 b - 0.231 c 0.500 0.077 d 0.444 0.192 e 0.056 0.269

f - 0.038 Pgd a 1.000 1.000 Me a 0.600 0.462

b 0.400 0.538 Sdh a - 0.600

b 1.000 0.400 Mdh a 0.182 -

b 0.818 0.750 c - 0.250

Idh a 1.000 0.250 b - 0.750

44

Chapter 2

The Leeu Gamka population had lower genetic polymorphism than the Richards

Bay population for all indices measured. For the Leeu Gamka population, the percentage

of polymorphic loci (%P) was 60, mean number of alleles per locus (A) = 2.3, mean

number of alleles per polymorphic locus (AP) = 2.86 and Shannon index (H) = 0.27. In

the Richards Bay population, % P = 80, A = 2.8, AP = 3, and H = 0.45.

Among- population polymorphisms

Of the 32 bands found in both populations, two (namely Pgd and Mdh-b) were

removed from the analysis because they were fixed in both populations. For 17 bands, χ2

tests for heterogeneity were significantly different between populations (p values ranged

from 0.045 - 0.001). Leeu Gamka population had several private alleles, namely Mdh-a,

Est-1a, Est-2a and Est-2b (13.33 % private alleles) while alleles Sdh-a, Mdh-c, Pgi-a,

Dia-a, Dia-b and Dia-f were private to the Richards Bay population (20 % private

alleles). Principal Coordinate Analysis (PCoA; Hogbin and Peakall, 1999) illustrates the

multivariate relationships among individuals as well as among populations (Figure 2).

45

Chapter 2

-1.5

-0.9

-0.3

0.3

0.9

1.5

-3 1

Axis 1

Axi

s 2Leeu Gamka Richards Bay

Figure 2: Principal Coordinate Analysis (PCoA) of Acacia karroo (18 individuals). Axis 1 extracted 37.79% of the variance among individuals and Axis 2 extracted 13.15% of the variance. Note that one individual from Leeu Gamka overlaps on both axes with individuals from Richards Bay. The first two axes extracted a total of 50.94 % of the variation among individuals (Axis

1: 37.79 %; Axis 2: 13.15 %), revealing little genetic overlap among populations.

Similarly, an AMOVA revealed that 42.24% of the total variance occurred among

populations and 57.76% occurred within populations (Table 3). AMOVA showed that

there was significant population differentiation (Φst = 0.422, p < 0.001), which is

consistent with the high Nei’s unbiased genetic distance (D = 0.26).

46

Chapter 2

Table 3: Analysis of molecular variance (AMOVA) within and among Acacia karroo populations Source of variation d.f.

SS

MS

Variance component

Total variance (%)

p

Among populations

1 34.22 34.22 3.30 42.24 <0.001

Within populations

16 72.22 4.51 4.51 57.76

Probability calculated by 1000 random permutations of individuals across populations.

Discussion

Population genetic variation

We found that these two populations diverged considerably, as indicated by the

large genetic distance between them (D = 0.26), which is a function of genetic identity (I

= 0.77). Gottlieb (1977) concluded that most conspecific populations have high mean

genetic identities, above 0.90, while congeneric plant species have reduced genetic

identities, varying from 0.50 to 0.60. This criterion would suggest that the Richards Bay

population has diverged to variety level, given that the Karoo populations are

documented as the “typical form” of the species (e.g. Ross, 1979). Similar results on

populations of A. karroo diverging to subspecies or variety level were pointed out in

Oballa’s study (1993), where values varied between 0.61 and 0.89. Those populations

include Prince Albert (close to Leeu Gamka) and Manzengwenya (close to Richards

Bay), which had a genetic identity of 0.84 in Oballa’s study (1993). The population

differentiation (Φst) in our study was 0.422, indicating a high proportion of genetic

variation due to population differentiation, which is consistent with the large genetic

47

Chapter 2

distance of 0.26. As with other plant species studied by Hamrick and Godt (1989) and

Macdonald et al. (2001), most genetic variation was found within populations, as

indicated by the analysis of molecular variance (57.76% variation within populations)

and which is consistent with the results of the Principal Coordinate analysis (Figure 2).

Interestingly, higher genetic diversity was observed in the Richards Bay

population (H = 0.45) than in Leeu Gamka (H = 0.27). This result is paradoxical because

the more stable forest environment at Richards Bay might be expected to have lower

genetic diversity than the highly variable arid Leeu Gamka environment (see e.g. Volis et

al. 2001; Hedrick et al., 1976; Hedrick, 1986). Both populations displayed very high

gene diversity compared with that found in tropical rainforest trees (mean H = 0.11 -

Hamrick and Loveless, 1986) and in the Australian Acacia species, A. crassicarpa (H =

0.13) and A. auriculiformis (H = 0.15) (Moran et al., 1989). An African species, Acacia

tortilis (Olngótie, 1992; see Oballa, 1993) and the closely-related Faidherbia albida (Joly

et al., 1992) have been shown to have high gene diversity (H = 0.45 in both cases), which

is similar to that recorded in the Richards Bay population. The Leeu Gamka population

had a gene diversity value approximately similar to what has been recorded in conifers (H

= 0.27 - Mitton, 1983) and in Acacia brevispica (H = 0.27, Harsh, 2003). The level of

polymorphism within each population (% P = 80 for Richards Bay and % P = 60 for

Leeu Gamka) was larger than those found in plant species in general (50%) (Hamrick and

Godt, 1989). This high level of genetic diversity is expected from widespread species, as

they occupy various habitats and encounter a variety of biotic and abiotic factors. It is

assumed that the level of genetic variation reflects an ability to adapt to changing

environmental conditions (Pigliucci and Schlichting, 1997). In another study, we (See

48

Chapter 2

Chapter 3) found that the coefficient of variation of a number of growth- and defence-

related traits was higher in Richards Bay than in Leeu Gamka. The reason for this is

uncertain but may reflect differences in spatial and temporal variability between these

habitats (see also Volis et al. 2001).

Recent reviews of the plant allozyme literature have shown that rare, endemic, or

narrowly-distributed plants tend to contain less allozyme variation than species with a

widespread distribution (Hamrick and Godt, 1989; Hamrick et al., 1992; Kang and

Chung, 2000). These results are particularly interesting because the efficiency of

allozymes in detecting variation is often considered to be low (ca. 30 % - Van Straalen

and Timmermans, 2002). Thus, the high population-level diversity recorded in this study

may be explained by the fact that A. karroo is predominantly an outcrosser, even though

it shows a high level of effective selfing (Ross, 1979; Oballa, 1993). In conservation

terms, the high level of genetic diversity observed within these populations is considered

to be encouraging because genetic diversity and fitness are often considered to be

positively correlated (Reed and Frankham 2003).

Divergence among populations has been established to be the result of many

factors, among which are genetic drift due to limited gene flow, breeding system of the

species (e.g. Barrett and Kohn, 1991; Hogbin and Peakall, 1999), and local adaptation via

genotype-by-environment interactions (e.g. Shrestha et al., 2002; Volis et al., 2002a,

2002b, 2002c). The Leeu Gamka population faces heavy grazing by livestock and wild

antelopes and low rainfall while the Richards Bay population occurs in an area of high

rainfall and has low or nonexistent grazing because they are in forest (Boyes, 2004). All

49

Chapter 2

these factors may have an impact on the genetic structure of these populations and may

lead to local adaptation via genotype-by-environment interactions.

In general, our evidence that these extreme populations differ genetically and

show a high level of genetic diversity conforms with the results of a study by Oballa

(1993) where high levels of genetic diversity between A. karroo populations of different

localities were detected, as well as Brain’s studies (1985; 1989) which showed genetic

differences between certain populations in different environments with a single enzyme

system. Inferences can now be drawn on the possible genetic basis of the phenomenal

phenotypic plasticity observed in populations of A. karroo by Archibald and Bond

(2003). Appropriate common-garden and reciprocal transplant experiments will be

needed to establish whether local adaptation is the main cause of population

differentiation (See Chapter 3).

Acknowledgements

We gratefully acknowledge funding for this research provided by the National

Research Foundation (to David Ward) and the Gabonese Government (to Georges

Bayonne Mboumba). Nir Peleg, Heidi Thunemann, Cara Nieuwoudt for technical

assistance in the laboratory. We especially thank Gavin Gouws and Sergei Volis for

helpful discussions.

50

Chapter 2

References

Archibald S., Bond, W.J., 2003. Growing tall vs. growing wide: tree architecture and

allometry of A. karroo in forest, savanna, and arid environments. Oikos 102, 3-14.

Barrett, S.C.H., Kohn, J.R., 1991. Genetics and evolutionary consequences of small

population size in plants: Implications for conservation. Falk, D. A. and Holsinger

K. E. (Eds). Genetics and conservation of rare plants. Oxford University Press.

New York, USA, pp. 3-30.

Barnes, R.D., Filer, D.L., Milton S.J., 1996. Acacia karroo: monograph and annotated

bibliography. Trop. For. Pap. 32. Oxford Forestry Institute.

Oxford, UK.

Boyes, L.J., 2004. Regeneration failure and the Acacia karroo successional pathway in

coastal dune forest in KwaZulu-Natal. Unpubl. BSc (Hons) thesis, University of

KwaZulu-Natal, Pietermaritzburg.

Bradshaw, A.D., 1965. Evolutionary significance of phenotypic plasticity in plants. Adv.

Genet. 13, 115-155.

Brain, P., 1985. Leaf peroxidase types in Acacia karroo. Geographical distribution and

influence of the environment. S. Afr. J. Bot. 52, 47-52.

Brain, P., 1989. Genetic races in a ring species of Acacia karroo. S. Afr. J. Sci. 85, 181-

185.

Brain, P., Harris, S.A., Barnes, R.D., 1996. Leaf peroxidase types in Acacia karroo

Hayne (Acacieae, Leguminosae): a range-wide study. Silvae Genet. 46, 88-94.

51

Chapter 2

Cardy, B.J., Stuber, C.W., Goodman, M.M., 1981. Techniques for starch gel

electrophoresis of enzymes from maize (Zea mayz L.). Institute of Statistics

Mimeograph Series No 1317. North Carolina State University, Raleigh, NC.

Cheliak, W.M., Pitel, J.A., 1984. Techniques for starch gel electrophoresis of enzymes

from forest tree species. Petawawa National Forestry Institute, Canada.

Excoffier L, Smouse PE, and Quattro JM (1992) Analysis of molecular variance inferred

from metric distances among DNA haplotypes : Application to human

mitochondrial DNA data. Genetics 131:479.491.

Hamrick, J.L., Mitton, J.B., Linhart, Y.B., 1981. Levels of genetic variation in trees:

influence of life history characteristics. In: Conkle, M.T. (Ed.), Proceeding of the

symposium on isozymes of North American forest trees and forest insects. July

27, 1979, Berkeley, California. pp. 35-41.

Hamrick, J.L., Godt, M.J.W., 1989. Allozyme diversity in plant species. In: Plant

population, breeding, and genetic resources. Brown, A.H.D., Clegg, M.T., Kahler,

A.L., Weir, B.S. (Eds). Sinauer Associates, Sunderland, Massachusetts, pp. 43-63.

Hamrick, J.L., Godt, M.J.W., Sherman-Broyles, S.L., 1992. Factors influencing levels of

genetic diversity in woody plant species. New Forests 6, 95-124.

Hamrick, J.L., Loveless, M.D., 1986. Isozyme variation in tropical trees: Procedures and

preliminary results. Biotropica 18, 201-207.

Harsh, J., 2003. Population genetic structure of Acacia brevispica from east Africa.

http://www.iusb.edu./~journal/2000/harsh.html, accessed on: 2003/05/16

52

Chapter 2

Hedrick, P.W., Ginevan, M.E., Ewing, E.P., 1976. Genetic polymorphism in

heterogeneous environments. Annu. Rev. Ecol. Syst. 7, 1-32.

Hedrick, P.W., 1986. Genetic polymorphism in heterogeneous environments: a decade

later. Annu. Rev. Ecol. Syst. 17, 535-566.

Hogbin, P.M., Peakall, R., 1999. Evaluation of the contribution of genetic research to the

management of the endangered plant Zieria prostrata. Conserv. Biol. 13, 514-

522.

Jeffrey, C.G., Gavin, F.M., 2000. Genetic tools: The use of biochemical and molecular

markers. In: Young, A., Boshier, D., Boyle, T., (Eds.), Forest conservation

genetics: principles and practice. CSIRO, Collinwood, Australia. pp 39-59.

Joly, H.I., Zeh-Nlo, M., Danthu, P., Aygalent, C., 1992. Population genetics of an

African acacia, Acacia albida. I. Genetic diversity of populations from West

Africa. Aust. J. Bot. 40, 59-73.

Kang, S.S., Chang, M.G., 2000. High levels of allozyme variation within populations and

low allozyme divergence within and among species of Hemerocallis (Liliaceae).

Am. J. Bot. 87, 1634-1646.

Kovach, W. 1999. Multi-variate Statistical Package 3.11b. Kovach Computing Services,

Aberystwyth, Wales, UK.

Macdonald, S.E., Thomas, B.R., Cherniawsky, D.M., Purdy, B.G., 2001. Managing

genetic resources of lodgepole pine in west-central Alberta: patterns of isozyme

variation in natural populations and effects of forest management. For. Ecol.

Manage. 152, 45-58.

53

Chapter 2

Markert, C.L., Faulhaber, I., 1965. Lactate dehydrogenase isozyme patterns of fish. J.

Exp. Zool. 159, 319-332.

Mateu-Andres, I., Segarra-Moragues, J.G., 2002. Patterns of genetic diversity in related

taxa of Antirrhinum L. assessed using allozymes. Biol. J. Linn. Soc. 79, 299-307.

Mitton, J.B., 1983. Conifers. In: Tanksley, S.D., Orton, T.J. (Eds), Isozymes in plant

genetics and breeding. Part B. Elsevier Science Publishers B.V., Amerstam. pp.

443-472.

Moran, G.F., Hopper, S.D., 1987. Conservation of the genetic resources of rare and

widespread eucalypts in remnant vegetation. In: Saunders, D.A., Arnold, G.W.,

Burbidg, A.A., Hopkins, A.J.M. (Eds), Nature conservation: the role of remnants

native vegetation. CSIRO and CALM. Australia. pp. 151-162.

Mueller, U.G., Wolfenbarger, L.L., 1999. AFLP genotyping and fingerprinting. Trends

Ecol. Evol. 14, 389-393.

Nei, M., 1972. Genetic distance between populations. Am. Nat. 106, 283-292.

Oballa, P.O., 1993. Genetic variation within Acacia karroo Hayne. Oxford England,

Unpubl. D. Phil. Oxford University, Oxford, UK.

Oballa, P.O., Olng’otie, P.A.S., 1994. Chromosome numbers in two African Acacia

species. Kew Bulletin 49, 107-113.

Olngótie, P.A.S., 1992. Acacia tortilis (Forsk) Hayne: A study of genetic diversity and

breeding systems. Unpub. D.Phil. Thesis, University of Oxford, Oxford, UK.

Pigliucci, M., Schlichting, C.D., 1997. On the limits of quantitative genetics for the study

of phenotypic evolution. Acta Biotheoret. 45, 143-160.

54

Chapter 2

Quilichini, A., Debussche, M., Thompson, J.D., 2003. Geographic differentiation of

morphological traits and isozymes in the mediterranean island endemic Anchusa

crispa: implications for the conservation of a protected species. Biol. Conserv.

118, 651-660.

Reed, D.H., Frankham, R., 2003. Correlation between fitness and genetic diversity.

Conserv. Biol. 17, 230-237.

Ridgeway, G.J., Sherburne S.W., Lewis R.D., 1970. Polymorphism in the esterase of

Atlantic herring. Trans. Am. Fish. Soc. 99, 147-151.

Ross, J.H., 1971. Acacia karroo in southern Africa. Bothalia 10, 385-401.

Ross, J.H., 1979. A conspectus of the African Acacia species. Mem. Bot. Surv. S. Afr.

44, 93-95.

Schlichting, C.D., 1986. The evolution of phenotypic plasticity in plants. Annu. Rev.

Ecol. Syst. 17, 667-693.

Shrestha, M.K., Golan-Goldhirsh, A., Ward, D., 2002. Population genetic structure in

isolated populations of Acacia raddiana investigated by random amplified

polymorphic DNA (RAPD) markers. Biol. Conserv. 108, 119-127.

Soltis, D.E., Haufler, C.H., Darrow, D.C., Gastony, G.J., 1983. Starch gel electrophoresis

of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining

schedules. Am. Fern. J. 73, 9-26.

Soltis, D.E., Soltis, P.S., 1989. Isozymes in plant biology. Department of Botany,

Washington State University.

55

Chapter 2

Van Straalen, N.M., Timmermans, M.J.T.N., 2002. Genetic variation in toxicant stressed

populations: An evaluation of the "genetic erosion" hypothesis. Hum. Ecol. Risk

Assess. 8, 983-1002.

Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S.M., Schlichting, C.D., Van

Tienderen, P., 1995. Adaptive phenotypic plasticity: consensus and controversy.

Trends Ecol. Evol. 10, 212-217.

Volis, S. Yakubov, B. Shulgina, I., Ward, D., Zur, V., Mendlinger, S., 2001. Tests for

adaptive RAPD variation in population genetic structure of wild barley, Hordeum

spontaneum Koch. Biol. J. Linn. Soc. 74, 289-303.

Volis, S., Mendlinger, S., Ward, D., 2002a. Adaptive traits of wild barley plants of

Mediterranean and desert origin. Oecologia 133, 131-138.

Volis, S., Mendlinger, S., Ward, D., 2002b. Differentiation in populations of Hordeum

spontaneum Koch along a gradient of environmental productivity and

predictability: life history and local adaptation. Biol. J. Linn. Soc. 77, 479-490.

Volis, S., Mendlinger, S., Turuspekov, Y., Esnazarov, U., 2002c. Phenotypic and

allozyme variation in Mediterranean and desert populations of wild barley,

Hordeum spontaneum. Evolution 56, 1403-1415.

Yeh, F.C., Rong-Cai, Y., Boyle, T., 1998. POPGENE version 1.31. Edmonton, Alberta,

Canada, University of Alberta, Center for International Forestry Research.

Yu, J., Mosjidis, J.A., Klingler, K.A., Woods, F.M., 2001. Isozyme diversity in North

American cultivated red clover. Crop Sci. 41, 1625-1628.

56

Chapter 3

Phenotypic plasticity and local adaptation in two extreme populations of

Acacia karroo

Submitted as:

Mboumba, G.B., Ward, D. Phenotypic plasticity and local adaptation in two extreme

populations of Acacia karroo. Journal of Forest Ecology and Management.

57

Chapter 3

Abstract

We investigated differences in plasticity as well as local adaptations of two

extreme (i.e. ecologically and geographically extreme) populations of the extremely

phenotypically-plastic tree species Acacia karroo (Leeu Gamka, arid Karoo and Richards

Bay, subtropical coastal forest) in a common-garden experiment. Seeds were grown at

different levels of water availability with respect to their original environments and

browsing by mammals was simulated in the greenhouse by clipping plants (leaves and

twigs). The results suggest that the populations investigated are phenotypically plastic in

response to these factors. In addition, there were large differences between populations in

their range of plastic responses to treatments. Of the two populations investigated, the

slow-growing one (Leeu Gamka) was phenotypically more plastic with regard to

defence-related traits (longer spines, more tannin) while the fast-growing one (Richards

Bay) was phenotypically more plastic regarding growth-related traits (taller, with longer

leaves). Patterns of performance revealed that the populations have pure strategies of

either growth (forest) and defence (arid). The interactions between populations and

environments in some traits indicated genetic differentiation in plastic responses between

populations and, consequently, that phenotypic plasticity is locally adaptive and not

merely due to environmental differences. The two populations appear to have pure

strategies; when environmental conditions were improved by addition of water, the forest

population increased investment in growth but not defence while the arid populations

increased defence production but not growth.

58

Chapter 3

Keywords: Acacia karroo, phenotypic plasticity, trade-off, local adaptation, G×E

interaction.

59

Chapter 3

Introduction

Adaptive phenotypic plasticity is the ability of a single genotype to produce an

array of phenotypes depending on the environmental extent (Schlichting & Levin, 1984;

Via et al., 1995). Plasticity is expressed continuously in all plants and is an essential part

of the mechanisms by which plants survive, capture resources, deter herbivores and

pathogens and produce offspring (Grime and Mackey, 2002). Many studies suggest

correlations between variation in phenotypic plasticity and genetic differentiation among

individuals within populations (Jasienski et al., 1996) and between populations of

conspecific species (Counts, 1993). However, morphological variation among

populations does not necessarily correlate with genetic variation because some species

lacking genetic variation display high variation in phenotypic plasticity (Taylor and

Aarssen, 1988; Cordell et al., 1998). Understanding the functional or adaptive nature of

phenotypic plasticity necessitates an assessment of variability in plant morphology and

physiology (Bloom et al, 1985; Sultan 1987; Tilman, 1988; Campbell and Grime, 1992).

Indeed, adaptive phenotypic plasticity implies that genotypes differ and that there is local

adaptation “where fitness is maximized in the local environment” (Schlichting and Levin,

1984).

Acacia karroo is widespread throughout Southern Africa and displays a

remarkable phenotypic plasticity over its geographical range (Ross, 1975, Archibald and

Bond, 2003). Previous studies suggest that the species is phenotypically very plastic with

regard to various selection pressures (Barnes et al., 1996; Archibald and Bond, 2003);

and is highly differentiated genetically (Brain, 1985; 1989; Oballa, 1993). It occupies a

wide range of different habitats from subtropical forest to semi-desert (Ross, 1979;

60

Chapter 3

Goldsmith and Carter, 1981, Acocks, 1988), is therefore exposed to temperature and

rainfall extremes (Brain et al., 1996) and is thus likely to be under considerable selective

pressure for local adaptation to particular parts of its geographical range. It is yet to be

established whether the genetic differentiation recorded is associated with local

adaptation and whether the phenotypic variation recorded in adult plants in the field (i.e.

not under controlled conditions - e.g. Archibald and Bond 2003) is related to genotypic

differentiation. We set out to test whether adaptive phenotypic plasticity occurs in this

species and whether phenotypic variation is consistent with local adaptation.

Strategies of growth and defence

African acacias evolved as savanna trees under intense pressure from mammalian

herbivores (Ross, 1979; Scholes and Walker, 1993; Ward and Young, 2002; Shrestha et

al., 2002). In general, plants occupying savannas experience slow growth due to nutrient-

poor soils and thus maintain high levels of secondary metabolites (tannins or alkaloids),

which make them less palatable to browsers than fast-growing plants on nutrient-rich

soils (Bryant et al., 1989). Studies on Acacia responses to herbivory have focused on dry,

tropical savannas, where the importance of browsing ungulates on plants is well

documented (Pellew, 1983; 1984, Gowda, 1997). Additionally, Acacia species are known

to have long thorns and high polyphenol (especially condensed tannins) contents that help

them to deter mammalian herbivores (Wrangham and Waterman, 1981; Cooper and

Owen-Smith, 1985, Ward and Young 2002). It is generally assumed that defence is costly

because investments in defence come at the expense of investments in growth and

reproduction (Briggs and Schultz, 1990; Palo et al., 1993; Ward and Young, 2002). This

61

Chapter 3

is also supported by a number of studies on Acacia drepanolobium, A. seyal (Young,

1987; Milewski et al., 1991), A. tortilis (Gowda, 1996; Rohner and Ward, 1997), and A.

raddiana (Rohner and Ward, 1997).

The aims of this study were twofold: (1) to investigate how plastic two

populations from extreme environments (semi-desert and coastal sub-tropical forest) are,

and (2) to investigate adaptation of the populations to local conditions. Coastal sub-

tropical forest populations grow up to 40 m tall and have small spines, and the semi-

desert populations are short (and are within the browsing height of mammals for most, if

not all, of their lives) and have large thorns (Ross, 1979). Chemical defences of these

populations have not yet been studied. A common-garden experiment was used to

address the question of local adaptation as such experiments have frequently been used to

demonstrate local adaptation for species from geographically-separated populations

(Bradshaw, 1984). We used the widely-supported resource availability hypothesis of

plant defence to develop predictions about the level of investment in defence and growth

in A. karroo populations (Coley et al., 1985). According to this hypothesis, resource

availability is the major factor affecting the amount and type of plant defence. When

plants are resource-limited, plants with inherently low growth rates will be favoured over

those with fast growth rates and that slow growth rates favour heavy investments in

defence. Thus, the optimal levels of defence investment should increase as the potential

growth rate declines (Coley et al., 1985). Accordingly, we predicted the following

differences between the populations:

(1) Individuals from the coastal subtropical forest (Richards Bay) population

encounter little mammalian herbivory and experience high environmental quality with a

62

Chapter 3

mean annual rainfall of 1200 mm. Therefore, they should be able to easily regrow parts

lost to herbivory. These individuals should thus grow fast and invest little in defence.

(2) In contrast, individuals from the semi-desert (Leeu Gamka) population are

browsed regularly by large mammalian herbivores and experience poor environmental

conditions with low rainfall and low soil nutrients, making the cost of regrowth high for

individuals in this population. Therefore, they should grow slowly, and invest heavily in

defence mechanisms based on secondary plant compounds

Materials and methods

Study site and populations

Acacia karroo plants from Leeu Gamka in the semi-desert Karoo in the centre of

South Africa (32.3˚S, 22.3˚E), grow under poor environmental conditions, with a mean

annual rainfall of 150 mm that mostly falls in winter, minimum monthly temperature of

5˚C and a maximum of 32˚C (Archibald and Bond, 2003). At this location, A. karroo

predominantly forms shrubs (with an approximate mean height of 3 m) that are grazed

heavily by livestock (sheep, cattle and goats) and wild herbivores such as kudu

Tragelaphus strepsiceros. The second population, from Richards Bay on the east coast of

South Africa (28.46˚S, 32.06˚E), grows in a subtropical coastal forest, a moist

environment with a mean annual rainfall of 1200 mm, with rain falling throughout the

year. Monthly, the minimum and maximum temperatures are respectively 10˚C and 28˚C

(Archibald and Bond, 2003). Individuals in this population grow into trees reaching

63

Chapter 3

heights of over 30 m, and face low or nonexistent browsing when adults (Brain et al.,

1996; Boyes, 2004).

Experimental treatment

Seeds were collected from 15 trees (randomly chosen) in each population in July

2003. Trees sampled were at least 10 m apart from each other to ensure that each tree

represented a unique genotype and that the individuals were not closely related to each

other. Seeds were placed in paper bags and brought to the laboratory one the same day.

Seeds were then dried at 25˚C to remove all unwanted organisms such as fungi or

bacteria, which may cause deterioration in seed quality. After drying, seeds were kept in

the fridge at 2- 8˚C. From each population, 30 seeds (from each mother tree) were tested

for viability and were pre-treated in boiling water for two hours and left for a day in tap

water to scarify the hard seed coat. In nature, the seed coat is broken by a number of

biotic and abiotic agents, such as mammals, microorganisms, soil acids, temperature

fluctuations, rain and winds (Barnes et al., 1996). After the pre-treatments described

above, seeds were planted into trays containing vermiculite and left in the nursery to

germinate. Watering was done once a day, until seedlings were large enough to undergo

the experimental treatments. In each, 12 seedlings from each of the ten selected from the

15 trees sampled per population were transplanted in plastic bags to undergo the

treatments (in total 240 seedlings). To simulate the major ecological differences between

the two populations investigated, water availability and browsing were manipulated. A

completely crossed experimental design was used, and seedlings were randomly allocated

to the four treatments. Of the 12 offspring per tree sampled, six (three clipped and three

64

Chapter 3

unclipped) were watered every three days (high water availability) and the remainder

every 10 days (low water availability). Clipping of the plants was used to imitate

browsing by mammals. The use of real herbivores is often laborious and may be virtually

impossible in many experimental setups (Lehtilä and Boalt, 2004). Plants were clipped 1

cm from the tip of leaves and young branches. Water soluble fertilizer “Vitagro” (N-P-K-

S-Zn-B: 254-30-89-24-0.445-0.405 g/kg) was used as a soil nutrient to maximize growth

rate. It was provided to all bags at a 1% concentration (0.5kg to 50l of water) for the first

4 weeks, 1.5% concentration for the 2 next weeks (until 6 weeks stage) and then at 2%

concentration for the remainder of the time.

The experiment was terminated after a year when above- and below-ground

biomasses were harvested. Soil was removed from the roots by washing before below-

ground biomass was determined, and plants were dried in oven at 70˚C for 24 h. The

following parameters were measured before harvesting the plants: plant height (vertically

from ground to highest leaf (cm)), stem diameter (greatest diameter of the main stem

above the ground (cm)), leaf length (mean of five leaves on a randomly-chosen branch),

spine number (total for each tree) and spine length (cm) (mean length of all spines on a

randomly-selected branch). All these parameters were measured according to Archibald

and Bond (2003).

Condensed tannins were determined by the proanthocyanidin method (Hagerman,

1995). We ground 0.02 g of dried plant material with a Wiley mill to pass through a 0.6

mm sieve and the absorbance was read at 550 nm with a spectrophotometer. The

concentration of condensed tannins as quebracho equivalents was calculated using this

65

Chapter 3

formula: = [{(0.0049*abs) – 0.0004}/ mass] * 100 mg/ml. Quebracho for the standard

was obtained from A. Hagerman (Miami University, Oxford, Ohio, USA).

Data analysis

StatSoft, Inc. (2004) software was used for all statistical analyses. To examine the

overall significance of the treatments, MANOVA was used to protect experiment-wise

error when examining many traits. Subsequently, a full factorial ANOVA (General

Linear Model), with mother tree (nested within populations) as a random effect, was

conducted on each dependent variable. Some variables were transformed to satisfy

assumptions of the analysis. Thus, plant height, leaf length, spine number, and above and

below ground biomass were log10-transformed and tannin concentrations were arcsine-

square root transformed. Stem diameter and spine length required no transformation. The

level of significance was 0.05. Correlations between various traits were obtained using a

Pearson’s product-moment correlation.

Results

For the sake of brevity, we describe below only those results where a significant

effect was detected.

Phenotypic plasticity

The overall multivariate analysis of variance (MANOVA) indicated significant

phenotypic plasticity for both growth-related and defence-related traits (Table 1). The

66

Chapter 3

traits that were the most plastic (i.e. showed greatest differences among treatment means)

were above- and below-ground biomass, and spine length. The least plastic traits were

stem diameter and spine number.

Table 1: Results of MANOVA on all quantitative traits measured.

Factor d.f. Wilks’ λ F P

Clipping 8 .894 1.95 .058

Populations 8 .431 21.64 <0.001

Watering 8 .823 3.52 .001

Clipping*Populations 8 .857 2.73 .008

Clipping*Watering 8 .962 .66 .730

Populations*Watering 8 .807 3.93 <.001

Clipping*Populations*Watering 8 .922 1.39 .208

Mother(Populations) 144 .195 1.69 <.001

Error 131

A significant population effect was observed for plant height (F = 38.54, p =

0.001), stem diameter (F = 8.11, p < 0.001), leaf length (F = 115.58, p < 0.001), spine

length (F = 41.61, P < 0.001), below-ground biomass (F = 76.36, p < 0.001) and tannin

concentration (F = 20.23, p = 0.001) (Table 2 (a and b)).

67

Chapter 3

Table 2 (a): Mean squares, F values and significance (p) in the analyses of variance of the two populations of A. karroo under different treatments. Tannin concentration Leaf length Factor d.f. F Mean Squares (p) d.f. F Mean squares (p)

Clipping 1 0.71 58 (0.4) 1 0.03 0.001 (0.870)

Population 1 13.81 1630 (0.001) 1 115.58 4.406 (0.001)

Water 1 1.28 103 (0.261) 1 1.44 0.033 (0.249)

Clipping*Pop 1 4.29 346 (0.040) 1 0.55 0.014 (0.459)

Clipping*Water 1 0.93 75 (0.34) 1 0.07 0.002 (0.796)

Pop*Water 1 0.7 56 (0.41) 1 0.56 0.014 (0.457)

Clip*Pop*Water 1 0.34 28 (0.560) 1 0 0.00001 (0.986)

Mother(Pop) 18 1.5 121 (0.098) 18 1.57 0.039 (0.076)

Error 138 80.58 156 0.025 Stem diameter Spine length Factor d.f. F Mean squares (p) d.f. F Mean squares (p)

Clipping 1 2.15 0.023 (0.145) 1 1.12 0.094 (0.291)

Population 1 8.11 0.220 (0.001) 1 41.61 8.688 (<0.001)

Water 1 5.92 0.064 (0.016) 1 3.06 0.256 (0.082)

Clipping*Pop 1 1.67 0.018 (0.198) 1 0.01 0.001 (0.908)

Clipping*Water 1 1.77 0.019 (0.185) 1 0.32 0.027 (0.571)

Pop*Water 1 1.56 0.017 (0.213) 1 7.87 0.658 (0.006)

Clip*Pop*Water 1 4.38 0.047 (0.038) 1 0.15 0.013 (0.698)

Mother(Pop) 18 2.62 0.028 (<0.001) 18 2.61 0.218 (<0.001)

Error 156 0.011 156 0.084

68

Chapter 3

Table 2 (b): Mean squares, F values and significance (p) in the analyses of variance of the two populations of A. karroo under different treatments.

Spine number Belowground biomass Factor d.f. F Mean squares (p) d.f. F Mean squares (p)

Clipping 1 0.14 0.003 (0.709) 1 0 0 (0.992)

Population 1 0.36 0.007 (0.617) 1 76.36 10.88 (<0.001)

Water 1 0.02 0.001 (0.881) 1 29.84 2.14 (<0.001)

Clipping*Pop 1 3.27 0.060 (0.073) 1 0.25 0.02 (0.618)

Clipping*Water 1 1.12 0.021 (0.292) 1 0.25 0.02 (0.62)

Pop*Water 1 6.09 0.113 (0.015) 1 0.87 0.06 (0.352)

Clip*Pop*Water 1 5.01 0.093 (0.027) 1 0.56 0.04 (0.475)

Mother(Pop) 18 1.44 0.027 (0.121) 18 2.06 0.15 (0.01)

Error 156 0.019 156 0.072 Aboveground biomass Plant height Factor d.f. F Mean squares (p) d.f. F Mean squares (p)

Clipping 1 0.04 0.002 (0.85) 1 8.25 0.131 (0.005)

Population 1 4.2 0.331 (0.054) 1 38.54 0.612 (0.001)

Water 1 21.17 1.087 (<0.001) 1 2.88 0.046 (0.092)

Clipping*Pop 1 0.08 0.004 (0.782) 1 1.54 0.024 (0.217)

Clipping*Water 1 1.64 0.084 (0.202) 1 3.85 0.061 (0.051)

Pop*Water 1 4.94 0.254 (0.028) 1 3.07 0.049 (0.082)

Clip*Pop*Water 1 0.61 0.031 (0.435) 1 1.81 0.029 (0.180)

Mother(Pop) 18 1.57 0.081 (0.073) 18 2.73 0.043 (<0.001)

Error 156 0.051 156 0.016

69

Chapter 3

Richards Bay plants were taller and had greater leaf length than Leeu Gamka

plants. Leeu Gamka plants had greater stem diameter, spine length, below-ground

biomass, and tannin concentration (Table 3).

Table 3: Phenotypic differences in various morphological traits between the two populations. Values are means (± 1 SD).

Populations Leeu Gamka Richards Bay p

Plant height (log10 ,cm) 1.311 (0.164) 1.443 (0.123) 0.001

Stem diameter (cm) 0.456 (0.125) 0.389 (0.108) 0.001

Spine length (cm) 1.259 (0.358) 0.801 (0.282) < 0.001

Spine number (log10) 1.739 (0.131) 1.761 (0.151) 0.617

Leaf length (log10 ,cm) 0.344 (0.206) 0.670 (0.114) 0.001

Aboveground biomass (log10, g) 0.23 (0.228) 0.141 (0.263) 0.054

Belowground biomass (log10, g) 0.232 (0.322) -0.28 (0.284) < 0.001

Condensed Tannin Concentration ( % Q.E.) 15.92 (10.92) 8.72 (7.36) 0.0014

A significant clipping effect was observed for plant height only (F = 8.25, p =

0.005). Clipped plants had lower mean height (22.39 cm) than unclipped ones (26.30

cm). There was a significant watering effect on stem diameter (F = 5.92, p = 0.016),

above-ground biomass (F = 21.17, p < 0.001) and belowground biomass (F= 29.84, p <

0.001) (Table 2 a and b) only. Plants had greater stem diameter with an increase in water

availability (high watering: 0.44 ± 0.013 cm vs. low watering: 0.40 ± 0.012 cm). Higher

70

Chapter 3

above- and below-ground biomass means were recorded when the level of water

increased (respectively high watering: 1.66 g ± 1.05 vs. low watering: 1.38 g ±1.07 and

1.12 g ±1.09 vs. 0.70 g ±1.09).

Coefficients of variation of growth- and defence-related traits within populations

were measured in the four treatments. The trend was for higher variation in Richards Bay

plants for spine length, spine number, tannin, stem diameter, above- and below-ground

biomass while Leeu Gamka plants exhibited higher variation in plant height and leaf

length (Table 4).

Table 4: Coefficients of variation (CV %) for growth- and defence-related traits

Plant height

Stem diameter

Leaf length

Spine length

Spine number

Above biomass

Below biomass

Tannin

Herbivory

RB 10 10.04 26.30 16.75 28.75 8.16 58.60

76.44

80.23

Herbivory

LG 10 14.83 29.95 83.01 28.37 6.29 43.04

54.87

61.31

Control

RB 10 7.69 32.91 14.90 39.77 7.87 53.51

41.83

74.99

Control

LG 10 10.73 26.49 50.06 22.74 6.42 47.82

52.63

79.59

Herbivory

RB 3 8.72 24.17 18.73 34.57 10.59 40.97

56.91

71.24

Herbivory

LG 3 11.02 24.43 35.65 32.37 7.91 44.85

59.39

62.53

Control

RB 3 5.16 25.12 16.82 36.26 7.21 53.79

86.88

99.47

Control

LG 3 10.66 23.19 49.82 22.43 7.67 48.49

64.12

61.76

71

Chapter 3

Trade-offs between growth and defence

The resource availability hypothesis predicts that there will be trade-offs in

investment between growth and defence-related traits (Coley et al., 1985). Consequently,

correlations between various traits were tested. There was a significantly negative

correlation between spine length and leaf length (r = -0.27, F = 13.64, p < 0.001).

However, spine length was strongly and positively correlated with other traits, viz. stem

diameter (r = 0.46, F = 47.70, p < 0.001), above-ground biomass (r = 0.44, F = 43.10, p <

0.001) and below-ground biomass (r = 0.59, F = 93.88, p < 0.001). No correlation was

found between spine length and plant height (r = 0.1, F = 1.74, p = 0.19). There was a

positive correlation between spine length and other defence-related traits, viz. tannin

concentration (r = 0.21, F = 7.51, p = 0.007) and spine number (r = 0.24, F = 10.54, p =

0.001).

There was a significant positive correlation between spine number and stem

diameter (r = 0.21, F = 8.2, p = 0.005). There was a significant positive correlation

between spine number and leaf length (r = 0.17, F = 5.2, p = 0.024). There was a

significant positive correlation between spine number and above-ground biomass (r =

0.54, F = 74.8, p < 0.001). There was a significant positive correlation between spine

number and below ground biomass (r = 0.24, F = 10.9, p = 0.001), and between spine

number and plant height (r = 0.44, F = 44.07, p < 0.001).

There was no significant correlation between tannin concentration and plant

height (r = 0.1, F = 1.48, p < 0.226) as well as between tannin concentration and stem

diameter (r = 0.05, F = 0.5, p = 0.46). However, there was a significant negative

correlation between leaf length and tannin concentration (r = -0.23, F = 9.00, p = 0.003).

72

Chapter 3

Tannin concentration was weakly positively correlated with above-ground biomass (r =

0.16, F = 4.3, p = 0.04). There was a significant positive correlation between tannin

concentration and below-ground biomass (r = 0.32, F = 18.44, p < 0.001). There was no

correlation between tannin concentration and spine number (r = 0.02, F = 0.04, p = 0.84).

Genotypic by environment interactions among populations

Reaction norms of populations for a number of traits cross. These results suggest

that there are significant genotype-by-environment interactions for these traits (Stearns,

1992; Schlichting and Pigliucci, 1998). There was a significant population*water

interaction effect on spine number (F = 6.10, p = 0.015), spine length (F = 7.87, p =

0.006) and above-ground biomass (F = 4.94, p < 0.028) (Table 2). The response with

regard to spine number for both populations was not significant (Figure 1).

High Low

Watering

1.66

1.72

1.78

1.84

log(

Spi

ne n

umbe

r)

Leeu Gamka Richards Bay

Figure 1: Interaction between populations and water levels for spine number.

73

Chapter 3

A significant increase in spine length (t = - 2.92, df = 35, p = 0.006) due to an

increase in water availability was only recorded in Leeu Gamka plants (Figure 2).

High Low

Watering

0.6

0.9

1.2

1.5

Spin

e le

ngth

(cm

)

Leeu Gamka Richards Bay

Figure 2: Interaction between populations and water levels for spine length.

Both populations increased their aboveground biomass as the water availability

increased. However, Richards Bay plants responded significantly whereas Leeu Gamka

plants did not (t = 4.69, df = 47, p < 0.001) (Figure 3).

74

Chapter 3

High LowWatering

-0.1

0.0

0.1

0.2

0.3

0.4

Log

(Abo

vebi

omas

s, g

) Richards Bay Leeu Gamka

Figure 3: Interaction between populations and water levels for aboveground biomass.

There was a population* clipping interaction effect on tannin concentration (F =

4.29, p = 0.04). Leeu Gamka had higher tannin concentrations in the control than when

clipped whereas Richards Bay had slightly higher tannin concentrations when clipped

than in the control (Figure 4). However, there was no significant difference between

treatments in both populations (this can be interpreted as the range of plasticity being

narrow) as indicated by a t test (Richards Bay population: t = 1.12, df = 35, p = 0.27;

Leeu Gamka: t = 1.46, df = 34, p = 0.15).

75

Chapter 3

Clipping Control

Treatments

4

10

16

22

Asin

(Tan

nin,

%Q

E)

Leeu Gamka Richards Bay

Figure 4: Interaction between populations and clipping for tannin concentration.

There was no clipping*water interaction effect on either morphological traits or

tannin concentration. However, there was a significant population*clipping*water

interaction effect on stem diameter (F = 4.38, p = 0.038) (Figure 5), and spine number (F

= 5.01, p = 0.027) (Figure 6). However, the interaction was stronger with a decrease in

water availability in both traits (Figure 5 and Figure 6). Leeu Gamka plants increased

stem diameter when clipped when water availability was high but decreased stem

diameter when clipped under low water availability. In contrast, Richards Bay plants

increased stem diameter with clipping under low and high water availability (Figure 5).

There were no significant differences between treatments in both populations (p > 0.05),

thus reflecting the narrow range of plasticity in populations for this trait. Neither

population increased spine number when clipped under conditions of high water

availability. However, under low water availability, the Richards Bay population

increased its investment in spine number while the Leeu Gamka population decreased

76

Chapter 3

spine number, although only the change in the Leeu Gamka population was significant (t

= -3.14, p = 0.005).

High watering

Clipping Control0.25

0.40

0.55

Stem

dia

met

er, c

m

Low watering

Clipping Control

Leeu Gamka Richards Bay

Figure 5: Interaction between populations, clipping and water levels for stem diameter.

High watering

Clipping Control1.55

1.65

1.75

1.85

log(

Spin

e nu

mbe

r)

Low watering

Clipping Control

Leeu Gamka Richards Bay

Figure 6: Interaction between populations, clipping and water levels for spine number.

77

Chapter 3

Discussion

The overall results suggest that the populations investigated are phenotypically

plastic. Individuals of the two populations differed greatly in their range of plastic

responses to treatments. Richards Bay plants invested more in plant height and leaf

length whereas Leeu Gamka plants invested more in stem diameter, spine length, below-

ground biomass and tannin concentration. These results are consistent with our

predictions. Archibald and Bond (2003) found similar results (for aboveground

parameters, excluding tannin which they did not measure) when working with wild

populations of A. karroo from arid shrubland and forest. Arid trees invested more in

vertical growth (i.e. plant height), had longer spines and greater stem diameter than forest

trees. It is generally assumed that plants adapted to resource-rich environments should

invest more in growth and plants adapted to resource-poor environments should invest

more in defence (Coley et al., 1985). The relationships found among these traits in A.

karroo are consistent with these strategies of investments. There were trade-offs between

leaf length and spine length, as well as between leaf length and tannin concentration.

However, most other traits were positively correlated, contra the resource availability

hypothesis. Water availability had an effect on two traits, viz. stem diameter and above-

ground biomass. Both traits increased as the water level increased. Contrary to the

findings of Shrestha et al. (2002) that an increase in level of water increased tannin

production in Acacia raddiana, there was no significant water effect on tannin

concentration. In our study, there was a significant clipping effect on plant height;

herbivory significantly decreased plant height.

78

Chapter 3

A significant population*water interaction indicated that an increase in water

availability was related to an increase of performance or investment, indicating that

populations have pure strategies. Richards Bay plants increased aboveground biomass but

not spine length whereas Leeu Gamka plants increased spine length but not aboveground

biomass in response to an increase in environmental condition (as indexed by water

availability); as a result, forest plants have a growth strategy and desert a defence

strategy.

There was a significant interaction between clipping and populations for tannin

concentration. Clipping increased tannin concentration in Richards Bay plants compared

to the control treatment, while it resulted in a decrease in tannin concentration in Leeu

Gamka plants with regard to the control treatment. Note that, even though there was a

decrease in tannin concentration in Leeu Gamka plants, the overall tannin concentration

was still higher than in Richards Bay plants (Figure 4).

The interaction between population*clipping*water for stem diameter as well as

for spine number was more pronounced with a decrease in water availability than for

stem diameter. In both populations, there was an increase in investment in spine number

after being exposed to clipping. Unlike Leeu Gamka plants, Richards Bay plants

increased their stem diameter and spine number after being clipped in the low water

treatment. This suggests that the regrowth of lost parts in Leeu Gamka plants is more

costly than in Richards Bay plants. Evidence for such a response has been demonstrated

in some woody species (e.g. Osteospermum sinuatum, Van der Heyden, 1992; Acacia

tortilis, du Toit et al., 1990) that responded to simulated browsing by rapid regrowth,

while other species had limited regrowth (e.g. Pteronia pallens, Stock et al., 1993; Acacia

79

Chapter 3

nigrescens, du Toit et al., 1990). This is also in accordance with Coley et al. (1985)’s

findings on resources availability of plants concerning fast-growing and slow-growing

plants. Coley and her colleagues suggested that slow-growing plants occur in poor

environments. Thus; the loss of leaves and twigs is associated with a loss of energy and

their replacement is costly. To remedy this situation, these plants tend to reduce the

turnover rate of such parts because it is advantageous in terms of energy wasted. In

contrast, fast-growing plants occurring in rich environments tend to maximize the rapid

turnover of leaves because younger leaves allow higher photosynthetic rates.

The significant interactions between populations and water, and between

populations, clipping and water indicate that there were genotype by environment

interactions. Consequently, local adaptation is likely for these traits. Leeu Gamka plants

showed local adaptation with regard to spine number, spine length and above-ground

biomass with an increase in water availability. They performed better in a richer

environment. Furthermore, performance in both environments for spine number and

above-ground biomass were not significantly different although performance differed

among environments for spine length. Our results agree with the results of Milton (1991),

who found that moisture might mediate selection for spinescence in moist, nutrient-rich

habitats in arid areas because mammalian herbivores concentrate there and plants are

consequently heavily broken, trampled and browsed. Richards Bay plants showed local

adaptation of the same traits. However, they performed well in the rich environment

(except for spine length and spine number where there were no significant differences).

Both populations showed local adaptation in terms of tannin concentration under the

clipping treatment. Leeu Gamka plants produced less tannin after being browsed than

80

Chapter 3

when browsing did not occur, while Richards Bay plants produced relatively more tannin

after browsing by herbivores. Finally, local adaptation was observed for stem diameter

and spine number in a three way interaction (populations*clipping *water). Leeu Gamka

plants performed better in a poor environment in terms of spine number with the absence

of herbivores than when they were present. While Richards Bay plants performed better

with the presence of herbivores than when they were absent. However, performances for

both populations were similar in a rich environment regardless of the treatment. Similar

performances were also observed with regard to stem diameter.

Of the two populations investigated, the slow-growing one (Leeu Gamka) was

phenotypically more plastic with regard to defence-related traits (longer spines, more

tannin) while the fast-growing one (Richards Bay) was phenotypically more plastic

regarding growth-related traits (taller, longer leaves). Furthermore Leeu Gamka plants

did not exhibit a large amount of variability in their phenotypes (see coefficient of

variation, CV %) as would be expected from plant population such as this from a highly

variable environment (e.g. Hedrick et al., 1976; Hedrick, 1986). However, this

observation is in accordance with the results of a previous study on molecular genetic

variation (See Chapter 2), which showed a lower genetic diversity in Leeu Gamka

population than in the Richards Bay population.

Acknowledgements

We would like to thank Dr Theron, Khanysile Mbatha, Donald Midocko Iponga

for technical assistance and all those who provided us with valuable suggestions and

81

Chapter 3

comments. This study was funded by National Research Foundation (NRF) South Africa

and the Gabonese Government.

References

Acocks, J.P.H., 1988. Veld types of South Africa. 3rd Edition. Mem. Bot. Surv. S. Afr.

57 Botanical Research Institute, South Africa. pp. 146.

Archibald, S. Bond, W. J., 2003. Growing tall vs. growing wide: tree architecture and

allometry of A. karroo in forest, savanna, and arid environments. Oikos 102, 3-14.

Barnes, R.D., Filer, D.L., Milton S.J., 1996. Acacia karroo: monograph and annotated

bibliography. Trop. For. Pap. 32. Oxford Forestry Institute, UK.

Bloom, A.J., Chapin, F.S. III, Mooney, H.A., 1985. Resource limitation in plants- an

economic analogy. Annu. Rev. Ecol. Syst. 16, 363-392.

Boyes, L.J., 2004. Regeneration failure and the Acacia karroo successional pathway in

coastal dune forest in KwaZulu-Natal. Unpubl. BSc (Hons) thesis, University of

KwaZulu-Natal, Pietermaritzburg.

Bradshaw, A.D., 1984. Ecological significance of genetic variation between populations.

In: Dirzo, R. and Sarukhan, J. (Eds), Perspectives in Plant Population Ecology.

Sinauer Associates, Sunderland, MA. pp. 213-228.

Brain, P., 1985. Leaf peroxidase types in Acacia karroo. Geographical distribution and

influence of the environment. S. Afr. J. Bot. 52, 47-52.

Brain, P., 1989. Genetic races in a ring species of Acacia karroo. S. Afr. Sci. 85, 181-

185.

82

Chapter 3

Brain, P., Harris, S.A., Barnes, R.D., 1996. Leaf Peroxidase Types in Acacia karroo

Hayne (Acacieae, Leguminosae): a range-wide study. Silvae Genet. 46, 88-94.

Briggs, M.A., Schultz, J.C., 1990. Chemical defence production in Lotus corniculatus. II.

Trade-offs among growth, reproduction, and defence. Oecologia 83, 32-37.

Bryant, J.P., Kuropat, P., Cooper, S.M., Frisby, K., Owen-Smith, N., 1989. Resource

availability hypothesis of plant antiherbivore defence tested in a South Africa

savanna ecosystem. Nature 340, 227-229.

Campbell, B.D., Grime, J.P., 1992. An experimental test of plant strategy theory. Ecology

73, 15-29.

Coley, P.D., Bryant, J.P. and Chapin, F.S. III, 1985. Resource availability hypothesis and

plant antiherbivore defence. Science 230, 895-899.

Cooper, S.M., Owen-Smith, N., 1985. Condensed tannins deter feeding by browsing

ungulates in a South African savanna. Oecologia 67, 142-146.

Cordell, S., Goldstein, G., Mueller-Dombois, D., Webb, D., Vitousek, P.M., 1998.

Physiological and morphological variation in Metrosideros polymorpha, a

dominant Hawaiian tree species, along an altitudinal gradient: the role of

phenotypic plasticity. Oecologia 113, 188-196.

Counts, R.L., 1993. Phenotypic plasticity and genetic variability in Zizania spp. along a

latitudinal gradient. Can. J. Bot. 71, 145-154.

Du Toit, J.T., Bryant, J.P., Frisby, K., 1990. Regrowth and palatability of Acacia shoots

following pruning by afican savanna browsers. Ecology 71, 149-154.

Goldsmith, B., Carter, D.T., 1981. The indigenous timbers of Zimbabwe. Forest

Commission, Zimbabwe. Res. Bull. 9, 9.

83

Chapter 3

Gowda, J.H., 1996. Plant defences: Influence on plant –animal interactions in eastern

Africa. PhD thesis. Swedish University of Agricultural Sciences, Umea.

Gowda, J.H., 1997. Physical and chemical response of juvenile Acacia tortilis trees to

browsing. Experimental evidence. Funct. Ecol. 11, 106-111.

Grime, J.P, Mackey, J.M.L., 2002. Role of plasticity in resource capture by plants. Evol.

Ecol. 16, 299-307.

Hagerman, A.E., 1995. Tannin analysis. Unpublished laboratory manual, Miami

University, Oxford, Ohio. pp 66.

Hedrick, P.W., Ginevan, M.E., Ewing, E.P., 1976. Genetic polymorphism in

heterogeneous environments. Annu. Rev. Ecol. Syst. 7, 1-32.

Hedrick, P.W., 1986. Genetic polymorphism in heterogeneous environments: a decade

later. Annu. Rev. Ecol. Syst. 17, 535-566.

Jasienski, M., Francisco, J.A., Bazzaz, F.A., 1996. Phenotypic plasticity and similarity of

DNA among genotypes of an annual plant. Heredity 78, 176-181.

Lehtilä, K., Boalt, E., 2004. The use and usefulness of artificial herbivory in plant-

herbivore studies. Ecol. Stud. 173, 257-275

Milewski, A.V., Young, T.P., Madden, D., 1991. Thorns as induced defences:

experimental evidence. Oecologia 86, 70-75.

Milton, S.J., 1991. Plant spinescence in arid southern Africa: does moisture mediate

selection by mammals? Oecologia 87, 279-287.

Oballa, P.O., 1993. Genetic variation within Acacia karroo Hayne. Unpubl. D. Phil.

Oxford University. Oxford England.

84

Chapter 3

Palo, R.T., Gowda, J., Hogberg, P., 1993. Species height and root symbiosis, two factors

influencing antiherbivore defence of woody plants in East African savanna.

Oecologia 93, 322-326.

Pellew, R.A., 1983. The impacts of elephant, giraffe and fire upon the Acacia tortilis

woodlands of the Serengeti. Afr. J. Ecol. 21, 41-74.

Pellew, R.A., 1984. The feeding ecology of a selective browser, the giraffe (Giraffa

camelopardalis tippelskirchi). J. Zool. London 202, 57-81.

Rohner, C., Ward, D., 1997. Chemical and mechanical defence against herbivory in two

sympatric species of desert Acacia. J. Veg. Sci. 8, 717-726.

Ross, J.H., 1975. Notes on African Acacia species. Bothalia 11, 443-447.

Ross, J.H., 1979. A conspectus of the African Acacia species. Mem. Bot. Surv. S. Afr.

44, 93-95.

Schlichting, C.D., Levin, D.A., 1984. Phenotypic plasticity of annual Phlox: tests of some

hypotheses. Am. J. Bot. 71, 252-260.

Schlichting, C.D., Pigliucci, M., 1998. Phenotypic evolution, a reaction morm

perspective. Sinauer, Sunderland, MA.

Scholes, R.J., Walker, B.H., 1993. An African Savanna: Synthesis of the Nylsvley Study.

Cambridge University Press, Cambridge, UK.

Shrestha, M.K., Golan-Goldhirsh, A., Ward, D., 2002. Population genetic structure in

isolated populations of Acacia raddiana investigated by random amplified

polymorphic DNA (RAPD) markers. Biol. Conserv. 108, 119-127.

StatSoft, Inc., 2004. STATISTICA (data analysis software system), version 7.

www.statsoft.com.

85

Chapter 3

Stearns, S.C., 1992. The evolution of life histories. Cambridge University Press,

Cambridge, UK.

Stock, W.D., Le Roux, D., Van der Heyden, F., 1993. Regrowth and tannin production in

woody and succulent Karoo shrubs in response to simulated browsing. Oecologia

96, 562-568.

Sultan, S.E., 1987. Evolutionary implications of phenotypic plasticity in plants. Evol.

Biol. 21, 127-178.

Taylor, D.R., Aarssen, L.W., 1988. An interpretation of phenotypic plasticity in

Agropyron repens (Graminae). Am. J. Bot. 75, 401-413.

Tilman, D., 1988. Plant strategies and the dynamics and structure of plant communities.

Princeton University Press, Princeton, NJ.

Van der Heyden, F., 1992. Effects of defoliation on regrowth and carbon budgets three

semi-arid karoo shrubs. Ph.D. thesis, University of Cape Town.

Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S.M., Schlichting, C.D., Van

Tienderen, P., 1995. Adaptive phenotypic plasticity: consensus and controversy.

Trends Ecol. Evol. 10, 212-217.

Ward, D., Young, T.P., 2002. Effects of large mammalian herbivores and ant symbionts

on condensed tannins of Acacia drepanolobium in Kenya. J. Chem. Ecol. 28, 921-

937.

Wrangham, R.W., Waterman, P.G., 1981. Feeding behaviour of vervet monkeys on

Acacia tortilis and Acacia xanthophloea: With special reference to reproductive

strategies and tannin production. J. Anim. Ecol. 50, 715-731.

86

Chapter 3

Young, T.P., 1987. An increased thorn length in Acacia drepanolobium- an induced

response to browsing. Oecologia 71, 434-438.

87

Chapter 4

Conclusions

Previous studies investigating phenotypic plasticity and population differentiation

in Acacia karroo have approached the topic from a variety of different angles (e.g. Brain

1986, 1989, Archibald and Bond 2003). We acknowledge that these approaches will

endow the literature with some substantiation of adaptive phenotypic plasticity in local

populations, although it remains difficult to gauge the precise effect of each factor, owing

to a large number of interacting variables that typify each environment. Our perceptions

of the nature of adaptive phenotypic plasticity in populations experiencing a variety of

selection pressures dictates that due consideration should be given to the age of the

individuals in addition to the environment under consideration. That is, inferences from

field observations are limited in their ability to discern the adaptive nature of phenotypic

plasticity (particularly in long-lived trees - cf. Archibald and Bond 2003) and require

controlled studies such as we have done (i.e. growing seeds of different provenance under

identical conditions or by reciprocal transplanting). The current study is also innovative

because it remedies the shortcomings of earlier studies of genetic differentiation (Brain,

1986; 1989) that merely considered a single protein. Additionally, it is the first attempt to

combine studies of phenotypic differentiation and genetic variation. Thus, this study

bridges the gaps in research on population differentiation based on phenotypic and

genetic variation by not only assessing differentiation at a large number of allozyme loci

but also by combining molecular genetic research with a common-garden experiment to

provide a better understanding of the phenomenon of adaptive phenotypic plasticity

88

Chapter 4

under controlled conditions. We do not claim to have covered all possible loopholes in

the study of phenotypic plasticity, but have certainly opened the door towards a fuller

understanding of this fascinating natural laboratory of the evolution of phenotypic

plasticity.

Our conclusions can be summarized according to our two main questions:

1) Do both populations differ genetically?

The analysis of the genetic variation within and between populations, at the

allozyme level (proteins), indicated that the two study populations are genetically highly

differentiated. The high genetic diversity uncovered falls in the range of populations on

the verge of moving towards species or variant level (e.g. Gottlieb, 1977). This is an

outstanding result in that it supports Swartz’s (1982) findings based on 19 morphological

characters, affirming that the Richards Bay population is a variety of A. karroo, although

our results do not show the complete differentiation of these extreme populations that

might be sufficient to raise the populations to species status (i.e. contra Swartz in Coates

Palgrave, 2002). We concluded that the degree of differentiation between these

populations is for the most part due to their geographic isolation as well as adaptation to

their local environments.

An argument has been put forward that maintains that allozyme markers detect

only 30% of genetic variation (Van Straalen and Timmermans, 2002). Our study used a

number of loci within the range usually considered to be sufficient to show significant

differentiation should it exist (12-20 loci; reviewed by Hamrick and Godt, 1989).

Consequently, we are confident that our results reliably represent significant

89

Chapter 4

differentiation of these populations. Future studies using more powerful techniques, and

with more populations, could determine more precisely the degree of population

differentiation and could examine the specific status of these populations. Nonetheless,

our results provide considerable evidence that the populations of A. karroo represent a

fascinating opportunity to examine evolution by natural selection in progress.

2) Do populations differ in phenotypic plasticity in response to experimental

manipulation?

The controlled experiments in the greenhouse demonstrated that the differences in

strategies of investments observed in the wild between the two populations did not

change under altered conditions. This implies that these investment strategies are

genetically based. The populations we studied have pure strategies of growth and

defence. That is, when environmental quality was improved (i.e. when water was added),

Richards Bay plants (mesic subtropical forest environment) increased investments in

growth (as indexed by aboveground biomass) and not defence (spine length) while Leeu

Gamka plants (arid environment) did the reverse i.e. they increased investments in

defence (spine length) and not growth (aboveground biomass). This reflects the broad-

spectrum trend of investment of forest and desert plants (in growth and defence,

respectively) as indicated by the resource availability hypothesis (Coley et al., 1985),

although that hypothesis specifically predicts a trade-off between growth and defence

(which we did not find). We also found significant interactions between populations and

environments in some traits, which indicate genetic differentiation in plastic responses

90

Chapter 4

between populations and therefore that phenotypic plasticity in the two A. karroo

populations is locally adaptive and not merely due to environmental differences.

Acacia karroo is not a species threatened by extinction and is unlikely to

disappear, unlike numerous other plant species that are overexploited. A causative factor

that frequently leads to the extermination of species is the loss of or very low genetic

diversity, which lessens the propensity to evolve and respond to environmental change

(e.g. Pigliucci and Schlichting, 1997). Our results show relatively high levels of genetic

diversity in A. karroo, leading us to conclude that there is no urgent need for the

conservation and genetic preservation of these populations. On the other hand, if there

was a need to set up a management plan, the results discussed here indicate that each

population should be conserved separately because they are genetically differentiated and

show evidence of significant genotype-by-environment interactions (i.e. are locally

adapted). Mixing populations would result in the loss of genotypes that are well adapted

to specific environments.

91

Chapter 4

References

Archibald, S. Bond, W. J., 2003. Growing tall vs. growing wide: tree architecture and

allometry of A. karroo in forest, savanna, and arid environments. Oikos 102, 3-14.

Brain, P., 1985. Leaf peroxidase types in Acacia karroo. Geographical distribution and

influence of the environment. S. Afr. J. Bot. 52, 47-52.

Brain, P., 1989. Genetic races in a ring species of Acacia karroo. S. Afr. Sci. 85, 181-

185.

Coates-Palgrave, K., 2002. Trees of southern Africa. 3rd Edition, Struik, Cape Town. pp.

19, 286.

Coley, P.D., Bryant, J.P. and Chapin, F.S. III, 1985. Resource availability hypothesis and

plant antiherbivore defence. Science 230, 895-899.

Gottlieb, L. DL, 1977. electrophoretic evidence and plant systematics. Ann. Missouri

Bot. Gard. 64, 161-180.

Hamrick, J.L., Godt, M.J.W., 1989. Allozyme diversity in plant species. In: Plant

population, breeding, and genetic resources. Brown, A.H.D., Clegg, M.T., Kahler,

A.L., Weir, B.S. (Eds). Sinauer Associates, Sunderland, Massachusetts, pp. 43-63.

Pigliucci, M., Schlichting, C.D., 1997. On the limits of quantitative genetics for the study

of phenotypic evolution. Acta Biotheoret. 45, 143-160.

Swartz, P.P., 1982. A numerical-taxonomical evaluation of the taxon Acacia karroo

Hayne in southern Africa. Unpl. MSc thesis, University of Pretoria South Africa.

[‘n Numeries-taksonomiese evaluering van die takson Acacia karroo Hayne in

suidelike Afrika.]

92

Chapter 4

Van Straalen, N.M., Timmermans, M.J.T.N., 2002. Genetic variation in toxicant stressed

populations: An evaluation of the "genetic erosion" hypothesis. Hum. Ecol. Risk

Assess. 8, 983-1002.

93