RELATIONSHIPS BETWEEN CATTLE GRAZING AND

RIO GRANDE WILD TURKEYS IN THE

SOUTHERN GREAT PLAINS

by

GALON I. HALL, II, B.S.

A THESIS

IN

WILDLIFE SCIENCE

Submitted to the Graduate Faculty of Texas Tech University in

Partial Fulfillment of the Requirements for

the Degree of

MASTER OF SCIENCE

Approved

Mark C. Wallace Co-Chairperson of the Committee

Warren B. Ballard Co-Chairperson of the Committee

Donald C. Ruthven, III

Accepted

John Borrelli Dean of the Graduate School

May, 2005

ACKNOWLEDGEMENTS

This project would not have been possible without funding from Texas Parks and

Wildlife Department, Department of Kansas Wildlife and Parks, the National Wild

Turkey Federation, and Texas State Chapter of the National Wild Turkey Federation. All

of which provided monies used for vehicles, technicians, and stipends! I would also like

to thank Houston Safari Club for their gracious scholarship, which helped personal ends

meet when things were tight. I also thank the many private landowners that gave us

unrestricted access to the turkeys on their property, especially Mr. Larry McLean and Dr.

Carl Clary. Without their cooperation, we would have nothing!

I would like to thank Dr. Mark C. Wallace and Dr. Warren B. Ballard for their

never ending comments on how to improve this study. It definitely would not be what it

is without their guidance and input. I also thank them for their friendship and advice

(sometimes even when I did not want it!) about life and being a wildlife biologist. I

would like to thank Mr. Donald C. Ruthven for constructive criticism and field support

while in Paducah, Texas. In addition, there were many other Texas Parks and Wildlife

employees that assisted with trapping and extrication of stuck vehicles, specifically Mr.

Bill Adams, Mr. Larry Jones, Mr. Fred Stice, and Mrs. Dana Wright. I also owe a large

debt of gratitude to my fellow graduate students that embarked on this turkey journey

with me and contributed their thoughts and suggestions at exactly the right moments:

John Brunjes, Matt Butler, Derrick Holdstock, Rachael Houchin, Ross Huffman, and

Richard Phillips. Plus, one student who was not a turkey, Brady McGee, thanks Brady

ii

for your friendship and input into this project! I thank Byron Buckley, Danny Ferris,

Meredith Greene, and Kelly Reyna for their unending positive attitudes while collecting

data and providing the information for this project. The newest group of graduate

students (Brian Peterson, Ryan Swearingin, Ryan Walker) have no idea what they are

getting into, but I promise each of you that I will always be a mouse click or phone call

away if you ever need any assistance. I owe a lot of the positive grammatical changes

made in this manuscript to Ms. Tina Brunjes, without her input, the thoughts and ideas

would still be illogical and incomplete. Dan (you too Liz!), there is no way I can thank

you enough for your friendship the past 3 years. What began in Research Methods,

carried over into the wilds of Texas and Europe, and will continue to whatever is next.

Hopefully, I will be able to turn that little duck call into a musical instrument one day,

like you can!

Finally, I would like to thank my family. Mom and Dad (Hall and Grimm!),

thanks for always being supportive and encouraging of Maggie and I as we figure this

stuff out. It is always awesome to see how blessed we truly are to have been raised in

Christian homes with parents that know no bounds to their love. Maggie, where to start?

Thanks for putting up with long field seasons and reminding me that it would not last

forever. Thanks for putting up with me on days when things just did not go right. You

sacrificed so much so we could do this and I am forever grateful. I love you!!!!

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ii

ABSTRACT vii

LIST OF TABLES ix

LIST OF FIGURES xiii

CHAPTER

I. INTRODUCTION 1

1.1 Literature Cited 4

II. RELATIONSHIPS BETWEEN CATTLE GRAZING AND FEMALE RIO GRANDE WILD TURKEY HABITAT USE AND NESTING ECOLOGY IN THE SOUTHERN GREAT PLAINS 6 2.1 Introduction 6 2.2 Study Areas 10 2.3 Methods 16 2.3.1 Capture and Telemetry 16 2.3.2 Nesting 17 2.3.3 Vegetation Measurements 18 2.3.4 Habitat Selection 20 2.3.5 Grazing and Nesting 22 2.3.6 Available Vegetative Cover 23 2.3.7 Non-nesting Pasture Selection 24

iv

2.4 Results 25 2.4.1 Capture and Telemetry 25

2.4.2 Habitat Selection 26 2.4.3 Grazing and Nesting 27 2.4.4 Available Vegetative Cover 28 2.4.5 Non-nesting Pasture Selection 29 2.5 Discussion 30 2.6 Literature Cited 35 III. RELATIONSHIPS BETWEEN CATTLE GRAZING AND MALE RIO GRANDE WILD TURKEY HOME RANGE SIZE AND PASTURE USE IN THE SOUTHERN GREAT PLAINS 66 3.1 Introduction 66 3.2 Study Areas 68 3.3 Methods 75 3.3.1 Capture and Telemetry 75 3.3.2 Vegetation Measurements 76 3.3.3 Home Range Size Calculation 78 3.3.4 Turkey Home Ranges and Cattle 79 3.3.5 Anthropogenic Food Sources 80 3.3.6 Individual Locations and Cattle 80 3.3.7 Available Displaying Habitat 81

v

3.4 Results 82 3.4.1 Capture and Telemetry 82

3.4.2 Home Range Size Calculation 82 3.4.3 Turkey Home Ranges and Cattle 83 3.4.4 Anthropogenic Food Sources 83

3.4.5 Individual Locations and Cattle 84

3.4.6 Available Displaying Habitat 84

3.5 Discussion 85 3.6 Literature Cited 89 APPENDIX 110

vi

ABSTRACT

Previous studies on the response of female and male turkeys to grazing have

produced conflicting results, warranting further investigation. Our objectives were to

quantify habitat use by female Rio Grande wild turkeys (Meleagris gallopavo

intermedia) during the nesting period and determine possible relationships between cattle

grazing and nesting site selection. We also wanted to investigate changes in space use

and pastures used by male Rio Grande wild turkeys in the presence and absence of cattle.

From 2000-2004, we located 351 nesting sites from radio-transmittered birds in

the Texas Panhandle and southwestern Kansas. A logistic regression model comparing

nesting sites to random sites indicated horizontal visual obstruction, vertical visual

obstruction, and percentage of bare ground provided the highest predictive power

(P ≤ 0.003) for nesting site selection. Agricultural and upland zones were used less than

available and riparian zones were used more than available (P < 0.001) for nesting;

grazed pastures were used less than available and non-grazed pastures were used more

than available (P < 0.05) for nesting. Statistical differences in measured vegetative

characteristics were found primarily in compositional components among vegetative

zones; upland zone nesting sites had a higher percent shrub component (P ≤ 0.001) and

riparian zone nesting sites had a higher percent grass component (P ≤ 0.001). There were

no significant differences in measured vegetative characteristics among pasture types, but

there were differences in what was available for nesting in grazed and non-grazed

pastures. Grazed pastures consistently had less grass cover (P ≤ 0.018) and more bare

vii

ground (P ≤ 0.043). Because of cattle impacts on grass availability, grazing would likely

have the highest impact on nesting in riparian zones due to the high use of grass in

riparian zones as Rio Grande turkey nesting cover. An appropriate grazing plan to

promote Rio Grande turkey nesting habitat would include grazing upland zones in the

spring, when it likely has little impact on nesting site selection, and grazing riparian areas

following nesting season.

We recorded telemetry locations of radio-transmittered male turkeys in the Texas

Panhandle and southwestern Kansas during the same period. Area-observation curves

indicated that ≥25 locations per bird were adequate for home range calculation. The

average home range size for adult male birds on all study sites was 1,830 ha and for

juvenile male birds on all study sites was 1,475 ha. Our analysis of home range sizes of

male turkeys at the Matador study area contained a lot of variation, and there were too

many confounding factors that influenced home range sizes. However, we did find that

52.6% of male home ranges contained a known anthropogenic food source. A more

effective analysis of cattle relationships involved comparing individual male locations

with cattle presence or absence. We found no selection for grazed or non-grazed pastures

(p > 0.05) by male Rio Grande turkeys. This differs from reported female pasture use

and indicated a difference between the sexes in response to grazing. Grazing at light to

moderate intensities with periods of rest did not affect male turkey pasture use and

continued to maintain open areas used by male turkeys for displaying purposes.

viii

LIST OF TABLES

2.1 Total number of Rio Grande wild turkey nests documented from January 2000 through August 2004 on 4 study areas in the Southern Great Plains. 42

2.2 Comparison of habitat characteristics measured at Rio Grande wild

turkey nesting (n = 351) and paired nest random sites (n = 345) from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). 43

2.3 Selection of vegetative zones for nesting sites by female Rio Grande

wild turkeys from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 2: χ² = 5.99; calculated χ² was 488.92. 44

2.4 Nest success associated with vegetative zones used by female Rio

Grande wild turkeys from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). 45

2.5 Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in each vegetative zone from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). 46

2.6 Selection of pasture for nesting sites by female Rio Grande wild turkeys

from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 10.51. 47

2.7 Nest success associated with pasture types used by female Rio Grande

wild turkeys from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 48

2.8 Comparison of habitat characteristics measured at Rio Grande wild

turkey nesting sites in each pasture type from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 49

ix

2.9 Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in upland grazed and upland non-grazed categories from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 50

2.10 Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in riparian grazed and riparian non-grazed categories from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 51

2.11 Nest success associated with the 4 categories used by female Rio

Grande wild turkeys from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 52

2.12 Comparison of habitat characteristics measured at grazed random

plots and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 53

2.13 Comparison of habitat characteristics measured at riparian random

plots and upland random plots from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). 54

2.14 Comparison of habitat characteristics measured at riparian grazed

random plots and riparian non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 55

2.15 Comparison of habitat characteristics measured at upland grazed

random plots and upland non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 56

2.16 Selection of pasture types by non-nesting female Rio Grande wild

turkeys using all locations from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 336.08. 57

x

2.17 Selection of pasture types by non-nesting female Rio Grande wild turkeys during the breeding season (1 April-31 August) from April 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 129.90. 58

3.1 Average breeding season (1 April -31 August) home range sizes of

male Rio Grande wild turkeys on 4 study areas from January 2000 through August 2004 in the southern Great Plains. 95

3.2 Average reported home range sizes in the scientific literature for

male turkeys from 1963 to 2005 (1963 -1975 list compiled by Brown 1980). 96

3.3 Selection of pasture types by male Rio Grande wild turkeys using all

locations from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall chi-square critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 0.12. 97

3.4 Selection of pasture types by male Rio Grande wild turkeys during the

breeding season (1 April-31 August) from April 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall chi-square critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 0.16. 98

3.5 Comparison of habitat characteristics measured at grazed random plots

and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). 99

A.1 Grazing rotations on the Matador Wildlife Management Area from January 2000 through August 2004. 111

A.2 Grazing rotations on the Matador study area adjacent land from January 2000 through August 2004. 112

A.3 Grazing rotations on the Gene Howe Wildlife Management Area from January 2000 through August 2004. 113

A.4 Grazing rotations on the Cimarron National Grasslands from January 2000 through August 2004. 114

xi

A.5 Availability of measured vegetative characteristics in grazed and non-grazed pastures from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84. 119

A.6 Availability of measured vegetative characteristics in riparian and

upland zones from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84. 122

xii

LIST OF FIGURES

1.1 State of Texas Rio Grande wild turkey distribution map, updated 2004 by Texas Parks and Wildlife Department. 5

2.1 Location of 4 study areas where relationships between cattle grazing

and female Rio Grande wild turkey habitat use and nesting ecology were studied in the southern Great Plains from January 2000-August 2004. 59

2.2 Description of vegetation plot used to measure data regarding female Rio Grande wild turkey habitat use and availability in the southern Great Plains from January 2000-August 2004. 60

2.3 Bar graph of nest initiation dates for female Rio Grande wild turkeys

from 4 study areas in the southern Great Plains. The years (2000-2004) and study areas (Matador, Salt Fork, Gene Howe, Cimarron) are pooled. The first number corresponds to the first known nest initiation day (1 April) and the last day corresponds to the last known nest initiation day (3 July). 61

2.4 Comparison of plot proportions with each respective class of grass cover percentages between grazed and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant. 62

2.5 Comparison of plot proportions with each respective class of forb

cover percentages between grazed and non-grazed random plots on 3 study areas in the Southern Great Plains from January 2000 through August 2004 (pooled over years and study areas). Bonferroni intervals that overlap are not significant. 63

2.6 Comparison of plot proportions with each respective class of grass cover percentages between upland and riparian random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant. 64

2.7 Comparison of plot proportions with each respective class of shrub cover percentages between upland and riparian random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant. 65

xiii

3.1 Location of 4 study areas where relationships between cattle grazing and male Rio Grande wild turkey home range size and pasture use were studied in the southern Great Plains from January 2000-August 2004. 100

3.2 Description of vegetation plot used to measure data regarding Rio Grande wild turkey habitat use and availability in the southern Great Plains from January 2000-August 2004. 101

3.3 Bar graph of nest initiation dates for female Rio Grande wild turkeys

from 4 study areas in the southern Great Plains. The years (2000-2004) and study areas (Matador, Salt Fork, Gene Howe, Cimarron) are pooled. The first number corresponds to the first known nest initiation day (1 April) and the last day corresponds to the last known nest initiation day (3 July). 102

3.4 Average area-observation curve describing the relationship between

the number of telemetry locations and home range estimates for radio-transmittered male Rio Grande wild turkeys in the southern Great Plains. 103

3.5 Male home range comparison between the treatment (cattle removed

2003-2004) and control groups with pooled adult and juvenile home range sizes on the Matador study area in Paducah, Texas. 104

3.6 Male home range comparison between the treatment (cattle removed

2003-2004) and control groups with juvenile home range sizes only on the Matador study area in Paducah, Texas. 105

3.7 Male home range comparison between the treatment (cattle removed

2003-2004) and control groups with adult home range sizes only on the Matador study area in Paducah, Texas. 106

3.8 Comparison of plot proportions with each respective class of shrub cover percentages between grazed and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant. 107

3.9 Comparison of plot proportions with each respective class of bare ground percentages between grazed and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant. 108

xiv

3.10 Selected sample of 4 male home ranges, showing use of anthropogenic food sources (gray areas) and home range size variability on the Matador study area. 109

A.1 Map of pasture boundaries on the Matador Wildlife Management Area. 115 A.2 Map of pasture boundaries on the Gene Howe Wildlife Management Area. 116 A.3 Map of pasture boundaries on the Cimarron National Grasslands. 117

xv

CHAPTER I

INTRODUCTION

The North American wild turkey (Meleagris gallopavo) is one of the management

success stories of federal and state wildlife agencies across North America. Populations

were at their lowest numbers toward the end of the 19th century; with most birds

surviving only in areas that included the most inaccessible cover (Kennamer et al. 1992).

Their population declines were caused by removal of trees and over harvest by settlers as

new areas of the United States were colonized (Kennamer et al. 1992). After forest

stands began to regenerate (both artificially and naturally) and predator control programs

to protect domestic livestock were initiated, wild turkey population numbers began to

increase. Active restoration programs began after World War II. Presently, viable wild

turkey populations are found throughout the contiguous United States, and Hawaii.

Considering they were historically found in only 39 states, turkey recovery is quite a

success story (Kennamer et al. 1992)

There are 5 wild turkey subspecies found in the United States: Florida wild

turkey (Meleagris gallopavo osceola), Eastern wild turkey (M.g. silvestris), Merriam's

wild turkey (M.g. merriami), Rio Grande wild turkey (M.g. intermedia), and Gould's wild

turkey (M.g. mexicana) (Kennamer et al. 1992). Of these 5 subspecies, 3 are currently

found in Texas: Rio Grande, Eastern, and Merriam's. The Rio Grande is similar in

appearance to other subspecies, but can be distinguished by coloration on the tips of the

1

tail feathers and upper tail coverts. These tips are tan in color in Rio Grande's, as

opposed to dark brown or white on other subspecies (Beasom and Wilson 1992).

As a result of the decline in turkey numbers in the late 1800's, first limits on

hunting wild turkeys in Texas were initiated, forbidding turkey hunting 5 months out of

the year. In 1903, a 25 turkey per day bag limit was declared, but there were limited

resources to enforce the law and turkey population numbers continued to decline

throughout most of their range as a result of continued overharvest. In 1919, bag limits

were reduced to 3 bearded gobblers per season, which helped turkey populations recover

(Suarez 2002). Today there are over 600,000 turkeys across Texas and Kansas, with the

Rio Grande subspecies being the most numerous (Kennamer et al. 1992).

In Texas, the increase in turkey numbers was greatly influenced by successful

restocking with wild-trapped birds. The first reported Texas trapping was in 1924 by

F.M. Cowsert, an employee of Texas Game, Fish and Oyster Commission (Glazener

1963, Beasom and Wilson 1992). Currently, the Rio Grande wild turkey is well

established across Texas (Figure 1.1). Rio Grande wild turkeys were reintroduced into

Kansas in 1964 and 1966 using Texas stock. Overall, Rio Grande turkey populations

have done well in Kansas because they are well adapted to open grassland areas dissected

with riparian vegetative communities (United States Department of Agriculture 1981).

The Rio Grande turkey occupies a wide variety of grassland, savanna, and shrubland

habitats across its range from Mexico to Kansas (Garrison et al. 1977).

In the 1990's on the southern Rolling Plains, it was estimated, based on annual

poult-hen counts from Texas Parks and Wildlife Department (TPWD) (G. Miller,

2

unpublished data, TPWD, Canyon, TX), that some Rio Grande populations were

declining and a large-scale research project was started at Texas Tech University.

Several factors were proposed that could be linked to the declining populations,

including; changes in land-use practices, loss of roost trees, advancing vegetative

succession, inadequate nesting habitat, and predation (Ballard et al. 2001).

Because of often-conflicting data regarding how livestock grazing may affect

ground nesting and foraging birds, such as turkeys, an in-depth study was needed on

possible relationships between livestock grazing and Rio Grande wild turkey nesting

ecology, habitat selection, and home range size and location. The following chapters

investigate relationships between Rio Grande wild turkeys and cattle grazing as a land-

use practice from January 2000 through August 2004 in the southern Great Plains. This

work represents my ability and desire to conduct scientific investigations and my skills in

statistical analysis of data and interpretation of results. This project was a collaborative

effort among many other graduate students and technicians and represents their hard

work as well. All chapters were written with the intention of publishing the information

in peer-reviewed journals following completion of this thesis. Chapter II investigates

relationships between cattle grazing and female Rio Grande wild turkey habitat use and

nesting ecology in the Southern Great Plains. Chapter III investigates relationships

between cattle grazing and male Rio Grande wild turkey home ranges and pasture use in

the Southern Great Plains

3

1.1 Literature Cited

Ballard, W. B., M. C. Wallace, J. H. Brunjes, T. Barnett, D. Holdstock, R. Phillips, B. Spears, M. Miller, B. Simpson, S. Sudkamp, R. Applegate, R. Gipson. 2001. Changes in land use patterns and their effects on Rio Grande turkeys in the Rolling Plains of Texas, Annual Report-2001. Department of Range, Wildlife, and Fisheries Management, Texas Tech University, Lubbock, Texas, USA.

Beasom, S. L., and D. Wilson. 1992. Rio Grande Turkey. Pages 306-330 in J.G. Dickson, editor. The wild turkey: biology and management. Stackpole Books, Mechanicsburg, Pennsylvania, USA.

Garrison, G. A., A. J. Bjugstad, D. A. Duncan, M. E. Lewis, and D. R. Smith. 1977. Vegetation and environmental features of forest and range ecosystems. United States Department of Agriculture, Agricultural Handbook 475.

Glazener, W. C., A. S. Jackson, and M. L. Cox. 1964. The Texas drop-net turkey trap. Journal of Wildlife Management 28: 280-287.

Kennamer, J. E., M. C. Kennamer, and R. Brenneman. 1992. History. Pages 6-17 in J.G. Dickson, editor. The wild turkey: biology and management. Stackpole Books, Mechanicsburg, Pennsylvania, USA.

Suarez, R. Texas Turkey Talk. 2002. Texas Parks and Wildlife Department, Austin, Texas, USA.

United States Department of Agriculture - Soil Conservation Service. 1981. Habitat management for turkeys. Salina, Kansas, USA.

4

Figure 1.1. State of Texas Rio Grande wild turkey distribution map, updated 2004 by Texas Parks and Wildlife Department.

5

CHAPTER II

RELATIONSHIPS BETWEEN CATTLE GRAZING AND

FEMALE RIO GRANDE WILD TURKEY HABITAT

USE AND NESTING ECOLOGY IN THE

SOUTHERN GREAT PLAINS

2.1 Introduction

Interactions between cattle grazing and various species of wildlife have been

extensively studied throughout the United States, particularly the effects of cattle

foraging on wild ungulate behavior (Bryant et al. 1981, Stover 1985, Kie et al. 1991, Loft

et al. 1991). Livestock grazing and trampling on a seasonal basis affected understory

vegetation that provided wildlife cover, but effects varied with vegetation type and

stocking rate (Loft et al. 1987).

Grazing has been used as a tool to manage herbaceous cover and improve forage

quality. Light to moderate summer cattle grazing improved habitat for mule deer

(Odocoileus hemionus) (Holechek et al.1982) and a site specific grazing plan was used to

manage waterfowl nesting cover in eastern Washington (Rees 1981). In northeastern

Oregon, livestock grazing was used to improve forage quality for elk (Cervis canadensis)

(Anderson and Scherzinger 1975).

The relationship between cattle and gallinaceous birds varied depending on bird

species and local habitat. Blue grouse (Dendragapus obscurus pallidus) density indices

were not different between grazed and non-grazed areas (Zwickel 1972) and even when a

6

pasture was overgrazed, grouse did not move from their traditional use areas to adjacent

non-grazed pastures or areas of taller grass (Nielson and Yde 1981). Investigations of the

effects of specific grazing regimes on bobwhite quail (Colinus virginianus) described

either positive or neutral effects of grazing on specific aspects of bobwhite ecology

(Baker and Guthery 1990, Bareiss et al. 1986, Wilkins and Swank 1992).

For the purposes of this investigation, we defined habitat as all abiotic and biotic

features of the environment (Litvaitis et al. 1996), specifically the spatial arrangement of

food, water, and cover. Habitat selection is different than use in that it implies a favoring

of 1 habitat among numerous alternative habitats. According to Johnson (1980), use is

selective if habitat types are used disproportionally to their availability.

Wild turkey (Meleagris gallopavo) nesting ecology, in terms of vegetative

characteristics, has been extensively studied. Nesting cover for Rio Grande turkeys on

the Welder Wildlife Refuge in south Texas was >45 cm in height where shrubs were

selected and in 50% of the cases where herbaceous cover was used (Ransom et al. 1987).

The vertical cover for Merriam's wild turkey nest sites in South Dakota was overhanging

vegetation within 1 m above the nest (Wertz and Flake 1988). Ground cover may be

important because it can inhibit nest depredation and loss of poults, as well as increase

survival of the hen (Glazener 1958, Beasom 1970). In Arkansas, Texas, South Dakota,

and Arizona, hens selected sites with greater visual obstruction around the nest (Badyaev

1995, Cook 1972, Day et al. 1991, Wakeling et al. 1998). In Colorado and South Dakota,

turkey nest habitat did not differ or differed only slightly between successful and

unsuccessful nests (Schmutz et al. 1989, Rumble and Hodorff 1993). Miller et al. (1995)

7

found that renesting Rio Grande females in south-central Kansas (after loss of first nest)

experienced higher survival rates than females incubating first nest attempts, possibly due

to the influence of rain on the vegetative communities providing taller and denser cover

for nesting.

There have been a few published manuscripts studying effects of different

livestock rotation methods or stocking densities on wild turkey ecology. Merriam's

turkeys (M.g. merriami) in New Mexico showed no preference for grazed or non-grazed

pastures in a rest-rotation system (Jones 1981). Walker (1948) indicated that 72% of

turkeys on the Edwards Plateau in south-central Texas, were concentrated on lightly

grazed or deferred pastures. Rio Grande turkeys (M.g. intermedia) in south Texas

avoided stock tanks at the hub of short duration grazing systems (Prasad and Guthery

1986). According to Schulz and Guthery (1987), horizontal visual obstruction was lower

in short duration grazing than in continuous grazing. Grazing effects on wild turkey

nesting site selection varied between several different studies (Merrill 1975, Ransom

et al. 1987, Wertz and Flake 1988). Rest-rotation and deferred-rotation grazing systems

can be beneficial to game birds because they provided pastures free from disturbance

during nesting and other critical seasons (Holechek et al. 1982). Bareiss et al. (1986)

compared impacts of short duration grazing and continuous grazing on available nesting

ground cover for wild turkeys and found that grazing treatment had no effect. However,

in both Texas and South Dakota, low, thorny brush may have protected residual cover

from grazing, neutralizing the effects of different stocking intensities on nesting wild

turkeys (Bareiss et al. 1986, Wertz and Flake 1988). Trampling of nests by cattle was

8

thought to impact nest success of ground nesting birds, but Koerth et al. (1983) found

trampling effects to be the same in short duration and continuous grazing situations.

Riparian zone use by cattle was linked to decreases in riparian vegetation

(Szaro and Pase 1983, Sedgwick and Knopf 1991) and small mammal density indices

(Kauffman et al. 1981). Cattle preferred stream riparian zones to upland range sites,

likely because of water availability and forage quality (Ames 1978, Pinchak et al. 1991,

Roath and Krueger 1982, Smith et al. 1992). Ammon and Stacey (1997) found that

vertical vegetation diversity was lower on grazed than on rested pastures. They

concluded that livestock grazing may not only affect availability of nesting substrates for

birds in riparian zones, but could also facilitate nest predation. Improper use of riparian

areas can result in altered plant composition and increased erosion (Knopf and Cannon

1982, Skovlin 1984). In Colorado, Schulz and Leininger (1990) found differences in

vegetative composition between cattle grazed (600 AUM's) and non-grazed zones

(exclosures established in 1956): shrub canopy cover and grass cover were greater in

grazing exclosures and bare ground was 4 times greater in grazed areas. In Texas, it was

recommended that riparian corridors be established as separate pastures so grazing can be

used as a tool and complete removal of cattle, when necessary, can be carried out

(Miller and Ray 1996).

The objectives of this study were to quantify habitat use by female Rio Grande

wild turkeys during the nesting period and determine possible relationships between

cattle grazing, nest success, and nesting site selection. We hypothesized that female Rio

Grande wild turkeys would select riparian zones for nesting sites in greater proportion to

9

availability than other vegetative zones because of the increased vegetative cover, which

would also result in increased nest success. We predicted that composition of visual

obstruction would vary, but overall vegetative structure associated with nests would

remain the same (structure has greater effect on nesting site selection than composition).

We also predicted the presence of cattle would directly affect the availability of nesting

substrate by reducing the amount of vegetative cover, but that female Rio Grande turkeys

would select nesting sites in grazed areas that were structurally the same as in non-grazed

areas.

2.2 Study Areas

Turkey movements were used to determine study area boundaries. Land use at

each of the study areas included production of cattle, cotton, wheat, and grain sorghum.

Grazing occurred at various levels on upland and riparian sites and cattle densities ranged

from 1.8 to 31 ha/animal unit (AU) (Table A.1-A.4; Figure A.1-A.3). All cattle operators

provided supplemental feed with hay bales or cottonseed cake pellets Data were

collected on female Rio Grande turkey nesting site selection and habitat use from January

2000 through August 2004.

Turkeys were captured on 4 study areas: 3 in the Texas Panhandle, and 1 in the

southwestern corner of Kansas (Figure 2.1). The southernmost site was the Matador

study area in Cottle County, Texas. It was located in the lower Rolling Plains and

consisted of 11,370 ha of public land (Matador Wildlife Management Area, MWMA)

with an additional 16,133 ha of adjacent private lands. Dry winters and hot summers

10

characterized the climate of Cottle County. Rainfall events usually occurred as

thunderstorms and average precipitation was 56.24 cm. However, monthly and annual

amounts were highly variable. Maximum precipitation usually occurred in May and June

and periods of drought were very common. The average growing season length was 219

days. The dominant soil association was Miles Springer with nearly level to strongly

sloping, deep, coarse textured and moderately coarse textured soils on outwash plains.

Secondary associations were Nobscot-Heatly with nearly level to gently sloping, deep,

coarse-textured soils on upland plains, and Woodward-Carey, with nearly level to gently

sloping, deep to moderately deep, medium-textured soils on upland plains. Soil series

included Quinlan (loamy, mixed, thermic, shallow; inceptisols), Woodward (coarse-silty,

mixed, thermic; inceptisols), Hilgrave (loamy-skeletal, mixed, thermic; alfisols), Yahola

(coarse-loamy, mixed calcareous, thermic; entisols), Enterprise (coarse-silty, mixed,

thermic; inceptisols), Lincoln (sandy, mixed, thermic; entisols), Miles (fine-loamy,

mixed, thermic; alfisols), and Springer (coarse-loamy, mixed, thermic; alfisols)

(Richardson et al. 1974). Primary woody vegetation included honey mesquite (Prosopis

glandulosa), redberry juniper (Juniperus pinchotii), netleaf hackberry (Celtis reticulata),

eastern cottonwood (Populus deltoides), salt cedar (Tamarix gallica), Chickasaw plum

(Prunus angustifolia), and sand sagebrush (Artemisia filifolia). Grasses found on the area

included little bluestem (Schizachyrium scoparium), sand dropseed (Sporobolus

cryptandrus), sideoats grama (Bouteloua curtipendula), purple threeawn (Aristida

purpurea), Japanese brome (Bromus japonicus) and plains bristlegrass (Setaria

leucopila). Primary forb species included western ragweed (Ambrosia psilostachya),

11

woolly plantain (Plantago insularis), annual sunflower (Helianthus annus), and lamb's

quarters (Chenopodium album). The confluence of the South Pease River and the Middle

Pease River occurred on the study area. The MWMA had a rest-rotation, cow-calf

operation grazing program from January 2000 through September 2002, and cattle were

removed from MWMA October 2002 through September 2004. Cattle stocking densities

ranged from 1.82-13.37 ha/AU. Adjacent lands also used forms of rotational grazing or

continuous grazing, and based the timing of cattle movement on available standing forage

and supplemented with cattle feed. The herd was primarily a cow/calf operation.

Densities on adjacent lands ranged from 2.09-31.54 ha/AU. On both the MWMA and

adjacent lands, pastures were grazed for at least 1 month before the herd was moved.

The Salt Fork study area was located in Collingsworth and Donley Counties, and

was bisected by the Salt Fork of the Red River. It was located near the Caprock

escarpment below the edge of the High Plains and was centered on private ranches with a

total of 17,000 ha. The Salt Fork study area was characterized by a dry-steppe climate

with mild winters. Average annual precipitation was 54.64 cm, most of which occurred

from April through October. The growing season averaged 206 days. Common soil

associations included Mobeetie-Veal-Potter (deep to very shallow, gently sloping to

steep, loamy soils; on uplands), Obaro-Aspermont-Quinlan (deep to shallow, gently

sloping to steep, loamy soils; on uplands), and Springer-Lincoln-Likes (Deep, nearly

level to sloping, sandy soils; on uplands and bottomlands). Soil series included Quinlan,

Springer, Miles, Enterprise, Wichita (fine, mixed, thermic; alfisols), Lutie (fine-silty,

mixed, thermic; mollisols), Ector (loamy-skeletal, carbonatic, thremic; mollisols), and

12

LaCasa (fine, mixed, thermic; mollisols) (McEwen et al. 1973). Primary woody

vegetation included eastern cottonwood, honey locust (Gleditsia triacanthos), black

locust (Robinia pseudo-acacia), salt cedar, sand sagebrush, and Chickasaw plum. Grass

species included sideoats grama, hairy grama (Bouteloua hirsuta), western wheatgrass

(Elytrigia smithii), and little bluestem. Forb species included western ragweed, annual

sunflower, and broom snakeweed (Gutierrezia sarothrae). No grazing information was

collected at the Salt Fork study area, so all grazing analyses excluded Salt Fork data.

The Gene Howe study area was located in Hemphill County, near Canadian,

Texas, and was bisected by the Canadian River. It was located in the Rolling Plains and

was centered on 2,180 ha of public land (Gene Howe Wildlife Management Area,

GHWMA) with access to 11,000 ha of adjacent private lands. Hemphill County was

characterized by a dry-steppe climate with 52.07 cm of average annual precipitation,

most frequently in the form of thunderstorms. Periods of 2-3 weeks with no rain were

fairly common and monthly periods of no rainfall were recorded. The average annual

growing season was 204 days. Soil associations included Tivoli-Springer (deep, sandy

soils on upland dunes and hummocks), Mobeetie-Berda-Potter (deep and very shallow,

gently sloping to steep, loamy soils on uplands), and Lincoln-Sweetwater (sandy and

loamy soils on bottomlands). Soil series included Linclon, Enterprise, Springer,

Sweetwater (fine-loamy over sandy or sandy-skeletal, mixed calcareous, thermic;

mollisols), Likes (mixed, thermic; entisols), Tivoli (mixed, thermic; entisols), Tipton

(fine-loamy, mixed, thermic; mollisols), and Bippus (fine-loamy, mixed, thermic;

mollisols) (Williams et al. 1974). Dominant woody vegetation included eastern

13

cottonwood, Russian olive (Eleagnus angustifolia), salt cedar, Chickasaw plum and sand

sagebrush. Grass species present on the area included little bluestem, sideoats grama,

Canada wildrye (Elymus canadensis), sand dropseed, purple threeawn, and western

wheatgrass. Primary forb species included western ragweed, annual sunflower, and

silverleaf nightshade (Solanum eleagnifolium). The GHWMA had a rest-rotation, cow-

calf operation grazing program from September 2001 through September 2004. Cattle

stocking densities ranged from 4.39-11.52 ha/AU. Pastures were grazed for 6 month

periods, then rested for 6 months before grazed again. No cattle information was

collected from adjacent lands.

The Cimarron study area was in the southwestern corner of Kansas near Elkhart,

Kansas in Morton County on 29,648 ha of public land (Cimarron National Grasslands,

CNG) and 15,000 ha of adjacent private land. The Cimarron River bisected the study

area. Climate of the Cimarron study area was semiarid with haphazard precipitation

events. Average annual precipitation was 42.60 cm, which mostly occurred in July.

Drought periods often extended past 30 days. Average length of the growing season was

214 days. Primary soil associations included Richfield-Ulysses (loamy soils of uplands),

Vona-Tivoli (rolling sandy land), and Otero-Lincoln (soils of the Cimarron River Valley

and adjacent slopes). Soil series included Tivoli, Lincoln, Dalhart (deep, dark, nearly

level and gently sloping sandy soils of the upland; regosols), Richfield (deep, dark, nearly

level upland soils; regosols), Vona (sandy soils of upland), Bridgeport (gently sloping,

north wall of Cimarron River; alluvial), Mansker (loamy soils of upland; calcisols), Potter

(shallow, overlie caliche or limestone; lithosols) (Dickey et al. 1963). Primary woody

14

species were eastern cottonwood and salt cedar in the riparian corridor, and sand

sagebrush on the uplands adjacent to the riparian corridor. Grasses found on the study

area included sand bluestem (Andropogon hallii), blue grama (Bouteloua gracilis),

sideoats grama, sand dropseed, sand lovegrass (Eragrostis trichodes), prairie sandreed

(Calamovilfa longifolia), and buffalo grass (Buchloe dactyloides) (Spears 2002).

Common forbs included broom snakeweed, western ragweed, and daisy fleabane

(Erigeron annus). The CNG had a rest-rotation, cow-calf operation grazing program

from January 2000-September 2004. Cattle stocking densities ranged from 2.52-10.1

ha/AU. Pastures were grazed for 1-2 months per rotation period and then rested for up to

1 year. No cattle information was collected from adjacent lands.

Hunting turkeys was allowed throughout the Texas Panhandle during our study,

with a maximum of 4 birds harvested per hunter in both the fall and spring season

combined. Fall hunting was split into 2 periods, a 4-week archery season from late

September to late October and a 9-week gun season from early November through early

January. Both sexes were legal during the fall season only (Texas Parks and Wildlife

Department 2004). The Kansas turkey hunting season occurred only in the spring and

went from early April through mid-May each year with a 1 bearded turkey bag limit

(Kansas Department of Wildlife and Parks 2004). Across the 5 years of the study,

hunters legally harvested a total of 49 transmittered male turkeys on private lands. On

the two wildlife management areas (MWMA, GHWMA), bag limits were restricted to

one bird during the spring season and hunts were by special permit only. Each spring on

MWMA, beginning in 2001, 5-10 permits were issued for male turkeys only. In 2001

15

through 2004, a total of 2, 3, 3, and 5 males, respectively, were harvested (Donald

Ruthven, Texas Parks and Wildlife Department, personal communication). On GHWMA

from 2000-2004, 4, 8, 7, 6, and 4 males, respectively, were harvested (Derrick Holdstock,

Texas Parks and Wildlife Department, personal communication). None of the

management area harvested turkeys were transmittered birds.

2.3 Methods

2.3.1 Capture and Telemetry

We trapped Rio Grande wild turkeys using rocket nets (Bailey et al. 1980) and

drop nets (Glazener et al. 1964) on sites baited with corn or grain sorghum from January

through March of each year. We captured additional birds each year to maintain 75

active transmitters at each study area. Upon capture, we recorded age and sex of each

bird (Pelham and Dickson 1992), and placed 110-gram backpack-style radio transmitters

from Advanced Telemetry Systems (ATS, Isanti, Minnesota, USA) or AVM Instruments

(Livermore, CA, USA) on selected individuals. Transmitters were equipped with a

mortality switch that activated after 8-hrs of inactivity. Birds were also fitted with Texas

Parks and Wildlife aluminum leg bands (size 8 for females) for further identification. We

located transmittered birds with ATS receivers, a hand-held 3-element yagi antenna, a

truck mounted omni-directional antenna, and a truck-mounted null-peak system

(Balkenbush and Hallett 1988, Samuel and Fuller 1996). We located each radio-tagged

turkey ≥2 times per week during the breeding season (1 April to 31 August) and once per

week throughout fall and winter (1 September to 31 March). We collected both visual

16

observations of transmittered birds and radio telemetry triangulation locations with a null

peak system (Samuel and Fuller 1996). Universal Transverse Mercator (UTM)

coordinates were collected on visual sightings using a Trimble Geoexplorer 2 or

Geoexplorer 3 (Trimble Navigation Limited, Sunnyvale, California, USA). The goal of

triangulation was to obtain at least 3 compass bearings, separated by >45° within 1 hour

total elapsed time from first bearing taken (White and Garrott 1990). We stratified

locations into 4 time periods based on turkey behavior: roosting (from dusk until dawn),

morning (AM) feeding (first 1/3 of daylight hours), midday (second 1/3 of daylight

hours), and afternoon (PM) feeding (third 1/3 of daylight hours). We used the computer

program Location of a Signal (LOAS, Ecological Software Solutions, Sacramento,

California) and the associated maximum likelihood estimator method to generate UTM

positions of the triangulated animals and a corresponding error ellipse. We determined

study area specific telemetry error ≥1 time per year by triangulating known location

transmitters. Each transmitter was placed at distances commonly associated with wild

turkey telemetry from our study areas (≤3 km). We also used known location

transmitters to adjust for system biases and to calibrate the truck mounted null-peak

system (White and Garrott 1990, Samuel and Fuller 1996).

2.3.2 Nesting

We determined onset of incubation when we had ≥3 locations of a hen in the

same place during the nesting season (Miller et al. 1998, Keegan and Crawford 1999).

Once birds were believed to be incubating, the nest was circled at a radius of ≈50 m to

17

locate the nest (Everett et al. 1980). We monitored each nesting hen to determine hatch

date and nest success. At 14 days, we walked-in and flushed the hen to determine the

number of eggs in the nest. We also floated ≥3 eggs to confirm nest initiation date and

obtain the best estimate of hatching date (Westerkov 1950). The nest was monitored at

least once every 3 days, until the hen vacated the nest area. Once vacated, the actual nest

bowl was found and nest outcome was determined based on shell remains (Schmutz and

Braun 1989). If only pieces of eggshells or crushed eggs remained, and a follow up flush

of the hen revealed no poults, we considered the nest unsuccessful. If there were eggshell

pieces indicating a possible hatch and a follow-up hen flush revealed at least 1 poult, we

considered the nest successful.

2.3.3 Vegetation Measurements

In order to describe vegetative characteristics at each nesting site, we measured

vegetation structure and composition after the hen vacated the nesting area. Along with

the nesting site plot, a paired nesting site random (PNR) plot located 50 m in a randomly

chosen cardinal direction was measured. In addition, random plots were measured 50 m

in a random cardinal direction from all associated observed turkey behavior locations. At

each plot, we established a 10 m x 20 m plot oriented north to south around the center

point (Figure 2.2). We measured percentage of vertical visual obstruction using an ocular

tube (i.e., 2-4 cm in diameter with crosshairs at one end) at 20 evenly spaced points

around the plot perimeter (Dueser and Shugart 1978). We recorded each reading as a hit

(i.e., canopy covered the crosshair) or a miss. We divided the number of hits by 20 and

18

multiplied by 100 for a percent estimate of vertical visual obstruction. We measured

percentage of horizontal visual obstruction using a visual obstruction pole (Robel et al.

1970) with 10 red and white 1-dm bands. We recorded the lowest Robel pole dm segment

completely visible at each measurement point, including center point. We placed the

pole along the plot centerline and observed it from a distance of 4 m and a height of 1 m

(Figure 2.2, point c to point d). We took readings along the centerline (Figure 2.2,

point a to point b) every 2 m, alternating sides of the centerline, for a total of 10

horizontal visual obstruction measurements (including a reading at 20 m). We classified

ground cover in the plot area into 5 categories: grass, shrub, bare ground, forb, and litter.

We estimated ground cover using the ocular tube along the 4 m chain between the Robel

and sighting pole. Ground cover classes immediately below the crosshairs were recorded

for that data point, from which a frequency estimate of percent ground cover of each class

was calculated (% grass cover, % shrub cover, % bare ground cover, % forb cover,

% litter cover). In 2003-2004, we adjusted vegetation measurement methods slightly to

improve sampling efficiency. We continued to measure vegetative characteristics at

nesting sites and PNR locations to quantify turkey habitat features. However, to assess

overall habitat availability, we added 200 randomly selected point vegetation plots at

each study area per year (2003-2004). We recorded 4 horizontal visual obstruction

measurements at the plot center point (1 in each cardinal direction) and reduced the

number of readings down the plot centerline to 5 alternating in east-west cardinal

directions. Thus, we recorded ground vegetation along 5 transects instead of 10. If the

point on the ground revealed a shrub, we recorded species and height class of the shrub.

19

We also measured vertical visual obstruction using a spherical densiometer held at 1 m

height above plot center point (Lemmon 1956). We took 1 reading in each of the 4

cardinal directions and averaged the 4 readings together to compute % vertical visual

obstruction.

2.3.4 Habitat Selection

We analyzed female Rio Grande wild turkey habitat selection at 2 scales. We

investigated third-order selection (Johnson 1980) of vegetation zones used for nesting

and fourth-order selection (Johnson 1980) based on nesting site habitat characteristics.

We compared vegetative characteristics between nesting sites and PNR locations

(with years and study sites pooled) using complete model and forward stepwise logistic

regression (P < 0.20 to enter or remove a variable; Hosmer and Lemeshow 1989) to

determine the best predictor of nesting sites relative to PNR locations. Vegetative

characteristics used to differentiate nesting sites from random sites were % vertical visual

obstruction, % horizontal visual obstruction, % grass cover, % shrub cover, % bare

ground cover, % forb cover, and % litter cover.

We identified vegetative types using aerial photos and vegetation maps at all 4

study areas. The resulting data was converted into a vegetative coverage layer in a

geographic information system (GIS) database (ArcView, ESRI, Los Angeles, California,

USA) (Brunjes 2005). For this analysis, we combined the 28 vegetative types defined in

the GIS database into 3 zones: urban-agriculture-Conservation Reserve Program lands

(UACRP), upland-non-riparian brushland and trees (UP), and riparian (RP). Availability

20

of each vegetative zone was calculated based on its accessibility to nesting turkeys

(Johnson 1980). The areas of available vegetative zones were calculated based on

research that indicated average dispersal distances were ≈7 km (Phillips 2004).

Therefore, we created a buffer of 7 km around 95% of all known locations and calculated

the area of each vegetative zone.

We classified each nesting site location into 1 of the 3 vegetative zones. Nests

located outside of the known vegetative zones were censored. We compared used to

available vegetative zones using chi-square tests combined with Bailey confidence

intervals (Neu et al. 1974, Bailey 1980, Byers et al. 1984, Cherry 1996). We compared

measured nesting site vegetative characteristics in each of the 3 zones using Kruskal-

Wallis H-tests (Zar 1999). Means were also provided for comparison to the literature.

The H statistics were corrected for ties, so the χ² value was reported. Nest success was

calculated by dividing successful nests by total number of nest attempts. We considered

a nest to be successful if ≥1 egg hatched (Badyaev 1995). We calculated nest success of

all nests in each zone and compared them using Bonferroni confidence intervals (Neu

et al. 1974, Byers et al. 1984) after removing nests that were censored due to human-

caused abandonment.

21

2.3.5 Grazing and Nesting

Grazing information was only available at the Matador, Gene Howe, and

Cimarron study areas. Data collected at these sites were used to assess relationships

between turkey nesting and cattle grazing. We defined grazing as presence of cattle in a

pasture after initiation of current year plant growth and before first nest initiation date

(1 April; Figure 2.3). We did not investigate Rio Grande turkey responses to specific

grazing regime systems because each study area used different grazing rotations. We

determined specific pasture use by cattle and timing of cattle rotations by interviewing

landowners and agency personnel (state and federal). We did not examine differences

between early and late nesting site locations because of the error associated with the

cattle locations; they were verified only to ≤1 week around the date given. We assigned

all vegetation plots to a grazing class, either non-grazed (plots that were non-grazed from

1 April to the time of plot measurement) or grazed (plots that were grazed at some period

between 1 April and time of plot measurement). Nests located in pastures with unknown

grazing intensities were censored. We calculated grazed and non-grazed areas to

determine amounts available for use by nesting Rio Grande turkeys each time the cattle

were moved from 1 pasture to another and compared them to areas actually used (Neu

et al. 1974, Bailey 1980, Byers et al. 1984, Cherry 1996). The areas of each pasture type

were cumulatively added because once a pasture was grazed for any length of time after 1

April, it was considered grazed until the beginning of the following year's growing

season. We then compared nesting site vegetative characteristics between grazed and

non-grazed pastures using Mann-Whitney U-tests (Zar 1999). Means were also provided

22

for comparison to the literature. We calculated nest success of all nests in each pasture

type and compared them using Bonferroni confidence intervals (Neu et al. 1974, Byers

et al. 1984) after removing censored nests (due to human-caused abandonment).

We classified nesting sites into 4 categories: upland grazed (UG), upland

non-grazed (UN), riparian grazed (RG), and riparian non-grazed (RN). We used

Mann-Whitney U-tests (Zar 1999) to compare the nesting site vegetative characteristics

in both grazing situations for UP and for RP zones to determine if differences in nesting

site vegetative structure and composition existed between each pair of categories (UG,

UN; RG, RN) and to determine if different structural components were used when cattle

were present (Bareiss et al. 1986, Holechek et al. 1982, Wertz and Flake 1988). We did

not compare all four categories together because we wanted to investigate differences not

accounted for by the vegetative zones. We calculated nest success of all nests in each

category and compared them using Bonferroni confidence intervals (Neu et al. 1974,

Byers et al. 1984) after removing nests that were censored due to human-caused

abandonment.

2.3.6 Available Vegetative Cover

We investigated differences in nesting habitat between vegetative zones and

pasture types by comparing random plots. We compared grazed random and non-grazed

random plots, riparian and upland random plots, riparian grazed and riparian non-grazed

random plots, and upland grazed and upland non-grazed random plots using

Mann-Whitney U-tests (Zar 1999).

23

We also compared availability of nesting habitat between pasture types and

vegetative zones. Each vegetative characteristic (% vertical visual obstruction,

% horizontal visual obstruction, % grass cover, % shrub cover, % bare ground, % forb

cover, and % litter cover) was grouped into classes, adhering to Cochran's rules regarding

χ2 analyses (Cochran 1952). We used χ2 goodness of fit tests (Conover 1999) to

investigate differences in distributions and used Bonferroni intervals to compare

individual classes within each measured vegetative characteristic if the goodness of fit

test was significant (Neu et al. 1974, Byers et al. 1984).

2.3.7 Non-nesting Pasture Selection

We also investigated the use of grazed and non-grazed pastures by females not in

the process of nest incubation during the year and breeding season. We collected

individual turkey locations and pasture use information to assess use of grazed and non-

grazed pastures in relation to their availability at the 3 study areas (Matador, Gene Howe,

Cimarron). We determined specific pasture use by cattle and timing of cattle rotations by

interviewing landowners and agency personnel (state and federal). We used ArcGIS to

clip pastures with a known grazing history that contained turkey locations. Each location

was then assigned to a grazed or non-grazed category. We defined grazing for individual

location use-availability analyses as presence of cattle in a pasture at the same time as

turkey locations were recorded. We removed all known nest related locations from our

locations database to prevent sampling the same information used in the nest habitat

analyses. We also removed all roost locations to eliminate repetitive locations in the

24

same pasture. We used ≤3 day locations per week for each bird, separated by at least 24

hours to maintain sampling independence (White and Garrott 1990). We identified

grazed and non-grazed areas each year to determine amount of each available for use by

female Rio Grande turkeys and compared that to areas used with a chi-square test (Neu

et al. 1974, Byers et al. 1984) and an associated Bailey confidence interval (Bailey 1980,

Cherry 1996). We calculated the areas of grazed and non-grazed pastures by summing

the amount of each, every time the cattle were moved from 1 pasture to another. In this

case, the areas were not cumulative and a pasture was considered grazed only if cattle

were present at the time the location was recorded because we were interested in use of

pastures when cattle were present. We examined use-availability for the entire year and

for the breeding season only.

We used SPSS for Windows (Release 9.0.0, 1998) for all statistical calculations.

This research was approved by the Texas Tech University Animal Care and Use

Committee (Protocol #'s 99917 and 01173B).

2.4 Results

2.4.1 Capture and Telemetry

Mean telemetry error polygon, calculated among all sites and seasons using

known location transmitters (n = 182), fell within 118 m of the true location. The

associated error polygon was 4.40 ha. Due to the large average pasture size (x̄ = 836 ha),

telemetry error was <1% of average pasture size and <1% of vegetative zone sizes.

25

2.4.2 Habitat Selection

We found 395 turkey nests (Table 2.1) and measured vegetative characteristics at

351 of those nesting sites between January 2000 and August 2004. We also measured

345 PNR plots between January 2000 and August 2004. The forward stepwise logistic

regression indicated % horizontal visual obstruction (P ≤ 0.001), % bare ground

(P < 0.003), and % vertical visual obstruction (P ≤ 0.001) were the best variables for

differentiating between nest locations and nearby random locations (Table 2.2). The

resulting complete model correctly predicted 67.8% of plots and the stepwise model

correctly predicted 67.2% of plots. The stepwise model was included because the 3

variables selected predicted the same percentage of plots correctly as when all variable

were entered. The orders of variables selected in the stepwise logistic regression were

% horizontal visual obstruction, % vertical visual obstruction, and % bare ground. The

average nesting site was characterized by 15% vertical visual obstruction, 39 cm of

horizontal visual obstruction, 43% grass cover, 19% shrub cover, 7% bare ground, 17%

forb cover, and 12% litter cover in the nest area (Table 2.2).

A total of 357 nests were classified into a vegetative zone, but only 351 nests had

measured vegetative characteristics. Female Rio Grande turkeys did not use vegetative

zones in proportion to their availability for nesting sites (χ² = 499.41, P < 0.001; Table

2.3). The RP (n = 140 nests) zones were selected for and UACRP (n = 21 nests) and UP

(n = 196 nests) zones were selected against (Table 2.3). The corresponding nest success

was higher (P < 0.05) in UACRP (45.0%) zones than in UP (29.3%) or RP (30.3%) zones

(Table 2.4). RP zone nests had higher (P ≤ 0.001) % vertical visual obstruction and

26

higher (P ≤ 0.001) % grass cover than UACRP and UP (Table 2.5). Nests in UP zones

had a higher (P ≤ 0.001) % shrub cover than the other 2 zones (Table 2.5). Nests in

UACRP had higher (P = 0.009) % bare ground than the other 2 zones and higher % litter

cover (P = 0.051) and % horizontal visual obstruction (P = 0.034) than the RP zone

(Table 2.5). The UACRP zone was removed from further analyses comparing nest site

vegetative characteristics because of the small number of nests (relative to UP and RP) in

UACRP zones.

2.4.3 Grazing and Nesting

A total of 181 nests were classified into a grazing class and had vegetative

measurements recorded. Female Rio Grande turkeys did not use pasture types in

proportion to their availability for nesting sites (χ² = 13.77, P < 0.005; Table 2.6).

Non-grazed pastures (n = 112 nests) were selected for and grazed (n = 69 nests) pastures

were selected against. However, nest success did not differ statistically (P > 0.05)

between grazed (29.9%) and non-grazed (29.2%) pastures (Table 2.7). Measured nesting

site vegetative characteristics also did not differ (P ≥ 0.311) between nests in grazed and

non-grazed pastures (Table 2.8).

We examined nesting site vegetative characteristics within vegetative zones

between grazed and non-grazed pasture types; upland grazed (n = 31 nests) versus upland

non-grazed (n = 56 nests), and riparian grazed (n = 38 nests) versus riparian non-grazed

(n = 56 nests). There were no statistical differences (P ≥ 0.169) between any of the

27

measured vegetative characteristics (Table 2.9, Table 2.10) or between nest success

(P > 0.05) in each category (Table 2.11).

2.4.4 Available Vegetative Cover

Random plots in non-grazed pastures (n = 924) had greater vertical visual

obstruction (P ≤ 0.001), grass cover (P ≤ 0.001), litter cover (P ≤ 0.001) and less bare

ground (P = 0.002), and forb cover (P ≤ 0.001) than random plots in grazed pastures

(n = 654; Table 2.12). Random plots in riparian zones (n = 809) had greater vertical

visual obstruction (P ≤ 0.001), grass cover (P ≤ 0.001), and less horizontal visual

obstruction (P ≤ 0.001), shrub cover (P ≤ 0.001), and forb cover (P ≤ 0.001) than random

plots in upland zones (n = 1622; Table 2.13). Random plots in riparian non-grazed

(n = 358) had greater grass cover (P = 0.009) and less bare ground (P = 0.035) than

random plots in riparian grazed (n = 229; Table 2.14). Random plots in upland

non-grazed (n = 556) had greater vertical visual obstruction (P ≤ 0.001), grass cover

(P = 0.018), litter cover (P ≤ 0.001) and less bare ground (P = 0.043) and forb cover

(P ≤ 0.001) than random plots in upland grazed (n = 422; Table 2.15).

Availability of grazed pasture vegetative characteristics differed (χ² ≥ 5.41,

P < 0.025; Table A.5) from non-grazed pastures. However, individual classes were

usually not different enough to be individually significant. Percentage of grass cover

differed (χ² = 20.42, P < 0.001) between grazed and non-grazed pastures and grazed

pastures consistently had a larger proportion of plots with 0-39% grass cover (Figure

28

2.4). Above 39% grass cover, plots in non-grazed pastures had a larger proportion of

high percentage grass cover (0.569) than grazed pastures (0.479; Figure 2.4). Individual

comparison of classes within overall forb composition indicated grazed pastures had a

larger proportion of plots with 30-39.9% forb cover available than non-grazed pastures

(Figure 2.5).

Vegetative characteristics in upland zones differed (χ² ≥ 27.82, P < 0.001;

Table A.6.) from riparian zone characteristics. Many of the individual comparisons were

also significant, specifically, in regards to grass and shrub composition. Upland zones

consistently had a larger proportion of plots with 0-59% grass cover (Figure 2.6). Above

69%, riparian zones had a larger proportion of plots with successively higher percentages

of grass cover (Figure 2.6). The proportion of plots with ≥10% shrub cover composition

was higher (P < 0.05) in upland zones (Figure 2.7).

2.4.5 Non-nesting Pasture Selection

An analysis of pasture use by non-incubating wild turkey female radio telemetry

locations for the entire year and breeding season compared to pasture availability

indicated that pastures were not used in proportion to their availability (annual

χ² = 336.08, P < 0.001, Table 2.16; breeding season χ² = 129.90, P < 0.001, Table 2.17).

Female Rio Grande wild turkeys avoided pastures with cattle and selected for pastures

without cattle.

29

2.5 Discussion

Percent horizontal visual obstruction was the measured vegetative characteristic

that best differentiated nesting sites from PNR sites. Horizontal visual obstruction was

cited as a principal component of nesting site selection in several previous studies

(Glazener 1958, Beasom 1970, Cook 1972, and Day et al. 1991). The remaining

variables selected by regression analysis were important for differentiating between

nesting and random sites, but were second in importance to the visual obstruction

component. The mean height of horizontal visual obstruction used for nesting sites was

39 cm, which falls within the ranges of other reported horizontal visual obstruction

averages (Cook 1972, Ransom 1987, Day et al. 1991, Wakeling et al. 1998).

Rio Grande wild turkey females selected riparian zones as nesting sites, which

indicated riparian zones play an important role in nesting ecology of Rio Grande wild

turkeys in the Southern Great Plains portion of Rio Grande turkey range. In riparian

zones, grass composition was higher (52.8%) and there was a greater percentage of

vertical visual obstruction (23.1%) providing more vertical cover for nesting turkeys.

Although there was a difference in percentage of vertical visual obstruction between

upland and riparian zone nesting sites, it may have been a result of vegetation differences,

more trees were present in riparian zones providing more vertical visual obstruction. In

the upland zones, there were a greater percentage of shrubs at the nesting site (25.1%).

These trends also held true for random points in both vegetative zones (Table 2.13,

Figure 2.6-2.7), suggesting the birds were not selecting shrubs in upland and grass in

riparian zones, they were simply using the components available after selection of a

30

vegetative zone. Day et al. (1991) found similar results for wild turkeys in South Dakota.

Structure of nest cover was not species specific and birds utilized the best features of

what was available. Although there were compositional differences in nesting cover

between zones, success did not differ, which indicated structure was more important than

specific composition, assuming herbaceous or woody visual obstruction was present.

This is an important issue for land managers of native rangelands. If efforts can be

directed at ensuring visual obstruction via shrubs in upland zones and via herbaceous

vegetation in riparian zones, efforts may be more efficient and beneficial for

establishment and maintenance of wild turkey nesting habitat.

Turkeys selected non-grazed pastures to nest in and avoided grazed pastures.

Previous studies indicated short duration grazing altered structure and composition of the

ground layer (Wilkins and Swank 1992) and affected vegetative screening (Schulz and

Guthery 1987). We expected visual obstruction and nest success to differ between nests

in grazed and non-grazed pastures, but our results did not support that prediction. This

indicated that female turkeys select the same kind of nesting sites regardless of whether

cattle were present. As long as there was herbaceous or shrub cover that provided 35-45

cm of horizontal visual obstruction, grazing activity did not affect selection of nesting

cover. Even when pastures were grazed, some sites remained that provided sufficient

cover for wild turkey nests (Holechek et al. 1982, Bareiss et al. 1986, Wertz and Flake

1988).

Similar patterns of grazed pasture use were also seen in the analysis of individual

telemetry locations of non-nesting birds. This may be a factor of nesting site selection

31

and previously described selection of riparian zones for nesting sites. Although the birds

were not known to establish a nest, the birds may have been searching for nesting habitat

or the nest was depredated before discovery, but there was still selection for riparian

zones consistent with nesting site selection data.

Grazed pastures had less vertical visual obstruction, grass cover, and more bare

ground than non-grazed pastures. Even though vegetation structure at nesting sites did

not differ among pasture types, grass composition was higher (P < 0.001) and shrub

composition was lower (P < 0.001) at nesting sites in riparian zones compared to upland

zones. A higher percentage of horizontal visual obstruction was provided by grass

(52.8%) in riparian zones than any other nesting site component. Under the grazing

intensities during our study, cattle reduced the availability of one nesting site component,

grass cover. Grazed pastures consistently had less grass cover available than non-grazed

pastures (Figure 2.4). Our data indicated a distinct crossover point of grazing effects on

grass availability (Figure 2.4). Non-grazed pastures had a higher proportion of plots with

≥40% grass cover than grazed pastures. Typical riparian zone Rio Grande turkey nesting

sites had 53% grass cover, which was more likely to occur in non-grazed pastures (0.569)

than in grazed pastures (0.479). This effect on grass availability impacted the nesting

characteristics in riparian zones the most due to the high association of riparian zone

nests with grass cover. If desired nesting habitat features are not easily accessible, more

time will have to be spent and distance traveled to find appropriate cover, increasing the

risk of mortality.

32

It is important to understand how these relationships of grazing and turkey nesting

ecology can be used to benefit nesting wild turkeys and maintain grazing practices. Land

managers that want to maintain turkey production, as well as maintain a profitable

livestock enterprise, can balance both with some practical solutions. An appropriate

grazing plan to promote Rio Grande turkey nesting habitat would include grazing upland

zones from 1 April to 31 August, because Rio Grande turkeys use shrub cover for nesting

sites in upland zones, followed by controlled riparian zone grazing after nesting season is

completed (1 September to 31 March). This would provide areas with increased forb

production (Figure 2.5), important for wild turkey brood rearing (Spears 2002), while

maintaining herbaceous cover availability for Rio Grande turkey nesting sites in riparian

zones.

A principle component of the recommended grazing plan is the ability to prevent

cattle from grazing in specific areas at certain times of the year. The ideal way would be

to fence riparian zones as separate pastures and control cattle presence completely.

However, this may not be the most cost effective or feasible solution. There are some

other solutions to alleviate grazing in riparian zones as detailed in "Beef, Brush and

Bobwhites: Quail Management in Cattle Country" (Guthery 1986). Cattle can be

concentrated with anything they like or need. For example, placing water sources and

supplemental feeding stations in strategic locations can concentrate most cattle in a

specific area. This method does not guarantee zero grazing pressure in rested zones, but

it removes most of the pressure and may act as a temporary solution until more

permanent exclosures can be established. Another added benefit from resting riparian

33

zones during the recommended period would be reducing disturbances to other game and

non-game species present, including bobwhite quail, grassland bird species, and various

amphibians.

34

2.6 Literature cited

Ames, C. R. 1978. Wildlife conflicts in riparian management: Grazing. Pages 39-51 in Importance, preservation and management of riparian habitat. USDA Forest Service General Technical Report RM-43.

Ammon, E. M., and P. B. Stacey. 1997. Avian nest success in relation to past grazing

regimes in a montane riparian system. Condor 99: 7-13.

Anderson, E. W., and R. J. Scherzinger. 1975. Improving quality of winter forage for elk by cattle grazing. Journal of Range Management 28: 120-125.

Badyaev, A. V. 1995. Nesting habitat and nesting success of eastern wild turkeys in the Arkansas Ozark Highlands. Condor 97: 221-232.

Bailey, B. J. R. 1980. Large sample simultaneous confidence intervals for the multinomial probabilities based on transformations of the cell frequencies. Technometrics 22: 583-589.

Bailey, W., D. Dennett, H. Gore, J. Pack, R. Simpson, and G. Wright. 1980. Pages 10-23 in Proceedings of the Fourth National Wild Turkey Symposium. National Wild Turkey Federation, March 2-5, 1980, Little Rock, Arkansas, USA.

Baker, D. L., and F. S. Guthery. 1990. Effects of continuous grazing on habitat and density of ground-foraging birds in south Texas. Journal of Range Management 43: 2-5.

Balkenbush, J. A., and D. L. Hallett. 1988. An improved vehicle-mounted telemetry system. Wildlife Society Bulletin 16: 65-67.

Bareiss, L., P. Schulz, and F. Guthery. 1986. Effects of short-duration and continuous grazing on bobwhite and wild turkey nesting. Journal of Range Management 39: 259-260.

Beasom, S. L. 1970. Turkey productivity in two vegetative communities in south Texas. Journal of Wildlife Management 34: 166-175.

Brunjes, J. H. 2005. The population biology and landscape ecology of Rio Grande wild turkeys in the rolling plains of Texas and Kansas. Dissertation, Texas Tech University, Lubbock, Texas, USA.

Bryant, F. C., F. Guthery, and W. Webb. 1981. Grazing management in Texas and its impact on selected wildlife. Pages 94-112 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

35

Byers, C. R., R. K. Steinhorst, and P. R. Krausman. 1984. Clarification of a technique for analysis of utilization-availability data. Journal of Wildlife Management 48: 1050-1053.

Cherry, S. 1996. A comparison of confidence interval methods for habitat use-availability studies. Journal of Wildlife Management 60: 653-658.

Cochran, W. G. 1952. The χ2 test of goodness of fit. Annals of Mathematical Statistics 23: 315-345.

Conover, W. J. 1999. Practical nonparametric statistics. John Wiley and Sons, New York, New York, USA.

Cook, R. L. 1972. A study of nesting turkeys in the Edwards Plateau of Texas. Pages 236-244 in Proceedings of the Southeastern Association of Game and Fish Commissioners, October 22-25, 1972, Knoxville, Tennessee, USA.

Day, K. S., L. D. Flake, and W. L. Tucker. 1991. Characteristics of wild turkey nests in a mixed-grass prairie-oak-woodland mosaic in the northern great plains, South Dakota. Canadian Journal of Zoology 69: 2840-2845.

Dickey, H. P., W. R. Swafford, and Q. L. Markley. 1963. Soil survey of Morton County, Kansas. United States Department of Agriculture, Soil Conservation Service and Forest Service, in cooperation with the Kansas Agricultural Experiment Station.

Dueser, R. D., and H. H. Shugart, Jr. 1978. Microhabitats in a forest-floor small mammal fauna. Ecology 59: 89-98.

Everett, D. D., D. W. Speake, and W. K. Maddox. 1980. Natality and mortality of a north Alabama wild turkey population. Pages 117-126 in Proceedings of the Fourth National Wild Turkey Symposium. National Wild Turkey Federation, March 2-5, 1980, Little Rock, Arkansas, USA.

Glazener, W. C. 1958. Wildlife management as a part of range management in Texas. The 11th annual meeting of the American society range management. 10 p. (mimeo).

Glazener, W. C., A. S. Jackson, and M. L. Cox. 1964. The Texas drop-net turkey trap. Journal of Wildlife Management 28: 280-287.

Guthery, F. S. 1986. Beef, brush and bobwhites: quail management in cattle country. Daesar Kleberg Wildlife Research Institute, Kingsville, Texas, USA.

Holdstock, D. 2004. Personal communication. Texas Parks and Wildlife Department, Canadian, Texas, USA.

36

Holechek, J., R. Valdez, S. Schemnitz, R. Pieper, and C. Davis. 1982. Manipulation of grazing to improve or maintain wildlife habitat. Wildlife Society Bulletin 10: 204-210.

Hosmer, D. W., and S. Lemeshow. 1989. Applied logistic regression. Second edition. John Wiley and Sons, New York, New York, USA.

Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61: 65-71.

Jones, K. H. 1981. Effects of grazing and timber management on Merriam's turkey habitat in mixed conifer vegetation of southcentral New Mexico. Thesis, New Mexico State University, Las Cruces, New Mexico, USA.

Kansas Department of Wildlife and Parks. 2004. 2004 Kansas hunting and furharvesting regulations. Kansas Department of Wildlife and Parks, Topeka, Kansas, USA.

Kauffman, J. B., W. Krueger, and M. Vavra. 1981. Impacts of a late season grazing scheme on nongame wildlife in a Wallowa Mountain riparian ecosystem. Pages 208-220 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

Keegan, T. W. and J. A. Crawford. 1999. Reproduction and survival of Rio Grande turkeys in Oregon. Journal of Wildlife Management 63: 204-210.

Kie, J., C. J. Evans, E. Loft, and J. Menke. 1991. Foraging behavior by mule deer: the influence of cattle grazing. Journal of Wildlife Management 55: 665-674.

Knopf, F. L., and R. W. Cannon. 1982. Structural resilience of a willow community to changes in grazing practices. Pages 198-209 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

Koerth, B. H., W. M. Webb, F. C. Bryant, and F. Guthery. 1983. Cattle trampling of simulated ground nests under short duration and continuous grazing. Journal of Range Management 36: 385-386.

Lemmon, P. E. 1956. A spherical densiometer for estimating forest overstory density. Forest Science 2: 314-320.

Litvaitis, J. A., K. Titus, and E. M. Anderson. 1996. Measuring vertebrate use of terrestrial habitats and foods. Pages 254-274 in T.A. Bookhout, editor. Research and management techniques for wildlife and habitats. Fifth edition. The Wildlife Society, Bethesda, Maryland, USA.

37

Loft, E., J. Menke, J. Kie, and R. Bertram. 1987. Influence of cattle stocking rate on the structural profile of deer hiding cover. Journal of Wildlife Management 51: 655-664.

Loft, E., J. Menke, and J. Kie. 1991. Habitat shifts by mule deer: the influence of cattle grazing. Journal of Wildlife Management 55: 16-26.

McEwen, H. F., F. F. Wheeler, and J. C. Williams. 1973. Soil survey of Collingsworth County, Texas. United States Department of Agriculture, Soil Conservation Service, in cooperation with the Texas Agricultural Experiment Station.

Merrill, L. B. 1975. Effect of grazing management practices on wild turkey habitat. Pages 108-112 in Proceedings of the Third National Wild Turkey Symposium. Texas Chapter of the Wildlife Society, February 13, 1975, Austin, Texas, USA.

Miller, D. A., L. W. Burger, B. D. Leopold, and G. A. Hurst. 1998. Survival and cause specific mortality of wild turkey hens in central Mississippi. Journal of Wildlife Management 62: 306-313.

Miller, G., and J. Ray. 1996. Managing riparian habitat for wildlife in northwest Texas. Texas Parks and Wildlife Department, Wildlife Division. 2 pages.

Miller, M. S., D. J. Buford, and R. S. Lutz. 1995. Survival of female Rio Grande turkeys during the reproductive season. Journal of Wildlife Management 59: 766-771.

Neu, C. W., R. C. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-availability data. Journal of Wildlife Management 38: 541-545.

Nielson, L., and C. Yde. 1981. The effects of rest-rotation grazing on the distribution of sharp-tailed grouse. Pages 147-165 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

Pelham, P. H., and J. G. Dickson. 1992. Physical Characteristics. Pages 32-45 in J.G. Dickson, editor. The wild turkey: biology and management. Stackpole Books, Mechanicsburg, Pennsylvania, USA.

Phillips, R. S. 2004. Movements, survival, and reproduction of Rio Grande wild turkeys in the Texas panhandle. Thesis, Texas Tech University, Lubbock, Texas, USA.

Pinchak, W. E., M. A. Smith, R. H. Hart, and J. W. Waggoner Jr. 1991. Beef cattle grazing distribution patterns on foothill range. Journal of Range Management 44:267-275.

38

Prasad, N. L. N. S., and F. Guthery. 1986. Wildlife use of livestock water under short duration and continuous grazing. Wildlife Society Bulletin 14: 450-454.

Ransom, D., O. J. Rongstad, and D. H. Rusch. 1987. Nesting ecology of Rio Grande turkeys. Journal of Wildlife Management 51: 435-439.

Rees, J. 1981. Potential uses of cattle grazing to manage waterfowl nesting cover on Turnbull National Wildlife Refuge. Pages 86-93 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

Richardson, W. E., J. Hajek, E. Blakley, and C. Neitsch. 1974. Soil survey of Cottle County, Texas. United States Department of Agriculture, Soil Conservation Service, in cooperation with the Texas Agricultural Experiment Station.

Roath, L. R., and W. C. Krueger. 1982. Cattle grazing influence in a mountain riparian zone. Journal of Range Management 35: 100-104.

Robel, R., J. Briggs, A. Dayton, and L. Hulbert. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management 23: 295-297.

Rumble, M. A., and R. A. Hodroff. 1993. Nesting ecology of Merriam's turkey in the Black Hills, South Dakota. Journal of Wildlife Management 57: 789-801.

Ruthven, D. 2004. Personal communication. Texas Parks and Wildlife Department, Paducah, Texas, USA.

Samuel, M. D., and M. R. Fuller. 1996. Wildlife Radiotelemetry. Pages 370-418 in T.A. Bookhout, editor. Research and management techniques for wildlife and habitats. Fifth edition. The Wildlife Society, Bethesda, Maryland, USA.

Schmutz, J. A., and C. E. Braun. 1989. Reproductive performance of Rio Grande wild turkeys. Condor 91: 675-680.

Schmutz, J. A., C. E. Braun, and W. F. Andelt. 1989. Nest habitat use of Rio Grande wild turkeys. Wilson Bulletin 101: 591-598.

Schulz, P., and F. Guthery. 1987. Effects of short duration grazing on wild turkey home ranges. Wildlife Society Bulletin 15: 239-241.

Schulz, T. T., and W. C. Leininger. 1990. Differences in riparian vegetation structure between grazed areas and exclosures. Journal of Range Management 43: 295-299.

39

Sedgwick, J. A., and F. L. Knopf. 1991. Prescribed grazing as a secondary impact in a western riparian flood plain. Journal of Range Management 44: 369-373.

Skovlin, J. M. 1984. Impacts of grazing on wetlands and riparian habitat: a review of our knowledge. Pages 1001-1103 in B. Delworth Gardner, committee chairman. Developing strategies for rangeland management. National Research Council/National Academy of Science, Westview Press, Boulder, Colorado, USA.

Smith M. A., J. D. Rodgers, J. L. Dodd, and Q. D. Skinner. 1992. Habitat selection by cattle along an ephemeral channel. Journal of Range Management 45: 385-390.

Spears, B. L. 2002. Wild turkey pre-flight chick habitat characteristics and survival. Thesis, Texas Tech University, Lubbock, Texas, USA.

Stover, S. L. 1985. Silviculture and grazing in the New Forest: rival land uses over nine centuries. Journal of Forest History 29: 32-42.

Szaro, R. C., and C. P. Pase. 1983. Short-term changes in a cottonwood-ash-willow association on a grazed and an ungrazed portion of Little Ash Creek in central Arizona. Journal of Range Management 36: 382-384.

Texas Parks and Wildlife Department. 2004. 2003-2004 Outdoor Annual. Texas Parks and Wildlife Department, Austin, Texas, USA.

Wakeling, B. F., S. S. Rosenstock, and H. G. Shaw. 1998. Forest stand characteristics of successful and unsuccessful Merriam's turkey nest sites in north-central Arizona. Southwestern Naturalist 43: 242-248.

Walker, E. A. 1948. Rio Grande turkey in the Edwards Plateau of Texas, 1946-1948. Texas Game, Fish, and Oyster Commission, Final Report Federal Aid Project 22-R.

Wertz, T. L., and L.D. Flake. 1988. Wild turkey nesting ecology in south central South Dakota. Prairie Naturalist 20: 29-37.

Westerkov, K. 1950. Methods for determining the age of game bird eggs. Journal of Wildlife Management 14: 56-67.

White, G. C., and R.A. Garrott. 1990. Analysis of wildlife radio-tracking data. Academic Press Inc., San Diego, California, USA.

Wilkins, R. N. and W.G Swank. 1992. Bobwhite habitat use under short duration and deferred-rotation grazing. Journal of Range Management 45: 549-553.

40

Williams, J. C., A. J. Welker, F. F. Wheeler, and H. F. McEwen. 1974. Soil survey of Hemphill County, Texas. United States Department of Agriculture, Soil Conservation Service, in cooperation with the Texas Agricultural Experiment Station.

Zar, J. H. 1999. Biostatistical analysis. Prentice-Hall, Upper Saddle River, New Jersey, USA.

Zwickel, F. C. 1972. Some effects of grazing on blue grouse during summer. Journal of Wildlife Management 36: 631-634.

41

42

Table 2.1. Total number of Rio Grande wild turkey nests documented from January 2000 through August 2004 on 4 study areas in the Southern Great Plains.

Year Study Site 2000 2001 2002 2003 2004 Matador 23 24 20 22 12 Salt Fork 9 24 36 13 15 Gene Howe 23 34 19 13 22 Cimarron 16 32 0a 17 21

aNo data collected at Cimarron for 2002

Table 2.2. Comparison of habitat characteristics measured at Rio Grande wild turkey nesting (n = 351) and paired nest random sites (n = 345) from January 2000 through August 2004 at 4 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE)a Logistic regression (LR) Forward stepwise LR Variable Nesting 50 m Random Estimate SE P Estimate SE P Vertical obstruction 14.8 (1.5) 9.3 (1.1) 0.012 0.004 ≤ 0.002 0.013 0.004 ≤ 0.001 Horizontal obstruction 38.8 (1.0) 26.5 (0.7) 0.510 0.073 ≤ 0.001 0.515 0.062 ≤ 0.001 Grass cover 42.7 (1.4) 45.5 (1.3) -0.279 0.202 0.168 N/Ab N/A N/A

43

Shrub cover 19.1 (1.1) 12.4 (0.8) -0.265 0.207 0.200 N/A N/A N/A Bare ground 7.4 (0.5) 12.6 (0.7) -0.545 0.225 0.015 -0.241 0.081 ≤ 0.003 Forb cover 16.8 (1.0) 17.0 (0.8) -0.314 0.206 0.126 N/A N/A N/A Litter cover 12.5 (0.7) 11.7 (0.7) -0.205 0.210 0.331 N/A N/A N/A a Means are provided for comparison to other literature. b N/A indicates that the variable was not included in the forward stepwise logistic regression equation.

Table 2.3. Selection of vegetative zones for nesting sites by female Rio Grande wild turkeys from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 2: χ² = 5.99; calculated χ² was 499.41. Vegetative Relative Expected Nests Observed zone Area (Ha) area nests observed proportion Bailey intervald UACRPa 73,770 0.251 89.52 21 0.059 0.042 ≤ P ≤ 0.080 - UPb 197,051 0.670 239.11 196 0.549 0.511 ≤ P ≤ 0.587 - RPc 23,381 0.079 28.37 140 0.392 0.356 ≤ P ≤ 0.430 + a urban areas/agriculture/Conservation Reserve Program lands 44

b upland zones/non-riparian brushland and trees c riparian zones d + indicates selection for, - indicates selection against

Table 2.4. Nest success associated with vegetative zones used by female Rio Grande wild turkeys from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). Vegetative # Nest # Nests # Successful # Unsuccessful Nest Bonferroni zone attempts censored nests nests successa interval UACRPb 21 1 9 11 45.0% A 38.6 ≤ P ≤ 51.4 UPc 196 5 56 135 29.3% B 23.4 ≤ P ≤ 35.2 RPd 140 8 40 92 30.3% B 24.4 ≤ P ≤ 36.2 a nest success followed by the same letter are not different at α = .05 b urban areas/agriculture/Conservation Reserve Program lands 45

c upland zones/non-riparian brushland and trees d riparian zones

Table 2.5. Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in each vegetative zone from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas).

Mean (SE) Mean Ranka Kruskal-Wallis Variable UACRPb (n = 21) UPc (n = 192) RPd (n = 138) UACRP UP RP χ² P Vertical obstruction 4.3 (2.3) 10.4 (1.8) 23.1 (2.7) 139.61A 156.36A 205.48B 29.07 ≤ 0.001 Horizontal obstruction 44.7 (5.0) 40.0 (1.3) 36.3 (1.5) 200.27A 183.56AB 157.68B 6.79 0.034 Grass cover 35.5 (5.2) 36.4 (1.8) 52.8 (2.2) 145.89A 150.27A 212.97B 32.70 ≤ 0.001 Shrub cover 11.0 (4.9) 25.1 (1.5) 12.0 (1.4) 103.32A 209.88B 136.59A 54.39 ≤ 0.001

46

Bare ground 16.0 (4.3) 7.5 (0.6) 5.9 (0.7) 217.48A 181.98B 157.10BC 9.32 0.009 Forb cover 16.7 (4.0) 16.0 (1.2) 17.7 (1.8) 172.59A 174.12A 175.33A 0.02 0.990 Litter cover 18.4 (3.8) 13.5 (1.0) 10.2 (0.9) 201.75A 182.31AB 159.18B 5.94 0.051 a mean ranks within rows, followed by the same letter, are not different at α = .05. b urban areas/agriculture/Conservation Reserve Program lands c upland zones/non-riparian brushland and trees d riparian zones

Table 2.6. Selection of pasture for nesting sites by female Rio Grande wild turkeys from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 13.77. Relative Expected Nests Observed Pasture Area (Ha) area nests observed proportion Bailey interval Grazed 391,713 0.519 93.94 69 0.381 0.344 ≤ P ≤ 0.421 - a

Non-grazed 363028 0.481 87.06 112 0.619 0.581 ≤ P ≤ 0.658 + a + indicates selection for, - indicates selection against 47

Table 2.7. Nest success associated with pasture types used by female Rio Grande wild turkeys from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). # Nest # Nests # Successful # Unsuccessful Nest Bonferroni Pasture attempts censored nests nests successa interval Grazed 69 2 20 47 29.9% A 22.1 ≤ P ≤ 37.6 Non-grazed 112 6 31 75 29.2% A 21.5 ≤ P ≤ 37.0 a nest success followed by the same letter are not different at α = .05 48

Table 2.8. Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in each pasture type from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable Grazed (n = 112) Non-grazed (n = 69) Grazed Non-grazed U P Vertical obstruction 22.2 (3.8) 23.4 (3.1) 89.88 91.69 3,786.5 0.811 Horizontal obstruction 35.5 (2.0) 35.8 (1.5) 89.36 92.01 3,751.0 0.741 Grass cover 51.2 (3.0) 47.7 (2.3) 96.03 87.90 3,517.0 0.311 Shrub cover 14.6 (2.1) 15.3 (1.7) 88.51 92.53 3,692.5 0.615 49 Bare ground 6.4 (0.9) 6.9 (0.7) 87.91 92.90 3,651.0 0.529 Forb cover 17.8 (2.4) 17.9 (1.9) 90.83 91.10 3,852.5 0.973 Litter cover 9.6 (1.3) 10.4 (1.0) 88.09 92.79 3,663.0 0.556

Table 2.9. Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in upland grazed and upland non-grazed categories from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable UGa (n = 31) UNb (n = 56) UG UN U P Vertical obstruction 13.5 (4.7) 19.6 (4.2) 41.97 44.32 794.0 0.646 Horizontal obstruction 39.2 (3.1) 33.8 (1.7) 48.40 40.88 693.0 0.183 Grass cover 46.4 (4.6) 42.5 (2.9) 46.25 42.03 757.5 0.455 50

Shrub cover 20.1 (3.4) 20.3 (2.6) 44.40 43.02 813.0 0.807 Bare ground 6.5 (1.1) 8.5 (1.1) 39.97 45.39 734.0 0.334 Forb cover 17.6 (2.9) 15.5 (2.2) 46.72 41.78 743.5 0.381 Litter cover 8.7 (2.0) 10.6 (1.4) 38.62 46.12 693.5 0.183 a upland grazed b upland non-grazed

Table 2.10. Comparison of habitat characteristics measured at Rio Grande wild turkey nesting sites in riparian grazed and riparian non-grazed categories from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable RGa (n = 38) RNb (n = 56) RG RN U P Vertical obstruction 29.0 (5.4) 27.4 (4.5) 47.78 47.30 1,061.5 0.931 Horizontal obstruction 33.1 (2.6) 37.5 (2.4) 42.91 50.75 893.5 0.169 Grass cover 55.2 (4.0) 53.6 (3.6) 48.68 46.66 1,026.5 0.724 51 Shrub cover 10.0 (2.3) 9.8 (1.8) 46.47 48.23 1,032.5 0.754 Bare ground 6.2 (1.4) 5.1 (0.8) 48.01 47.14 1,052.5 0.875 Forb cover 17.7 (3.6) 20.2 (3.0) 44.55 49.59 957.5 0.377 Litter cover 10.5 (1.7) 10.0 (1.5) 48.87 46.53 1,019.0 0.680 a riparian grazed b riparian non-grazed

Table 2.11. Nest success associated with the 4 categories used by female Rio Grande wild turkeys from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). # Nest # Nests # Successful # Unsuccessful Nest Bonferroni Pasture attempts censored nests nests successa interval Upland grazed 31 0 10 21 32.3% A 23.4 ≤ P ≤ 41.1 Upland non-grazed 56 1 15 40 27.3% A 18.8 ≤ P ≤ 35.7 Riparian grazed 38 2 10 26 27.8% A 19.3 ≤ P ≤ 36.3 52

Riparian non-grazed 56 5 16 35 31.4% A 22.6 ≤ P ≤ 40.2 a nest success followed by the same letter are not different at α = .05

Table 2.12. Comparison of habitat characteristics measured at grazed random plots and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable Grazed (n = 654) Non-grazed (n = 924) Grazed Non-grazed U P Vertical obstruction 8.5 (0.8) 11.9 (0.8) 736.56 826.97 267,528.0 ≤ 0.001 Horizontal obstruction 22.2 (0.5) 22.5 (0.4) 766.14 806.03 286,873.5 0.086 Grass cover 40.4 (0.9) 44.6 (0.8) 740.76 823.99 270,275.0 ≤ 0.001 Shrub cover 10.2 (0.5) 9.7 (0.4) 794.06 786.27 299,168.0 0.736 53

Bare ground 16.8 (0.6) 14.9 (0.5) 831.96 759.45 274,382.0 0.002 Forb cover 20.7 (0.7) 15.7 (0.5) 872.75 730.58 247,701.5 ≤ 0.001 Litter cover 10.7 (0.5) 13.1 (0.5) 738.24 825.78 268,626.5 ≤ 0.001

Table 2.13. Comparison of habitat characteristics measured at riparian random plots and upland random plots from January 2000 through August 2004 on 4 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable Riparian (n = 809) Upland (n = 1622) Riparian Upland U P Vertical obstruction 13.5 (0.9) 6.3 (0.5) 1,402.25 1,123.11 505,425.0 ≤ 0.001 Horizontal obstruction 21.3 (0.5) 24.6 (0.4) 1,082.05 1,282.81 547,736.5 ≤ 0.001 Grass cover 46.2 (0.9) 40.7 (0.6) 1,317.52 1,165.36 573,966.5 ≤ 0.001 Shrub cover 6.6 (0.4) 11.6 (0.3) 985.20 1,331.11 469,385.0 ≤ 0.001 54

Bare ground 16.2 (0.6) 15.2 (0.3) 1,189.45 1,229.24 634,623.5 0.187 Forb cover 16.2 (0.6) 18.4 (0.4) 1,136.18 1,255.81 591,524.0 ≤ 0.001 Litter cover 13.7 (0.6) 12.7 (0.3) 1,185.73 1,231.10 631,610.0 0.132

Table 2.14. Comparison of habitat characteristics measured at riparian grazed random plots and riparian non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable RGa (n = 229) RNb (n = 358) RG RN U P Vertical obstruction 12.7 (1.5) 14.2 (1.3) 289.38 296.95 39,934.0 0.568 Horizontal obstruction 21.8 (1.0) 22.4 (0.6) 278.68 303.80 37,483.0 0.079 Grass cover 44.0 (1.7) 49.5 (1.3) 271.27 308.54 35,785.0 0.009 55 Shrub cover 9.1 (0.9) 6.9 (0.5) 300.00 290.16 39,616.5 0.480 Bare ground 16.3 (1.1) 13.5 (0.8) 312.40 282.23 36,778.5 0.035 Forb cover 17.8 (1.2) 15.8 (0.9) 307.86 285.14 37,818.0 0.113 Litter cover 12.3 (0.9) 13.2 (0.8) 291.92 295.33 40,514.0 0.811 a riparian grazed b riparian non-grazed

Table 2.15. Comparison of habitat characteristics measured at upland grazed random plots and upland non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable UGa (n = 422) UNb (n = 556) UG UN U P Vertical obstruction 6.3 (0.9) 10.6 (0.9) 444.97 523.30 98,523.0 ≤ 0.001 Horizontal obstruction 22.4 (0.6) 22.7 (0.5) 479.00 497.47 112,884.0 0.310 Grass cover 38.6 (1.1) 41.8 (0.9) 464.95 508.13 106,957.0 0.018 56 Shrub cover 10.9 (0.6) 11.7 (0.5) 479.93 496.76 113,278.0 0.354 Bare ground 16.9 (0.7) 15.6 (0.6) 510.41 473.63 108,493.5 0.043 Forb cover 22.3 (0.9) 15.8 (0.7) 553.27 441.10 90,406.5 ≤ 0.001 Litter cover 9.9 (0.6) 12.9 (0.6) 443.83 524.16 98,045.5 ≤ 0.001 a upland grazed b upland non-grazed

Table 2.16. Selection of pasture types by non-nesting female Rio Grande wild turkeys using all locations from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 336.08. Relative Expected Locations Observed Pasture Area (Ha) area locations observed proportion Bailey interval Grazed 366,009 0.349 2,790.30 2,009 0.251 0.220 ≤ P ≤ 0.284 - a

Non-grazed 682,578 0.651 5,203.70 5,985 0.749 0.715 ≤ P ≤ 0.779 + a + indicates selection for, - indicates selection against

57

Table 2.17. Selection of pasture types by non-nesting female Rio Grande wild turkeys during the breeding season (1 April- 31 August) from April 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 129.90. Relative Expected Locations Observed Pasture Area (Ha) area locations observed proportion Bailey interval Grazed 254,197 0.337 1,784.03 1,392 0.263 0.231 ≤ P ≤ 0.296 - a

Non-grazed 500,545 0.663 3,512.97 3,905 0.737 0.703 ≤ P ≤ 0.768 + a + indicates selection for, - indicates selection against

58

S

SFo

Cimarron: Cimarron National Grasslands, Elkhart, KGene Howe: Gene Howe Wildlife Management Area, CSalt Fork: Salt Fork of the Red River, Clarendon, TexMatador: Matador Wildlife Management Area, Padu Figure 2.1. Location of 4 study areas where relationshipsfemale Rio Grande wild turkey habitat use and nesting ecsouthern Great Plains from January 2000-August 2004.

59

Cimarron

Gene Howe

alt rk

r

Matado

ansas anadian, Texas as cah, Texas

between cattle graology were studied

K

CO

NM

X

T

zi i

OK

ng and n the

Figure 2.2. Description of vegetation plot used to measure data regarding female Rio Grande wild turkey habitat use and availability in the southern Great Plains from January 2000-August 2004.

60

Dates of nest initiation

0

2

4

6

8

10

12

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95

Calendar days (1 April-3 July)

Cou

nt o

f Nes

ts

Figure 2.3. Bar graph of nest initiation dates for female Rio Grande wild turkeys from 4 study areas in the southern Great Plains. The years (2000-2004) and study areas (Matador, Salt Fork, Gene Howe, Cimarron) are pooled. The first number corresponds to the first known nest initiation day (1 April) and the last number corresponds to the last known nest initiation day (3 July).

61

Grass Cover

0.000

0.050

0.100

0.150

0.200

0.250

0-9.9

10-19

.9

20-29

.9

30-39

.9

40-49

.9

50-59

.9

60-69

.9

70-79

.9

80-89

.9

90-10

0

Class (% grass cover)

Prop

ortio

n of

Plo

tss

GrazedNon-Grazed

Figure 2.4. Comparison of plot proportions with each respective class of grass cover percentages between grazed and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant.

62

Forb Cover

0.000

0.050

0.100

0.150

0.200

0.250

0.300

0.350

0.400

0.450

0.500

0-9.9

10-19

.9

20-29

.9

30-39

.9

40-49

.9

50-59

.9

60-69

.9

70-79

.9 >80

Class (% forb cover)

Prop

ortio

n of

Plo

tss

GrazedNon-Grazed

Figure 2.5. Comparison of plot proportions with each respective class of forb cover percentages between grazed and non-grazed random plots on 3 study areas in the Southern Great Plains from January 2000 through August 2004 (pooled over years and study areas). Bonferroni intervals that overlap are not significant.

63

Grass Cover

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0-9.9

10-19

.9

20-29

.9

30-39

.9

40-49

.9

50-59

.9

60-69

.9

70-79

.9

80-89

.9

90-10

0

Class (% grass cover)

Prop

ortio

n of

Plo

tss

UplandRiparian

Figure 2.6. Comparison of plot proportions with each respective class of grass cover percentages between upland and riparian random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant.

64

Shrub Cover

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0-9.910-19.9

20-29.930-39.9

40-49.950-59.9

60-69.970-79.9

Class (% shrub cover)

Prop

ortio

n of

Plo

tss

UplandRiparian

Figure 2.7. Comparison of plot proportions with each respective class of shrub cover percentages between upland and riparian random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant.

65

CHAPTER III

RELATIONSHIPS BETWEEN CATTLE GRAZING AND

MALE RIO GRANDE WILD TURKEY HOME

RANGE SIZE AND PASTURE USE IN

THE SOUTHERN GREAT PLAINS

3.1 Introduction

Evidence of competition between 2 species can be seen if niche shifts occur

following introduction or removal of 1 of the species (Diamond 1978). Niche areas most

likely segregated to reduce competition are habitat, food, and time of activity (Schoener

1974). Home range composition is influenced by many factors including competition,

and quality of food, water, and cover. Even if diet overlap is low, competition may still

be an observable factor in other aspects of interspecific relationships (Rosenzweig 1981).

Interactions between cattle and gallinaceous birds have been extensively studied

throughout the United States. Blue grouse (Dendragapus obscurus pallidus) density

indices were not different in grazed or non-grazed areas (Zwickel 1972) and even when a

pasture was overgrazed, grouse did not move from their traditional use areas into adjacent

non-grazed pastures or areas of taller grass (Nielson and Yde 1981). Several studies

described either positive or neutral effects of grazing on specific aspects of bobwhite

quail (Colinus virginianus) ecology. Density responses of bobwhites to grazing pressure

were variable (Baker and Guthery 1990) and short duration grazing was predicted to

improve bobwhite habitat, at least for the short-term (Wilkins and Swank 1992).

66

Bobwhite quail response to grazing on Texas rangeland varied depending on rainfall

during continuous grazing systems (Bryant et al. 1981).

Wild turkey (Meleagris gallopavo) research has, for the most part, centered on

turkey population dynamics and nesting habitat characteristics (Everett et al. 1980,

Bidwell et al. 1989, Miller et al. 1999). Few published manuscripts report effects of

varying livestock rotation methods and stocking densities on wild turkey ecology.

Merriam's turkeys (M.g. merriami) in New Mexico showed no preference for grazed or

non-grazed rest rotation pastures (Jones 1981). Walker (1948) indicated that 72% of Rio

Grande turkeys (M.g. intermedia) on the Edwards Plateau in central Texas were

concentrated on lightly grazed or deferred pastures, which was only 27% of the study

area. Rio Grande turkeys avoided stock tanks at the hub of short duration grazing

systems (Prasad and Guthery 1986) but short duration cattle grazing did not affect turkey

home range size, relative to continuous grazing in south Texas (Schulz and Guthery

1987).

Food availability is a major determinant of wild turkey home range size and

habitat use (Mosby and Handley 1943, Wheeler 1948, Miller et al. 1985, Kurzejeski and

Lewis 1990). When food was abundant, seasonal movements were reduced (Korschgen

1967, Ligon 1946). In south-central Texas, supplemental feeding tended to concentrate

daily Rio Grande turkey movements (Thomas et al. 1966). Grazing at light to moderate

stocking rates (35-55% removal) had little to no effect on important seed producing

plants in the southeastern U.S. (Lewis and Harshbarger 1986), but in the southwestern

U.S., the effects of grazing were much more detrimental to individual plants and habitat,

67

especially in riparian zones (Davis 1981). Schulz and Guthery (1987) suggested that

short duration grazing did not affect food supplies and therefore did not affect Rio

Grande turkey home ranges in southern Texas.

Our objectives were to calculate home range sizes and compare them to cattle

presence to assess possible space use relationships. Specifically, how did home range

sizes change when cattle were removed completely. We also compared the usefulness of

home range calculations and use-availability analyses (Neu et al. 1974, Bailey 1980,

Byers et al. 1984, Cherry 1996) to quantify male Rio Grande turkey space use in the

presence and absence of cattle. We predicted male Rio Grande wild turkeys would

exhibit reduced home range sizes when cattle were removed from the study site. We also

hypothesized male turkeys would use grazed pastures in higher proportion than their

availability due to the increased availability of displaying areas as a result of cattle

grazing (Holdstock 2003).

3.2 Study Areas

Turkey movements were used to determine study area boundaries. Land use at

each of the study areas included production of cattle, cotton, wheat, and grain sorghum.

Grazing occurred at various levels on upland and riparian sites and cattle densities ranged

from 1.8 to 31 ha/animal unit (AU) (Table A.1-A.4 and Figure A.1-A.3). All cattle

operators provided supplemental feeding with hay bales and/or cottonseed cake pellets

Data were collected on male Rio Grande turkey home range and pasture use from January

2000 through August 2004.

68

Turkeys were captured on 4 study areas: 3 in the Texas Panhandle, and 1 in the

southwestern corner of Kansas (Figure 3.1). The Salt Fork study area was located in

Collingsworth and Donley Counties in Texas, and was bisected by the Salt Fork River. It

was located near the Caprock escarpment below the edge of the High Plains and was

centered on private ranches with a total of 17,000 ha. The Salt Fork study area was

characterized by a dry-steppe climate with mild winters. Average annual precipitation

was 54.64 cm, most of which occurred from April through October. The growing season

averaged 206 days. Common soil associations included Mobeetie-Veal-Potter (deep to

very shallow, gently sloping to steep, loamy soils; on uplands), Obaro-Aspermont-

Quinlan (deep to shallow, gently sloping to steep, loamy soils; on uplands), and Springer-

Lincoln-Likes (Deep, nearly level to sloping, sandy soils; on uplands and bottomlands).

Soil series included Quinlan, Springer, Miles, Enterprise, Wichita (fine, mixed, thermic;

alfisols), Lutie (fine-silty, mixed, thermic; mollisols), Ector (loamy-skeletal, carbonatic,

thremic; mollisols), and LaCasa (fine, mixed, thermic; mollisols) (McEwen et al. 1973).

Primary woody vegetation included eastern cottonwood (Populus deltoides), honey locust

(Gleditsia triacanthos), black locust (Robinia pseudo-acacia), salt cedar (Tamarix

gallica), sand sagebrush (Artemisia filifolia), and Chickasaw plum (Prunus angustifolia).

Grass species included sideoats grama (Bouteloua curtipendula), hairy grama (Bouteloua

hirsuta), western wheatgrass (Elytrigia smithii), and little bluestem (Schizachyrium

scoparium). Forb species included western ragweed (Ambrosia psilostachya), annual

sunflower (Helianthus annus), and broom snakeweed (Gutierrezia sarothrae). No

69

grazing information was collected at the Salt Fork study area, so all grazing analyses

excluded Salt Fork data.

The Gene Howe study area was located in Hemphill County, near Canadian,

Texas, and was bisected by the Canadian River. It was located in the Rolling Plains and

was centered on 2,180 ha of public land (Gene Howe Wildlife Management Area,

GHWMA) with access to 11,000 ha of adjacent private lands. Hemphill County was

characterized by a dry-steppe climate with 52.07 cm of average annual precipitation,

most frequently in the form of thunderstorms. Periods of 2-3 weeks with no rain were

fairly common and monthly periods of no rainfall were recorded. The average annual

growing season was 204 days. Soil associations included Tivoli-Springer (deep, sandy

soils on upland dunes and hummocks), Mobeetie-Berda-Potter (deep and very shallow,

gently sloping to steep, loamy soils on uplands), and Lincoln-Sweetwater (sandy and

loamy soils on bottomlands). Soil series included Linclon, Enterprise, Springer,

Sweetwater (fine-loamy over sandy or sandy-skeletal, mixed calcareous, thermic;

mollisols), Likes (mixed, thermic; entisols), Tivoli (mixed, thermic; entisols), Tipton

(fine-loamy, mixed, thermic; mollisols), and Bippus (fine-loamy, mixed, thermic;

mollisols) (Williams et al. 1974). Dominant woody vegetation included eastern

cottonwood, Russian olive (Eleagnus angustifolia), salt cedar, Chickasaw plum and sand

sagebrush. Grass species present on the area included little bluestem, sideoats grama,

Canada wildrye (Elymus canadensis), sand dropseed (Sporobolus cryptandrus), purple

threeawn (Aristida purpurea), and western wheatgrass. Primary forb species included

western ragweed, annual sunflower, and silverleaf nightshade (Solanum eleagnifolium).

70

The GHWMA had a rest-rotation, cow/calf operation grazing lease from September 2001

through September 2004. Cattle stocking densities ranged from 4.39-11.52 ha/AU.

Pastures were grazed for 6 month periods, then rested for 6 months before grazed again.

No cattle information was collected from adjacent lands.

The Cimarron study area was in the southwestern corner of Kansas near Elkhart,

Kansas in Morton County on 29,648 ha of public (Cimarron National Grasslands, CNG)

and 15,000 ha of adjacent private land. The Cimarron River bisected the study area.

Climate of the Cimarron study area was semiarid with haphazard precipitation events.

Average annual precipitation was 42.60 cm, which mostly occurred in July. Drought

periods often extended past 30 days. Average length of the growing season was 214

days. Primary soil associations included Richfield-Ulysses (loamy soils of uplands),

Vona-Tivoli (rolling sandy land), and Otero-Lincoln (soils of the Cimarron River Valley

and adjacent slopes). Soil series included Tivoli, Lincoln, Dalhart (deep, dark, nearly

level and gently sloping sandy soils of the upland; regosols), Richfield (deep, dark, nearly

level upland soils; regosols), Vona (sandy soils of upland), Bridgeport (gently sloping,

north wall of Cimarron River; alluvial), Mansker (loamy soils of upland; calcisols), Potter

(shallow, overlie caliche or limestone; lithosols) (Dickey et al. 1963). Primary woody

species were eastern cottonwood and salt cedar in the riparian corridor, and sand

sagebrush on the uplands adjacent to the riparian corridor. Grasses found on the study

area included sand bluestem (Andropogon hallii), blue grama (Bouteloua gracilis),

sideoats grama, sand dropseed, sand lovegrass (Eragrostis trichodes), prairie sandreed

(Calamovilfa longifolia), and buffalo grass (Buchloe dactyloides) (Spears 2002).

71

Common forbs included broom snakeweed, western ragweed, and daisy fleabane

(Erigeron annus). The CNG had a rest-rotation, cow/calf operation grazing lease from

January 2000-September 2004. Cattle stocking densities ranged from 2.52-10.1 ha/AU.

Pastures were grazed for 1-2 months per rotation period and then rested for up to 1 year.

No cattle information was collected from adjacent lands.

The southernmost site was the Matador study area in Cottle County, Texas. It

was located in the lower Rolling Plains and consisted of 11,370 ha of public land

(Matador Wildlife Management Area, MWMA) with an additional 16,133 ha of adjacent

private lands. Dry winters and hot summers characterized the climate of Cottle County.

Rainfall events usually occurred as thunderstorms and average precipitation was 56.24

cm. However, monthly and annual amounts were highly variable. Maximum

precipitation usually occurred in May and June and periods of drought were very

common. The average growing season length was 219 days. The dominant soil

association was Miles Springer with nearly level to strongly sloping, deep, coarse

textured and moderately coarse textured soils on outwash plains. Secondary associations

were Nobscot-Heatly with nearly level to gently sloping, deep, coarse-textured soils on

upland plains, and Woodward-Carey, with nearly level to gently sloping, deep to

moderately deep, medium-textured soils on upland plains. Soil series included Quinlan

(loamy, mixed, thermic, shallow; inceptisols), Woodward (coarse-silty, mixed, thermic;

inceptisols), Hilgrave (loamy-skeletal, mixed, thermic; alfisols), Yahola (coarse-loamy,

mixed calcareous, thermic; entisols), Enterprise (coarse-silty, mixed, thermic;

inceptisols), Lincoln (sandy, mixed, thermic; entisols), Miles (fine-loamy, mixed,

72

thermic; alfisols), and Springer (coarse-loamy, mixed, thermic; alfisols) (Richardson et

al. 1974). Primary woody vegetation included honey mesquite (Prosopis glandulosa),

redberry juniper (Juniperus pinchotii), netleaf hackberry (Celtis reticulata), eastern

cottonwood, salt cedar, Chickasaw plum, and sand sagebrush. Grasses found on the area

included little bluestem, sand dropseed, sideoats grama, purple threeawn, Japanese brome

(Bromus japonicus) and plains bristlegrass (Setaria leucopila). Primary forb species

included western ragweed, woolly plantain (Plantago insularis), annual sunflower, and

lamb's quarters (Chenopodium album). The confluence of the South Pease River and the

Middle Pease River occurred on the study area. The MWMA had a rest-rotation,

cow/calf operation grazing lease from January 2000 through September 2002, and cattle

were removed from MWMA October 2002 through September 2004. Cattle stocking

densities ranged from 1.82-13.37 ha/AU. Adjacent lands also used forms of rotational

grazing or continuous grazing, and based the timing of cattle movement on available

standing forage and supplemented with cattle feed. The herd was primarily a cow/calf

operation. Densities on adjacent lands ranged from 2.09-31.54 ha/AU. On both the

MWMA and adjacent lands, pastures were grazed for at least 1 month before the herd

was moved. Drought conditions led to cattle being removed from the MWMA

(treatment) from October 2002 through September 2004. Adjacent lands served as the

control because cattle were not removed. This enabled us to compare Rio Grande wild

turkey home range sizes on pastures before and after cattle removal. The Matador study

area was the only site used for comparison of home ranges between treatment and

control.

73

Hunting turkeys was allowed throughout the Texas Panhandle during our study,

with a maximum of 4 birds harvested per hunter in both the fall and spring season

combined. Fall hunting was split into 2 periods, a 4-week archery season from late

September to late October and a 9-week gun season from early November through early

January. Both sexes were legal during the fall season only (Texas Parks and Wildlife

Department 2004). The Kansas turkey hunting season occurred only in the spring and

went from early April through mid-May each year with a 1 bearded turkey bag limit

(Kansas Department of Wildlife and Parks 2004). Across the 5 years of the study,

hunters legally harvested a total of 49 transmittered male turkeys on private lands. On

the two wildlife management areas (MWMA, GHWMA), bag limits were restricted to

one bird during the spring season and hunts were by special permit only. Each spring on

MWMA, beginning in 2001, 5-10 permits were issued for male turkeys only. In 2001

through 2004, a total of 2, 3, 3, and 5 males, respectively, were harvested (Donald

Ruthven, Texas Parks and Wildlife Department, personal communication). On GHWMA

from 2000-2004, 4, 8, 7, 6, and 4 males, respectively, were harvested (Derrick Holdstock,

Texas Parks and Wildlife Department, personal communication). None of the

management area harvested turkeys were transmittered birds.

74

3.3 Methods

3.3.1 Capture and Telemetry

We trapped Rio Grande wild turkeys using rocket nets (Bailey et al. 1980) and

drop nets (Glazener et al. 1964) on sites baited with corn or grain sorghum from January

through March of each year. We captured additional birds each year to maintain 75

active transmitters at each study area. Upon capture, we recorded age and sex of each

bird (Pelham and Dickson 1992), and placed 110-gram backpack-style radio transmitters

from Advanced Telemetry Systems (ATS, Isanti, Minnesota, USA) or AVM Instruments

(Livermore, CA, USA) on selected individuals. Transmitters were equipped with a

mortality switch that activated after 8-hrs of inactivity. Birds were also fitted with Texas

Parks and Wildlife aluminum leg bands (size 9 for males) for further identification. We

relocated transmittered birds with ATS receivers, a hand-held 3-element yagi antenna, a

truck-mounted omni-directional antenna, and a truck-mounted null-peak system

(Balkenbush and Hallett 1988, Samuel and Fuller 1996). We located each radio-tagged

turkey ≥2 times per week during the breeding season (1 April to 31 August), and once per

week throughout fall and winter (1 September to 31 March). We collected both visual

observation of transmittered birds and radio telemetry triangulation locations with a null

peak system (Samuel and Fuller 1996). Universal Transverse Mercator (UTM)

coordinates were collected on visual sightings using a Trimble Geoexplorer 2 or

Geoexplorer 3 (Trimble Navigation Limited, Sunnyvale, California, USA) Global

Positioning System (GPS). The goal of triangulation was to obtain at least 3 compass

bearings, separated by >45° within 1 hour total elapsed time from first bearing taken

75

(White and Garrott 1990). We stratified locations into 4 time periods based on turkey

behavior: roosting (from dusk until dawn), morning (AM) feeding (first 1/3 of daylight

hours), midday (second 1/3 of daylight hours), and afternoon (PM) feeding (third 1/3 of

daylight hours). We used the computer program Location of a Signal (LOAS, Ecological

Software Solutions, Sacramento, California) and the associated maximum likelihood

estimator method to generate UTM positions of the triangulated animals and a

corresponding error ellipse. We determined study area specific telemetry error ≥1 time

per year by triangulating known location transmitters. Each transmitter was placed at

distances commonly associated with wild turkey telemetry from our study areas (≤3 km).

We also used known location transmitters to adjust for system biases and to calibrate the

truck mounted null-peak system (White and Garrott 1990, Samuel and Fuller 1996).

3.3.2 Vegetation Measurements

In order to describe vegetative characteristics of grazed and non-grazed pastures,

we recorded vegetation measurements at random (PR) locations derived from visual

observations of turkeys across 3 of the 4 study sites (grazing information was not

available at the SFRR study area to classify the plot into a grazed or non-grazed pasture).

Each plot was 50 m in a random cardinal direction from the observed turkey location. At

each plot, we established a 10 m x 20 m plot oriented north to south around the center

point (Figure 3.2). We measured percentage of vertical visual obstruction using an ocular

tube (i.e., 2-4 cm in diameter with crosshairs at one end) at 20 evenly spaced points

around the plot perimeter (Dueser and Shugart 1978). We recorded each reading as a hit

76

(i.e., canopy covered the crosshair) or a miss. We divided the number of hits by 20 and

multiplied by 100 for a percent estimate of vertical visual obstruction. We measured

horizontal visual obstruction using a visual obstruction pole (Robel et al. 1970) with 10

red and white 1-dm bands. We recorded the lowest Robel pole dm segment completely

visible at each measurement point, including center point. We placed the pole along the

plot centerline and observed it from a distance of 4 m and a height of 1 m (Figure 3.2,

point c to point d). We took readings along the centerline (Figure 3.2, point a to point b)

every 2 m, alternating sides of the centerline, for a total of 10 horizontal visual

obstruction measurements (including a reading at 20 m). We classified ground cover in

the plot into 5 categories: grass, shrub, bare ground, forb, and litter. We estimated

ground cover using the ocular tube along the 4 m chain between the Robel and sighting

pole. Ground cover classes immediately below the crosshairs were recorded for that data

point, from which a frequency estimate of percent ground cover of each class was

calculated (% grass cover, % shrub cover, % bare gound, % forb cover, % litter cover).

In 2003-2004, we adjusted vegetation measurement methods slightly to improve

sampling efficiency. We continued to measure vegetative characteristics at PR locations

to quantify turkey microhabitat features. However, to assess overall habitat availability,

we added 200 randomly selected point vegetation plots at 3 study areas (Matador,

Gene Howe, Cimarron) per year (2003-2004). We recorded 4 horizontal visual

obstruction measurements at the plot center point (1 in each cardinal direction) and

reduced the number of readings down the plot centerline to 5 alternating in east-west

cardinal directions. Thus, we recorded ground vegetation along 5 transects instead of 10.

77

If the point on the ground revealed a shrub, we recorded species and height class of the

shrub. We also measured vertical visual obstruction using a spherical densiometer held at

1 m height above plot center point (Lemmon 1956). We took 1 reading in each of the 4

cardinal directions and averaged the 4 readings together to compute % vertical visual

obstruction.

3.3.3 Home Range Size Calculation

In order to accurately ascertain the number of locations necessary for effective

home range calculation, we developed an area-observation curve (Odum and Kuenzler

1955) using a random sample of 10 birds (each study site, year, age, and sex class were

sampled evenly) with ≥30 locations. We formed a 95% MCP around increments of 5

locations using the Animal Movements extension in ArcView (ESRI, Los Angeles,

California, USA). Previous studies have divided the year into 4 periods: spring

(1 February-31 May), summer (1 June-30 Sep), and fall/winter (1 October-31 January)

(Godwin et al. 1994, Godwin et al. 1995, Palmer et al. 1996), but we did not have

sufficient locations in each of the 4 seasons for effective analysis. Thus, we divided the

year into 2 periods based on turkey biology and previous research: the breeding season

(1 April to 31 August, based on the earliest female nest initiation day; Figure 3.3) and

fall/winter season (1 September to 31 March). Males were visibly breeding during the

first 3 months of the defined breeding season, but in order to provide sufficient locations

for home range analysis and to be consistent with female turkey biology, we extended

this period to 31 August for male turkeys. Previous research also indicated the average

78

dispersal date of juvenile Rio Grande wild turkeys was 15 March to 15 April (Phillips

2004). By dividing daily locations at April 1, we would remove some home range

variation due to dispersal and make relationships to grazing easier to determine.

Breeding season locations were the only locations used for home range analyses. Home

range sizes were determined by analyzing the LOAS generated locations of each

transmittered bird using the Animal Movements extension in ArcView (ESRI, Los

Angeles, California, USA). We used 95% fixed kernel home range estimations (Worton

1989, Seaman and Powell 1996, Worton 1995) to calculate breeding season home range

sizes of male turkeys. The descriptive characteristics of breeding season home ranges for

adult and juvenile turkeys from all study areas were documented and compared to

previous studies.

3.3.4 Turkey Home Ranges and Cattle

We compared calculated male turkey home ranges to pasture use by cattle at the

Matador study area. We did not investigate specific grazing regime responses because

Matador adjacent lands had different grazing rotation systems than MWMA and we were

interested in trends at Matador study area across years. We determined specific pasture

use by cattle and timing of cattle rotations by interviewing landowners and state agency

personnel. The treatment and control were grazed from 2000-2002 and cattle were

removed from the treatment area October 2002 to August 2004. We compared male Rio

Grande turkey home ranges on the treatment and control from 2000-2004. We used

79

Mann-Whitney U-tests (Zar 1999) for each comparison. Means were also provided for

comparison to the literature.

3.3.5 Anthropogenic Food Sources

We defined male turkey home ranges with ≥1 known anthropogenic food source

(planted agriculture field or wildlife food plot) using a previously created vegetative

cover layer in a geographic information system (GIS) database (ArcView, WSRI, Los

Angeles, California, USA) (Brunjes 2005). We calculated home ranges, and compared

each home range to the vegetative layer to determine if man-made food sources were

present within the home range boundaries.

3.3.6 Individual Locations and Cattle

We collected individual turkey locations and pasture use information to assess use

of grazed and non-grazed pastures in relation to their availability at the 3 study areas

(Matador, Gene Howe, Cimarron). We determined specific pasture use by cattle and

timing of cattle rotations by interviewing landowners and agency personnel (state and

federal). We used ArcGIS to clip pastures with a known grazing history that contained

turkey locations. Each location was then assigned to a grazed or non-grazed category.

We defined grazing for individual location use-availability analyses as presence of cattle

in a pasture at the same time as turkey locations were recorded. Roost locations were

deleted from the dataset to prevent over-sampling in one specific area because the birds

returned to the same roost on a regular basis. We also used ≤3 day locations per week for

80

each bird, separated by at least 24 hours to maintain sampling independence (White and

Garrott 1990). We identified grazed and non-grazed areas each year to determine amount

of each available for use by male Rio Grande turkeys and compared that to areas used

with a chi-square test (Neu et al. 1974, Byers et al. 1984) and an associated Bailey

confidence interval (Bailey 1980, Cherry 1996). We calculated the areas of grazed and

non-grazed pastures by summing the amount of each, every time the cattle were moved

from 1 pasture to another. We examined use-availability for the entire year and for the

breeding season only.

3.3.7 Available Displaying Habitat

We assigned all random vegetation plots from 3 study areas (Matador,

Gene Howe, Cimarron) a grazing class; either non-grazed (plots that were non-grazed

from 1 April to the time of plot measurement) or grazed (plots that were grazed at some

period between 1 April and time of plot measurement). We then compared vegetative

characteristics between grazed and non-grazed pastures using Mann-Whitney U-tests

(Zar 1999). Means were also provided for comparison to the literature.

We also compared availability of nesting habitat between pasture types. Each

vegetative characteristic (% vertical visual obstruction, % horizontal visual obstruction,

% grass cover, % shrub cover, % bare ground, % forb cover, and % litter cover) was

grouped into classes, adhering to Cochran's rules regarding χ2 analyses (Cochran 1952).

We used χ2 goodness of fit tests (Conover 1999) to investigate differences in

distributions and used Bonferroni intervals to compare individual classes within each

81

measured vegetative characteristic if the goodness of fit test was significant (Neu et al.

1974, Byers et al. 1984).

We used SPSS for Windows (Release 9.0.0, 1998) for all statistical calculations.

This research was approved by the Texas Tech University Animal Care and Use

Committee (Protocol #'s 99917 and 01173B).

3.4 Results

3.4.1 Capture and Telemetry

Mean telemetry error polygon, calculated among all sites and seasons using

known location transmitters (n = 182), fell within 118 m of the true location. The

associated error polygon was 4.40 ha. Due to the large average pasture size (x̄ = 836 ha),

telemetry error was <1% of average pasture size and <1% of vegetative zone size.

3.4.2 Home Range Size Calculation

The number of individual locations for each randomly selected bird ranged from

30-65. The area-observation curve (Figure 3.4) indicated that 25 locations accounted for

86% of home range area. This was the point where the asymptote increased <5% in

home range size for every increment of 5 locations. Based on these findings, 25 locations

were used as the minimum number of locations for breeding season home range analyses,

birds with <25 locations were excluded from analysis. A total of 210 male home ranges

were used in analyses. Average home range sizes ranged from 475 ha to 4,683 ha

82

(Table 3.1). The pooled average home range size for adult male turkeys was 1,830 ha

and the average pooled home range size for juvenile turkeys was 1,475 ha.

3.4.3 Turkey Home Ranges and Cattle

Home ranges at the Matador study area (n = 76) exhibited variation between

treatment and control. There were no differences between 2003-2004 treatment and

control (P = 0.270) or between years in the treatment (P = 0.596; Figure 3.5). In

2000-2002, average control home range size was larger (P ≤ 0.001) than average

treatment home range size (Figure 3.5). Average home range size on the control

decreased (P = 0.003) between year periods (Figure 3.5). The same patterns were present

when average juvenile male home range size was compared separately (Figure 3.6).

However, adult male average home range size on the control between year periods was

not statistically different (P = 0.131; Figure 3.7).

3.4.4 Anthropogenic Food Sources

Of the 76 MWMA male turkeys with sufficient daily locations for home range

calculation, 52.6% contained all or part of a known anthropogenic food source within the

calculated boundaries.

83

3.4.5 Individual Locations and Cattle

Individual male Rio Grande wild turkey radio telemetry locations were distributed

proportionately to occurrence of pasture types (P > 0.5) for the entire year (χ² = 0.12;

Table 3.3) and breeding season (χ² = 0.16; Table 3.4). Associated Bailey confidence

intervals around proportion of locations observed in each individual pasture type during

the year (Table 3.3) and for the breeding season (Table 3.4) supported the overall

χ² analysis, no selection was evident (P > 0.5).

3.4.6 Available Displaying Habitat

Random plots in non-grazed pastures (n = 924) had greater vertical visual

obstruction (P ≤ 0.001), grass cover (P ≤ 0.001), litter cover (P ≤ 0.001) and less bare

ground (P = 0.002), and forb cover (P ≤ 0.001) than random plots in grazed pastures

(n = 654; Table 3.5).

Availability of grazed pasture vegetative characteristics differed (χ² ≥ 5.41,

P < 0.025; Table A.5) from non-grazed pastures. However, individual classes were

usually not different enough to be individually significant. Both shrub (Figure 3.8) and

bare ground (Figure 3.9) cover were not significantly different (P > 0.05) in each class,

even though the bare ground trend showed grazed pastures had a larger proportion of

plots with 20-49.9% bare ground.

84

3.5 Discussion

Previous research recommended a minimum of 30, preferably ≥50, locations used

for home range estimation (Otis and White 1999, Seaman et al. 1999) and that home

range estimates varied depending on number of observations, time of year, density of

vegetation, geographic location, and methods of calculation (Brown 1980). Our results

indicated that ≥25 locations provided sufficient locations to reduce the asymptote slope to

<5% increase for each 5 location increment. The reduced number of needed locations

compared to previous studies was likely a result of our division of seasons. By dividing

the year into 2 seasons, based on initiation of breeding activities and completion of

dispersal activities, home range variation due to dispersal was removed. Our study had

larger home range sizes than previously reported (Table 3.2). In addition to reasons for

variation described by Brown (1980), some previously reported home range estimates did

not use radio telemetry; data gathering and calculations were based on visual

observations and band recovery (Lewis 1963, Ellis and Lewis 1967, Davis 1973).

Furthermore, most studies were conducted east of the Mississippi River on Eastern wild

turkeys, where greater availability of roost trees may have reduced turkey movement

distances.

Use of home ranges to indicate changes in utilization of space with and without

the presence of cattle was not the most effective method of detailing differences, under

the circumstances of our study, due to confounding factors. Home ranges were highly

variable regardless of age class, year group, or treatment and other factors appear to have

masked our treatment effects. Variation in home range sizes under the different

85

situations was likely due to factors such as physiological condition, social status

(Badyaev 1996), and food availability (Mosby and Handley 1943, Wheeler 1948, Miller

et al. 1985, Kurzejeski and Lewis 1990) not directly measured in this study. Where cattle

were fed supplemental sources, turkeys were often observed foraging in the food remains,

which may artificially effect turkey home range sizes (Thomas et al. 1966). There were

also situations where turkeys traveled considerable distances to use planted artificial food

sources (Figure 3.10), which inflated overall home range averages. When all Matador

study area individual turkey home ranges were analyzed for anthropogenic food source

use, 52.6% of the home ranges had a known anthropogenic source. However, unknown

anthropogenic sources, such as supplemental cattle feeding stations and wild game

feeders, which could not be accounted for post hoc, were likely part of more home ranges

than we were able to document. Turkey use of anthropogenic food sources needs further

research because they may impact home range size, survival, and reproduction. Another

possible reason for home range variation was our grazing intensities. Even though we

had a grazed treatment and a non-grazed treatment, our grazing densities were low at

times, and may have had no influence on turkey movements or home ranges. Schulz and

Leininger (1990) observed changes in vegetative composition and structure when zones

were consistently grazed heavily (<4 ha/AUM). There was likely a threshold of grazing

pressure that would impact male turkey habitat, but grazing in this study with intensities

and periods of rest we experienced did not breach that point.

Our results may also be a result of scale differences in our comparisons. Home

ranges did not take into account direct contact of turkeys and cattle because they

86

investigated larger periods of time. There may have been a more immediate interaction

that would have been seen in individual bird response to cattle locations on a daily basis,

a much smaller scale, but the home range scale was useful at determining use of

anthropogenic food sources that are more time stable and do not fluctuate from day to

day.

Due to variance present in our Matador study area home range data, we examined

male Rio Grande turkey individual radio telemetry locations in relation to cattle presence

and absence. Previous studies indicated turkeys selected for lightly grazed or deferred

pastures (Walker 1948) or turkeys exhibited no preference (Jones 1981). Individual

locations of male turkeys showed no selection for grazed or non-grazed pastures, which

differed from female turkey pasture use (Hall 2005). Males did not select against grazed

pastures and may have used them for breeding behaviors, especially displaying.

Displaying sites used by male Rio Grande turkeys were characterized by open areas with

less shrubs than neighboring areas (Holdstock 2003). We hypothesized that availability

of displaying site characteristics would be different in each pasture type. Even though

there was more average bare ground in grazed pastures (P = 0.002), the availability of

individual classes of bare ground and shrub cover were no different (P > 0.05) in grazed

and non-grazed pastures. Based on our data with the grazing intensities present on our

study sites, male turkeys used both pasture types in proportion to their availability and

there was not an increased availability of displaying site characteristics in either pasture.

In fact, grazing likely served as a tool to maintain open areas important for breeding,

feeding, and brood rearing activities (Spears 2002).

87

Based on the HR and individual location analyses used, individual location

analyses provided more detailed information about use of pastures than the home range

analyses. This is important for land managers that want to maintain turkey populations

and a profitable cattle enterprise. Grazing at light to moderate levels with periods of rest

in upland sites during the breeding season (1 April-31 August) not only would allow for

standing forage use by nesting females in riparian zones but also would maintain open

areas used by male turkeys during the breeding season for displaying purposes.

88

3.6 Literature Cited

Badyaev, A. V., W. J. Etges, T. E. Martin. 1996. Ecological and behavioral correlates of variation in seasonal home ranges of wild turkeys. Journal of Wildlife Management 60: 154-164.

Bailey, B. J. R. 1980. Large sample simultaneous confidence intervals for the multinomial probabilities based on transformations of the cell frequencies. Technometrics 22: 583-589.

Bailey, W., D. Dennett, H. Gore, J. Pack, R. Simpson, and G. Wright. 1980. Pages 10-23 in Proceedings of the Fourth National Wild Turkey Symposium. National Wild Turkey Federation, March 2-5, 1980, Little Rock, Arkansas, USA.

Baker, D. L., and F. S. Guthery. 1990. Effects of continuous grazing on habitat and density of ground-foraging birds in south Texas. Journal of Range Management 43: 2-5.

Balkenbush, J. A., and D. L. Hallett. 1988. An improved vehicle-mounted telemetry system. Wildlife Society Bulletin 16: 65-67.

Barwick, L. H., and D. W. Speake. 1973. Seasonal movements and activities of wild turkey gobblers in Alabama. Pages 125-134 in G.C. Sanderson and H.C. Schultz, eds. Wild turkey management: current problems and programs. University of Missouri Press, Columbia, Missouri, USA.

Bidwell, T. G., S. D. Shalaway, O. E. Maughan, L. G. Talent. 1989. Habitat use by female eastern wild turkeys in southeastern Oklahoma. Journal of Wildlife Management 53: 34-39.

Brown, E. K. 1980. Home range and movements of wild turkeys -- a review. Pages 251-261 in Proceedings of the Fourth National Wild Turkey Symposium. National Wild Turkey Federation, March 2-5, 1980, Little Rock, Arkansas, USA.

Brunjes, J. H. 2005. The population biology and landscape ecology of Rio Grande wild turkeys in the rolling plains of Texas and Kansas. Dissertation, Texas Tech University, Lubbock, Texas, USA.

Bryant, F. C., F. Guthery, and W. Webb. 1981. Grazing management in Texas and its impact on selected wildlife. Pages 94-112 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

89

Byers, C. R., R. K. Steinhorst, and P. R. Krausman. 1984. Clarification of a technique for analysis of utilization-availability data. Journal of Wildlife Management 48: 1050-1053.

Cherry, S. 1996. A comparison of confidence interval methods for habitat use-availability studies. Journal of Wildlife Management 60: 653-658.

Cochran, W. G. 1952. The χ2 test of goodness of fit. Annals of Mathematical Statistics 23: 315-345.

Conover, W. J. 1999. Practical nonparametric statistics. John Wiley and Sons, New York, New York, USA.

Davis, J. 1981. Livestock vs. riparian habitat management-there are solutions. Pages 175-184 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

Davis, J. R. 1973. Movements of wild turkeys in southwestern Alabama. Pages 135-140 in G.C. Sanderson and H.C. Schultz, eds. Wild turkey management: current problems and programs. University of Missouri Press, Columbia, Missouri, USA.

Diamond, J. M. 1978. Niche shifts and the rediscovery of interspecific competition. American Scientist 66: 322-331.

Dickey, H. P., W. R. Swafford, and Q. L. Markley. 1963. Soil survey of Morton County, Kansas. United States Department of Agriculture, Soil Conservation Service and Forest Service, in cooperation with the Kansas Agricultural Experiment Station.

Dueser, R. D., and H. H. Shugart, Jr. 1978. Microhabitats in a forest-floor small mammal fauna. Ecology 59: 89-98.

Ellis, J. E., and J. B. Lewis. 1967. Mobility and annual range of wild turkeys in Missouri. Journal of Wildlife Management 31: 568-581.

Everett, D. D., D. W. Speake, W. K. Maddox. 1980. Natality and mortality of a north Alabama wild turkey population. Pages 117-126 in Proceedings of the Fourth National Wild Turkey Symposium. National Wild Turkey Federation, March 2-5, 1980, Little Rock, Arkansas, USA.

Fleming, W. H. 1975. Study of home ranges and gobbling activities of wild turkeys during the 1973 breeding season. Thesis, Clemson University, Clemson, South Carolina, USA.

90

Glazener, W. C., A. S. Jackson, and M. L. Cox. 1964. The Texas drop-net turkey trap. Journal of Wildlife Management 28: 280-287.

Godwin, K. D., G. A. Hurst, B. D. Leopold. 1994. Movements of wild turkey gobblers in central Mississippi. Proceedings of the Southeastern Association of Fish and Wildlife Agencies 48: 117-121.

Godwin, K. D., G. A. Hurst, B. D. Leopold. 1995. Size and percent overlap of gobbler home ranges and core-use areas in central Mississippi. Pages 45-52 in Proceedings of the Seventh National Wild Turkey Symposium. National Wild Turkey Federation, May 24-26, 1995, Rapid City, South Dakota, USA.

Hall, G. I. 2005. Relationships between cattle grazing and Rio Grande wild turkeys in the Southern Great Plains. Thesis, Texas Tech University, Lubbock, Texas, USA.

Hoffman, R. W. 1991. Spring movements, roosting activities, and home-range characteristics of male Merriam's wild turkey. Southwestern Naturalist 36: 332-337.

Holdstock, D. P. 2003. Survival, movement, and habitat selection of male Rio Grande wild turkeys in the Texas panhandle and southwestern Kansas. Thesis, Texas Tech University, Lubbock, Texas, USA.

Holdstock, D. 2004. Personal communication. Texas Parks and Wildlife Department, Canadian, Texas, USA.

Jones, K. H. 1981. Effects of grazing and timber management on Merriam's turkey habitat in mixed conifer vegetation of southcentral New Mexico. Thesis, New Mexico State University, Las Cruces, New Mexico, USA.

Kansas Department of Wildlife and Parks. 2004. 2004 Kansas hunting and furharvesting regulations. Kansas Department of Wildlife and Parks, Topeka, Kansas, USA.

Korschgen, L. J. 1967. Feeding habits and food. Pages 137-198 in O.H. Hewitt, editor. The wild turkey and its management. The Wildlife Society, Washington, DC. USA.

Kurzejeski, E. W., and J.B. Lewis. 1990. Home ranges, movements and habitat use of wild turkey hens in northern Missouri. Pages 67-71 in Proceedings of the Sixth National Wild Turkey Symposium. National Wild Turkey Federation, March 1990, Charleston, South Carolina, USA.

Lemmon, P. E. 1956. A spherical densiometer for estimating forest overstory density. Forest Science 2: 314-320.

91

Lewis, C., and T. Harshbarger. 1986. Burning and grazing effects on bobwhite foods in the southeastern Coastal Plain. Wildlife Society Bulletin 14: 455-459.

Lewis, J. C. 1963. Observations on the winter range of wild turkeys in Michigan. Journal of Wildlife Management 29: 44-52.

Ligon, J. S. 1946. History and management of Merriam's wild turkey. New Mexico Game and Fish Commission Bulletin, New Mexico, USA.

McEwen, H. F., F. F. Wheeler, and J. C. Williams. 1973. Soil survey of Collingsworth County, Texas. United States Department of Agriculture, Soil Conservation Service, in cooperation with the Texas Agricultural Experiment Station.

Miller, B. K., P. D. Major, S. E. Backs. 1985. Pages 233-244 in Proceedings of the Fifth National Wild Turkey Symposium. National Wild Turkey Federation, June, 1985, Des Moines, Iowa, USA.

Miller, D. A., G. A. Hurst, B. D. Leopold. 1999. Habitat use of Eastern wild turkeys in central Mississippi. Journal of Wildlife Management 63: 210-222.

Mosby, H. S., and C. O. Handley. 1943. The wild turkey in Virginia. Virginia Commission of Game and Inland Fish, Richmond, Virginia, USA.

Neu, C. W., R. C. Byers, and J. M. Peek. 1974. A technique for analysis of utilization-availability data. Journal of Wildlife Management 38: 541-545.

Nielson, L., and C. Yde. 1981. The effects of rest-rotation grazing on the distribution of sharp-tailed grouse. Pages 147-165 in Proceedings of the wildlife-livestock relationships symposium. Forest, Wildlife, and Range Experiment Station, University of Idaho, April 20, 1981, Moscow, Idaho, USA.

Odum, E. P., and E. J. Kuenzler. 1955. Measurement of territory and home range size in birds. Auk 72: 128-137.

Otis, D. L., and G. C. White. 1999. Autocorrelation of location estimates and the analysis of radiotracking data. Journal of Wildlife Management 63: 1039-1044.

Palmer, W. E., K. D. Godwin, G. A. Hurst, and D. A. Miller. 1996. Effects of prescribed burning on wild turkeys. Pages 228-236 in Transactions of the 61st North American Wildlife and Natural Resources Conference. Wildlife Management Institute, March 22-27, 1996, Tulsa, Oklahoma, USA.

Pelham, P. H., and J. G. Dickson. 1992. Physical characteristics. Pages 32-45 in J.G. Dickson, editor. The wild turkey: biology and management. Stackpole Books, Mechanicsburg, Pennsylvania, USA.

92

Phillips, R. S. 2004. Movements, survival, and reproduction of Rio Grande wild turkeys in the Texas panhandle. Thesis, Texas Tech University, Lubbock, Texas, USA.

Prasad, N. L. N. S., and F. Guthery. 1986. Wildlife use of livestock water under short duration and continuous grazing. Wildlife Society Bulletin 14: 450-454.

Raybourne, J. W. 1968. Telemetry of turkey movements. Thesis, Virginia Polytechnical Institute, Blacksburg, Virginia, USA.

Richardson, W. E., J. Hajek, E. Blakley, and C. Neitsch. 1974. Soil survey of Cottle County, Texas. United States Department of Agriculture, Soil Conservation Service, in cooperation with the Texas Agricultural Experiment Station.

Robel, R., J. Briggs, A. Dayton, and L. Hulbert. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management 23: 295-297.

Rosenzweig, M. L. 1981. A theory of habitat selection. Ecology 62: 327-335.

Ruthven, D. 2004. Personal communication. Texas Parks and Wildlife Department, Paducah, Texas, USA.

Samuel, M. D., and M. R. Fuller. 1996. Wildlife radiotelemetry. Pages 370-418 in T.A. Bookhout, editor. Research and management techniques for wildlife and habitats. Fifth edition. The Wildlife Society, Bethesda, Maryland, USA.

Schoener, T. W. 1974. Resource partitioning in ecological communities. Science 185: 27-39.

Schulz, P., and F. Guthery. 1987. Effects of short duration grazing on wild turkey home ranges. Wildlife Society Bulletin 15: 239-241.

Schulz, T. T., and W. C. Leininger. 1990. Differences in riparian vegetation structure between grazed areas and exclosures. Journal of Range Management 43: 295-299.

Seaman, D. E., J. J. Millspaugh, B. J. Kernohan, G. C. Brundige, K. J. Raedeke, R. A. Gitzen. 1999. Effects of sample size on kernel home range estimates. Journal of Wildlife Management 63: 739-747.

Seaman, D. E., and R. A. Powell. 1996. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology 77: 2075-2085.

Speake, D. W., L. H. Barwick, H. O. Hillestad, W. Stickney. 1969. Some characteristics of an expanding turkey population. Proceedings of the Southeast Association Game and Fish Commission 23: 46-58.

93

Spears, B. L. 2002. Wild turkey pre-flight chick habitat characteristics and survival. Thesis, Texas Tech University, Lubbock, Texas, USA.

Texas Parks and Wildlife Department. 2004. 2003-2004 Outdoor Annual. Texas Parks and Wildlife Department, Austin, Texas, USA.

Thomas, J. W., C. V. Hoozer, R. G. Marburger. 1966. Wintering concentrations and seasonal shifts in range in the Rio Grande turkey. Journal of Wildlife Management 30: 34-49.

Walker, E. A. 1948. Rio Grande turkey in the Edwards Plateau of Texas, 1946-1948. Texas Game, Fish, and Oyster Commission, Final Report Federal Aid Project 22-R.

Wheeler, R. J. 1948. The wild turkey in Alabama. Alabama Department of Conservation, Alabama, USA.

White, G. C., and R. A. Garrott. 1990. Analysis of wildlife radio-tracking data. Academic Press Inc., San Diego, California, USA.

Wilkins, R. N. and W. G Swank. 1992. Bobwhite habitat use under short duration and deferred-rotation grazing. The Journal of Range Management 45: 549-553.

Williams, J. C., A. J. Welker, F. F. Wheeler, and H. F. McEwen. 1974. Soil survey of

Hemphill County, Texas. United States Department of Agriculture, Soil Conservation Service, in cooperation with the Texas Agricultural Experiment Station.

Worton, B. J. 1989. Kernel methods for estimating the utilization distribution in home-range studies. Ecology 70: 164-168.

Worton, B. J. 1995. Using monte carlo simulation to evaluate kernel-based home range

estimators. Journal of Wildlife Management 59: 794-800. Zar, J. H. 1999. Biostatistical analysis. Prentice-Hall, Upper Saddle River, New Jersey,

USA. Zwickel, F. C. 1972. Some effects of grazing on blue grouse during summer. Journal of

Wildlife Management 36: 631-634.

94

Table 3.1. Average breeding season (1 April -31 August) home range sizes of male Rio Grande wild turkeys on 4 study areas from January 2000 through August 2004 in the southern Great Plains. Study Site Age n Average m2 Average acre Average hectare

Gene Howe Adult 32 4,750,4623 1,174 475

Gene Howe Juvenile 26 5,992,287 1,481 599

Cimarron Adult 19 46,824,518 11,571 4,683

Cimarron Juvenile 10 32,855,063 8,119 3,286

Matador Adult 43 10,904,222 2,695 1,090

Matador Juvenile 33 12,879,516 3,183 1,288

Salt Fork Adult 27 10,732,855 2,652 1,073

Salt Fork Juvenile 20 7,283,096 1,800 728

Pooled Adult 121 18,303,015 4,523 1,830

Pooled Juvenile 89 14,752,491 3,645 1,475

95

96

Table 3.2. Average reported North American home range sizes for male turkeys from 1963 to 2005 (1963 -1975 list compiled by Brown 1980). Source (ha) n subspecies Season Location Lewis 1963 683 34 Eastern winter Michigan Ellis and Lewis 1967 553 4 Eastern annual range Missouri Raybourne 1968 198 4 Eastern 2 months Virginia Barwick and Speake 1973 398 12 Eastern annual range Alabama 171 6 Eastern autumn Alabama 270 6 Eastern winter Alabama 204 6 Eastern spring Alabama 133 6 Eastern summer Alabama Davis 1973 244 15 Eastern annual Alabama Fleming 1975 95 8 Eastern spring South Carolina Speake et al. 1975 350 16 Eastern spring-summer Alabama Hoffman 1991 520 11 Merriam's; adult 1 April-15 June Colorado 1230 8 Merriam's; juvenile 1 April-15 June Colorado Godwin et al. 1995 1130 53 Eastern spring Mississippi 653 53 Eastern summer Mississippi 1134 53 Eastern fall-winter Mississippi Hall et al. 2005 1830 121 Rio Grande; adult spring-summer Texas 1475 89 Rio Grande; juvenile spring-summer Texas

Table 3.3. Selection of pasture types by male Rio Grande wild turkeys using all locations from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall chi-square critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 0.12. Relative Expected Locations Observed Pasture Area (Ha) area locations observed proportion Bailey interval Grazed 366,009 0.349 1,541.06 1,530 0.347 0.311 ≤ P ≤ 0.382 NS a

Non-grazed 682,579 0.651 2,873.94 2,885 0.653 0.617 ≤ P ≤ 0.687 NS a NS indicates no statistical difference

97

Table 3.4. Selection of pasture types by male Rio Grande wild turkeys during the breeding season (1 April-31 August) from April 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). The overall chi-square critical value with α = .05, df = 1: χ² = 3.84; calculated χ² was 0.16. Relative Expected Locations Observed Pasture Area (Ha) area locations observed proportion Bailey interval Grazed 254,197 0.337 928.89 919 0.333 0.299 ≤ P ≤ 0.368 NS a

Non-grazed 500,544 0.663 1,829.11 1,839 0.667 0.631 ≤ P ≤ 0.700 NS a NS indicates no statistical difference

98

Table 3.5. Comparison of habitat characteristics measured at grazed random plots and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the southern Great Plains (pooled over years and study areas). Mean (SE) Mean Rank Mann-Whitney Variable Grazed (n = 654) Non-grazed (n = 924) Grazed Non-grazed U P Vertical obstruction 8.5 (0.8) 11.9 (0.8) 736.56 826.97 267,528.0 ≤ 0.001 Horizontal obstruction 22.2 (0.5) 22.5 (0.4) 766.14 806.03 286,873.5 0.086 Grass cover 40.4 (0.9) 44.6 (0.8) 740.76 823.99 270,275.0 ≤ 0.001 Shrub cover 10.2 (0.5) 9.7 (0.4) 794.06 786.27 299,168.0 0.736 Bare ground 16.8 (0.6) 14.9 (0.5) 831.96 759.45 274,382.0 0.002 Forb cover 20.7 (0.7) 15.7 (0.5) 872.75 730.58 247,701.5 ≤ 0.001 Litter cover 10.7 (0.5) 13.1 (0.5) 738.24 825.78 268,626.5 ≤ 0.001

99

S

SFo

Cimarron: Cimarron National Grasslands, Elkhart, KGene Howe: Gene Howe Wildlife Management Area, CSalt Fork: Salt Fork of the Red River, Clarendon, TexMatador: Matador Wildlife Management Area, Padu Figure 3.1. Location of 4 study areas where relationshipsmale Rio Grande wild turkey home range size and pasturesouthern Great Plains from January 2000-August 2004.

100

Cimarron

Gene Howe

alt rk

r

Matado

ansas anadian, Texas as cah, Texas

between cattle gra use were studied

K

CO

NM

X

T

ziin

OK

ng and the

Figure 3.2. Description of vegetation plot used to measure data regarding Rio Grande wild turkey habitat use and availability in the southern Great Plains from January 2000-August 2004.

101

Dates of nest initiation

0

2

4

6

8

10

12

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95

Calendar days (1 April-3 July)

Cou

nt o

f Nes

ts

Figure 3.3. Bar graph of nest initiation dates for female Rio Grande wild turkeys from 4 study areas in the southern Great Plains. The years (2000-2004) and study areas (Matador, Salt Fork, Gene Howe, Cimarron) are pooled. The first number corresponds to the first known nest initiation day (1 April) and the last number corresponds to the last known nest initiation day (3 July).

102

Average Area-Observation Curve

0102030405060708090

100110

5 10 15 20 25 30 35 40 45 50 55 60 65

Telemetry Locations

% H

ome

Ran

ge

Figure 3.4. Average area-observation curve describing the relationship between the number of telemetry locations and home range estimates for radio-transmittered male Rio Grande wild turkeys in the southern Great Plains.

103

Treatment Control

2000-2002 Grazed

(424 ± 54 ha) n = 19

Not Equal P ≤ 0.001

Grazed (2210 ± 474 ha)

n = 25

Equal P = 0.596 Not Equal

P = 0.003

2003-2004 Non-Grazed

(637 ± 131 ha) n = 13

Equal P = 0.270

Grazed (938 ± 206 ha)

n = 19

Figure 3.5. Male home range comparison between the treatment (cattle removed 2003-2004) and control groups with pooled adult and juvenile home range sizes on the Matador study area in Paducah, Texas.

104

Treatment

Control

2000-2002 Grazed

(475 ± 94 ha) n = 7

Not Equal P ≤ 0.001

Grazed (2643 ± 691 ha)

n = 11

Equal P = 0.867 Not Equal

P = 0.004

2003-2004 Non-Grazed

(635 ± 187 ha) n = 8

Equal P = 0.694

Grazed (717 ± 103 ha)

n = 7

Figure 3.6. Male home range comparison between the treatment (cattle removed 2003-2004) and control groups with juvenile home range sizes only on the Matador study area in Paducah, Texas.

105

Treatment Control

2000-2002 Grazed

(394 ± 67 ha) n = 12

Not EqualP ≤ 0.001

Grazed (1869 ± 658 ha)

n = 14

Equal P = 0.279 Equal

P = 0.131

2003-2004 Non-Grazed

(639 ± 190 ha) n = 5

Equal P = 0.574

Grazed (1067 ± 319 ha)

n = 12

Figure 3.7. Male home range comparison between the treatment (cattle removed 2003-2004) and control groups with adult home range sizes only on the Matador study area in Paducah, Texas.

106

Shrub Cover

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0-9.910-19.9

20-29.930-39.9

40-49.950-59.9 >60

Class (% shrub cover)

Prop

ortio

n of

Plo

tss

GrazedNon-Grazed

Figure 3.8. Comparison of plot proportions with each respective class of shrub cover percentages between grazed and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant.

107

Bare Ground

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0-.9

10-19

.9

20-29

.9

30-39

.9

40-49

.9

50-59

.9

60-69

.9

70-79

.9 >80

Class (% bare ground)

Prop

ortio

n

GrazedNon-Grazed

Figure 3.9. Comparison of plot proportions with each respective class of bare ground percentages between grazed and non-grazed random plots from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). Bonferroni intervals that overlap are not significant.

108

Figure 3.10. Selected sample of 4 male home ranges, showing use of anthropogenic food sources (gray areas) and home range size variability on the Matador study area.

109

110

APPENDIX

Table A.1. Grazing rotations on the Matador Wildlife Management Area from January 2000 through August 2004. Pasture Areaa Date Densityb Date Density Date Density Date Density North Middle 2324 10/1/00-11/15/00 3.72 6/1/01-7/31/01 6.94 9/15/01-10/31/01 6.94 6/15/02-6/30/02 6.94 9/1/02-9/31/02 7.00 Sisk Pen 2570 4/1/01-5/31/01 7.67 6/1/01-7/15/01 4.11 8/1/01-9/15/01 7.67 3/1/02-6/15/02 7.67 Aermotor 1940 1/1/01-3/31/01 5.79 4/1/01-5/31/01 3.10 11/1/01-2/28/02 5.79 South Middle 1137 1/15/01-3/31/01 1.82 7/15/01-1/31/02 3.39 6/1/02-8/31/02 3.39 OX 694.5 1/1/00-2/28/01 8.68 8/15/01-10/31/01 11.58 Dogleg 669.8 11/15/00-1/15/01 2.46 5/15/02-7/15/02 11.16 Suitcase 501.6 3/1/01-5/31/01 6.27 11/1/01-2/14/02 8.36 Entrance 392.0 1/1/01-8/15/01 4.90 2/15/02-5/15/02 6.53 Shorty 351.2 11/15/00-1/15/01 2.46 7/15/02-9/31/02 5.85 Headquarters 334.3 8/15/01-9/30/02 13.37 Bull 240.0 Mouth of River 112.2 Cow Trap 83.31 111 Charlie FPc 1.560 Parson's FP 7.77 Slim FP 4.050 Turkey FP 2.880 Cow Hollow FP 1.930 Shorty FP 1.820 Sisk Pen FP 0.680 aarea of pasture in hectares bdensity of cattle in each pasture during the specified time frame in hectares per animal unit cfood Plot

Table A.2. Grazing rotations on the Matador study area adjacent land from January 2000 through August 2004. Pasture Areaa Date Densityb Date Density Date Density Date Density Date Density 1 4100 5/00-10/01 22.16 10/01-8/02 21.24 8/02-1/03 14.49 1/03-3/03 26.45 3/03-8/04 15.13 2 1606 5/00-10/01 12.35 10/01-8/02 12.85 8/02-1/03 22.30 1/03-3/03 22.30 3/03-8/04 9.07 3 997.7 5/00-10/01 11.09 10/01-8/02 10.73 8/02-1/03 12.79 1/03-3/03 12.79 3/03-8/04 7.67 4 571.7 5/00-10/01 9.22 10/01-8/02 11.43 8/02-1/03 22.87 1/03-3/03 22.87 3/03-8/04 10.59 5 504.6 5/00-10/01 9.34 10/01-8/02 8.55 8/02-1/03 6.47 1/03-3/03 20.18 3/03-8/04 9.52 6 387.4 5/00-10/01 7.31 10/01-8/02 7.04 8/02-1/03 18.45 1/03-3/03 19.37 3/03-8/04 15.50 7 3470 5/00-10/01 29.91 10/01-8/02 31.54 1/03-3/03 31.54 3/03-8/04 19.49 8 184.0 5/00-10/01 3.07 10/01-8/02 3.12 8/02-1/03 2.09 1/03-3/03 36.80 9 1259 5/00-10/01 30.06 1/03-3/03 9.19 3/03-8/04 14.30 10 513.6 5/00-10/01 4.21 10/01-8/02 4.18 8/02-1/03 8.70 11 442.0 8/02-1/03 14.73 1/03-3/03 14.73 12 1834 8/02-1/03 28.72 1/03-3/03 28.72 13 133.9 5/00-10/01 10.30 10/01-8/02 8.92 14 1142 112 15 275.0 16 116.6 17 115.8 18 99.08 19 32.53 aarea of pasture in hectares bdensity of cattle in each pasture during the specified time frame in hectares per animal unit

Table A.3. Grazing rotations on the Gene Howe Wildlife Management Area from January 2000 through August 2004. Pasture Areaa Date Densityb Date Density Date Density Middle 369.4 9/1/01-2/28/02 10.86 9/1/02-2/28/03 10.86 9/1/03-2/28/04 10.86 North 359.7 9/1/01-2/28/02 10.90 9/1/02-2/28/03 10.90 9/1/03-2/28/04 10.90 DeArment 345.5 9/1/01-2/28/02 11.52 9/1/02-2/28/03 11.52 9/1/03-2/28/04 11.52 Meadow 250.9 3/1/02-8/31/02 4.40 3/1/03-8/31/03 5.22 3/1/04-8/31/04 5.22 Persimmon 208.2 9/1/01-2/28/02 11.57 9/1/02-2/28/03 11.57 9/1/03-2/28/04 11.57 Hay Meadow 122.7 3/1/02-8/31/02 4.54 3/1/03-8/31/03 5.33 3/1/04-5/31/04 5.33 North Williams 105.5 9/1/01-2/28/02 10.55 9/1/02-2/28/03 10.55 9/1/03-2/28/04 10.55 Upper Meadow 105.5 3/1/02-8/31/02 4.39 3/1/03-8/31/03 5.27 3/1/04-8/31/04 5.27 Bunkhouse 75.51 3/1/02-8/31/02 4.44 3/1/03-5/31/03 5.39 3/1/04-8/31/04 5.39 East Bull 103.4 3/1/03-8/31/03 5.17 3/1/04-8/31/04 5.17 South Williams/ Williams West Bull 94.58 East Bull 39.54 113 aarea of pasture in hectares bdensity of cattle in each pasture during the specified time frame in hectares per animal unit

Table A.4. Grazing rotations on the Cimarron National Grasslands from January 2000 through August 2004. Pasture Areaa Date Densityb Date Density Date Density Date Density Date Density River NE 702.7 5/1-6/1/00 9.08 10/1-10/31/01 9.08 6/15-7/15/03 3.66 5/3-6/3/04 3.55 Bridge West 1016.4 5/1-6/1/00 9.08 10/1-10/31/01 9.08 8/15-10/31/04 5.13 Bridge East 1393.8 5/1-6/1/00 9.08 10/1-10/31/01 9.08 6/3-8/15/04 7.04 S Lowe E 1421.8 7/14-8/14/00 10.1 5/1-6/1/01 10.11 5/15-6/15/03 6.24 9/15-10/31/03 6.24 5/4-7/15/04 5.38 S Lowe W 1329.4 7/14-8/14/00 10.1 5/1-6/1/01 10.11 6/15-9/15/03 5.83 7/15-10/1/04 5.03 River E 960.3 7/14-8/14/00 10.1 5/1-6/1/01 10.11 5/15-6/15/03 5.00 10/1-10/31/04 3.64 E Art S 1105.7 7/1-8/14/00 9.27 5/1-6/1/01 9.24 7/15-9/30/03 5.67 5/5-7/15/04 5.5 E Art N 1338.6 7/1-8/14/00 9.27 5/1-6/1/01 9.24 5/15-7/15/03 6.86 9/30-10/31/03 6.86 7/15-10/1/04 River Mid E 669.9 7/1-8/14/00 9.27 5/1-6/1/01 9.24 7/15-8/15/03 3.49 10/1-10/31/04 3.33 W Art S 1548.7 5/1-6/6/00 9.10 10/1-10/30/01 9.60 5/15-6/15/03 5.51 8/15-10/10/03 5.51 7/30-10/31/04 6.17 W Art N 967.4 5/1-6/6/00 9.10 10/1-10/30/01 9.60 6/15-8/15/03 3.44 10/10-10/31/03 3.44 6/3-7/30/04 3.85 River Mid W 631.8 5/1-6/6/00 9.10 10/1-10/30/01 9.60 8/15-9/15/03 3.29 5/3-6/3/04 2.52 Point E 1575.9 5/15-7/25/03 4.01 5/6-7/10/04 4.11 Point W 2216.0 5/11-7/18/00 6.62 7/1-8/1/01 6.54 7/25-10/31/03 5.64 7/10-10/31/04 5.79 River W 1306.6 5/11-7/18/00 6.62 7/1-8/1/01 6.54 9/15-10/31/03 6.81 114 Steer N 1326.6 5/15-8/15/03 4.65 5/6-8/6/04 3.65 Steer S 1349.9 8/15-10/31/03 4.74 8/6-10/31/04 3.72 College Green 1023.6 5/15-6/15/03 5.42 9/20-10/31/03 5.42 5/5-6/5/04 5.17 9/15-10/31/04 5.17 College E 908.6 6/15-8/15/03 4.81 7/15-9/25/04 4.59 College W 919.8 8/15-9/20/03 4.87 6/5-7/15/04 4.65 81 E 909.6 5/15-7/15/03 2.71 5/3-7/3/04 2.71 81 W 713.2 8/15-9/15/03 2.13 8/3-9/10/04 2.13 81 Middle 1246.3 7/15-8/15/03 3.72 9/15-10/31/03 3.72 7/3-8/3/04 3.72 9/10-10/31/04 3.72 N Lowe W 1134.1 6/15-8/10/03 4.26 9/15-10/31/03 4.26 5/4-7/15/04 4.46 N Lowe N 906.4 5/15-6/15/03 3.41 9/15-10/31/04 3.57 N Lowe E 652.7 8/10-10/31/03 2.45 7/15-9/15/04 2.57 Stev Cnty 40 64.8 5/15-7/15/03 2.59 5/3-7/3/04 2.59 Stev Cnty river 308.9 7/15-10/31/03 6.18 7/3-10/31/04 5.62 aarea of pasture in hectares bdensity of cattle in each pasture during the specified time frame in hectares per animal unit

115

Figure A.1. Map of pasture boundaries on the Matador Wildlife Management Area.

116

Figure A.2. Map of pasture boundaries on the Gene Howe Wildlife Management Area.

117

Figure A.3. Map of pasture boundaries on the Cimarron National Grasslands.

Figure A.3. Continued

118

Table A.5. Availability of measured vegetative characteristics in grazed and non-grazed pastures from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.841. Grazed Non-grazed Class (%) Proportion Bonferroni interval Proportion Bonferroni interval Vertical Obstruction (overall χ2 = 19.411) 0-9.9* 0.789 0.756 ≤ P ≤ 0.822 0.706 0.675 ≤ P ≤ 0.736 10-19.9 0.070 0.050 ≤ P ≤ 0.091 0.080 0.062 ≤ P ≤ 0.098 20-29.9 0.032 0.018 ≤ P ≤ 0.046 0.062 0.045 ≤ P ≤ 0.078 30-39.9 0.023 0.011 ≤ P ≤ 0.035 0.045 0.031 ≤ P ≤ 0.060 40-49.9 0.023 0.011 ≤ P ≤ 0.035 0.029 0.018 ≤ P ≤ 0.041 50-59.9 0.018 0.008 ≤ P ≤ 0.029 0.016 0.008 ≤ P ≤ 0.025 60-69.9 0.008 0.001 ≤ P ≤ 0.015 0.011 0.004 ≤ P ≤ 0.018 70-79.9 0.006 0.000 ≤ P ≤ 0.012 0.012 0.005 ≤ P ≤ 0.019 80-89.9 0.008 0.001 ≤ P ≤ 0.015 0.010 0.003 ≤ P ≤ 0.016 90-100 0.023 0.011 ≤ P ≤ 0.035 0.029 0.018 ≤ P ≤ 0.041 Horizontal Obstruction (overall χ2 = 6.561) 0-10 0.531 0.491 ≤ P ≤ 0.571 0.484 0.450 ≤ P ≤ 0.517 10.1-20 0.254 0.219 ≤ P ≤ 0.289 0.294 0.264 ≤ P ≤ 0.325 20.1-30 0.113 0.088 ≤ P ≤ 0.139 0.121 0.099 ≤ P ≤ 0.143 30.1-40 0.055 0.037 ≤ P ≤ 0.073 0.058 0.043 ≤ P ≤ 0.074 40.1-50 0.024 0.012 ≤ P ≤ 0.037 0.024 0.014 ≤ P ≤ 0.034 50.1-60 0.011 0.002 ≤ P ≤ 0.019 0.013 0.005 ≤ P ≤ 0.021 >60 0.012 0.003 ≤ P ≤ 0.021 0.005 0.000 ≤ P ≤ 0.010 Grass Cover (overall χ2 = 20.417) 0-9.9* 0.106 0.081 ≤ P ≤ 0.130 0.062 0.045 ≤ P ≤ 0.078 10-19.9 0.113 0.088 ≤ P ≤ 0.139 0.104 0.083 ≤ P ≤ 0.12420-29.9 0.135 0.107 ≤ P ≤ 0.162 0.123 0.101 ≤ P ≤ 0.146 30-39.9 0.168 0.138 ≤ P ≤ 0.198 0.143 0.119 ≤ P ≤ 0.166 40-49.9 0.118 0.092 ≤ P ≤ 0.144 0.145 0.121 ≤ P ≤ 0.169 50-59.9 0.124 0.097 ≤ P ≤ 0.150 0.126 0.103 ≤ P ≤ 0.148 60-69.9 0.109 0.084 ≤ P ≤ 0.134 0.127 0.104 ≤ P ≤ 0.149 70-79.9 0.064 0.045 ≤ P ≤ 0.084 0.093 0.073 ≤ P ≤ 0.113 80-89.9 0.038 0.023 ≤ P ≤ 0.054 0.053 0.038 ≤ P ≤ 0.068 90-100 0.026 0.013 ≤ P ≤ 0.039 0.025 0.014 ≤ P ≤ 0.035

119

Table A.5 continued. Grazed Non-grazed Class (%) Proportion Bonferroni interval Proportion Bonferroni interval Shrub Cover (overall χ2 = 5.407) 0-9.9 0.602 0.563 ≤ P ≤ 0.642 0.593 0.560 ≤ P ≤ 0.626 10-19.9 0.222 0.188 ≤ P ≤ 0.255 0.237 0.208 ≤ P ≤ 0.266 20-29.9 0.101 0.077 ≤ P ≤ 0.125 0.109 0.088 ≤ P ≤ 0.130 30-39.9 0.038 0.023 ≤ P ≤ 0.054 0.030 0.019 ≤ P ≤ 0.042 40-49.9 0.023 0.011 ≤ P ≤ 0.035 0.018 0.009 ≤ P ≤ 0.027 50-59.9 0.003 0.000 ≤ P ≤ 0.007 0.008 0.002 ≤ P ≤ 0.013 >60 0.011 0.002 ≤ P ≤ 0.019 0.004 0.000 ≤ P ≤ 0.009 Bare Ground (overall χ2 = 10.746) 0-9.9 0.385 0.346 ≤ P ≤ 0.424 0.449 0.416 ≤ P ≤ 0.483 10-19.9 0.272 0.236 ≤ P ≤ 0.308 0.264 0.234 ≤ P ≤ 0.294 20-29.9 0.153 0.124 ≤ P ≤ 0.182 0.140 0.116 ≤ P ≤ 0.163 30-39.9 0.102 0.078 ≤ P ≤ 0.127 0.073 0.055 ≤ P ≤ 0.090 40-49.9 0.049 0.032 ≤ P ≤ 0.066 0.036 0.023 ≤ P ≤ 0.048 50-59.9 0.017 0.007 ≤ P ≤ 0.027 0.018 0.009 ≤ P ≤ 0.027 60-69.9 0.012 0.003 ≤ P ≤ 0.021 0.009 0.002 ≤ P ≤ 0.015 70-79.9 0.005 0.000 ≤ P ≤ 0.010 0.005 0.000 ≤ P ≤ 0.010 >80 0.005 0.000 ≤ P ≤ 0.010 0.006 0.001 ≤ P ≤ 0.012 Forb Cover (overall χ2 = 44.113) 0-9.9* 0.327 0.290 ≤ P ≤ 0.365 0.437 0.404 ≤ P ≤ 0.471 10-19.9 0.223 0.190 ≤ P ≤ 0.257 0.253 0.224 ≤ P ≤ 0.283 20-29.9 0.177 0.147 ≤ P ≤ 0.208 0.143 0.119 ≤ P ≤ 0.166 30-39.9* 0.124 0.097 ≤ P ≤ 0.150 0.069 0.052 ≤ P ≤ 0.086 40-49.9 0.067 0.047 ≤ P ≤ 0.087 0.044 0.030 ≤ P ≤ 0.058 50-59.9 0.032 0.018 ≤ P ≤ 0.046 0.035 0.022 ≤ P ≤ 0.047 60-69.9 0.032 0.018 ≤ P ≤ 0.046 0.012 0.005 ≤ P ≤ 0.019 70-79.9 0.009 0.002 ≤ P ≤ 0.017 0.004 0.000 ≤ P ≤ 0.009 >80 0.008 0.001 ≤ P ≤ 0.015 0.002 0.000 ≤ P ≤ 0.005

120

Table A.5 continued. Grazed Non-grazed Class (%) Proportion Bonferroni interval Proportion Bonferroni interval Litter (overall χ2 = 27.158) 0-9.9* 0.596 0.557 ≤ P ≤ 0.636 0.510 0.476 ≤ P ≤ 0.543 10-19.9 0.205 0.173 ≤ P ≤ 0.237 0.252 0.223 ≤ P ≤ 0.281 20-29.9 0.113 0.088 ≤ P ≤ 0.139 0.098 0.078 ≤ P ≤ 0.119 30-39.9* 0.040 0.024 ≤ P ≤ 0.055 0.075 0.057 ≤ P ≤ 0.092 40-49.9 0.028 0.014 ≤ P ≤ 0.041 0.034 0.021 ≤ P ≤ 0.046 50-59.9 0.008 0.001 ≤ P ≤ 0.015 0.022 0.012 ≤ P ≤ 0.031 60-69.9 0.009 0.002 ≤ P ≤ 0.017 0.003 0.000 ≤ P ≤ 0.007 >70 0.002 0.000 ≤ P ≤ 0.005 0.006 0.001 ≤ P ≤ 0.012 * Individual class comparison is different between grazed and non-grazed

121

Table A.6. Availability of measured vegetative characteristics in riparian and upland zones from January 2000 through August 2004 on 3 study areas in the Southern Great Plains (pooled over years and study areas). The overall χ² critical value with α = .05, df = 1: χ² = 3.841. Upland Riparian Class (%) Proportion Bonferroni interval Proportion Bonferroni interval Vertical Obstruction (overall χ2 = 113.682) 0-9.9* 0.855 0.837 ≤ P ≤ 0.872 0.676 0.642 ≤ P ≤ 0.710 10-19.9* 0.040 0.030 ≤ P ≤ 0.050 0.090 0.070 ≤ P ≤ 0.111 20-29.9* 0.026 0.018 ≤ P ≤ 0.034 0.056 0.039 ≤ P ≤ 0.072 30-39.9* 0.019 0.012 ≤ P ≤ 0.026 0.043 0.029 ≤ P ≤ 0.058 40-49.9 0.014 0.008 ≤ P ≤ 0.020 0.032 0.019 ≤ P ≤ 0.045 50-59.9* 0.009 0.004 ≤ P ≤ 0.014 0.026 0.014 ≤ P ≤ 0.037 60-69.9 0.007 0.003 ≤ P ≤ 0.011 0.015 0.006 ≤ P ≤ 0.024 70-79.9* 0.004 0.001 ≤ P ≤ 0.008 0.019 0.009 ≤ P ≤ 0.028 80-89.9 0.006 0.002 ≤ P ≤ 0.009 0.017 0.008 ≤ P ≤ 0.027 90-100 0.020 0.013 ≤ P ≤ 0.028 0.026 0.014 ≤ P ≤ 0.037 Horizontal Obstruction (overall χ2 = 34.587) 0-10* 0.440 0.415 ≤ P ≤ 0.465 0.554 0.518 ≤ P ≤ 0.590 10.1-20 0.278 0.255 ≤ P ≤ 0.301 0.241 0.210 ≤ P ≤ 0.272 20.1-30 0.139 0.122 ≤ P ≤ 0.157 0.109 0.086 ≤ P ≤ 0.131 30.1-40 0.073 0.060 ≤ P ≤ 0.086 0.054 0.038 ≤ P ≤ 0.071 40.1-50 0.037 0.027 ≤ P ≤ 0.047 0.019 0.009 ≤ P ≤ 0.028 50.1-60 0.018 0.011 ≤ P ≤ 0.025 0.012 0.004 ≤ P ≤ 0.020 60.1-70 0.010 0.005 ≤ P ≤ 0.015 0.004 0.000 ≤ P ≤ 0.008 70.1-80 0.002 0.000 ≤ P ≤ 0.005 0.002 0.000 ≤ P ≤ 0.006 >80 0.002 0.000 ≤ P ≤ 0.005 0.005 0.000 ≤ P ≤ 0.010

122

Table A.6. continued. Upland Riparian Class (%) Proportion Bonferroni interval Proportion Bonferroni interval Grass Cover (overall χ2 = 63.146) 0-9.9 0.070 0.057 ≤ P ≤ 0.083 0.074 0.055 ≤ P ≤ 0.093 10-19.9* 0.131 0.114 ≤ P ≤ 0.148 0.083 0.063 ≤ P ≤ 0.103 20-29.9 0.134 0.117 ≤ P ≤ 0.152 0.127 0.103 ≤ P ≤ 0.151 30-39.9 0.160 0.141 ≤ P ≤ 0.178 0.145 0.119 ≤ P ≤ 0.170 40-49.9 0.142 0.125 ≤ P ≤ 0.160 0.126 0.102 ≤ P ≤ 0.150 50-59.9 0.133 0.116 ≤ P ≤ 0.150 0.109 0.086 ≤ P ≤ 0.131 60-69.9 0.119 0.103 ≤ P ≤ 0.135 0.114 0.091 ≤ P ≤ 0.137 70-79.9* 0.064 0.051 ≤ P ≤ 0.076 0.115 0.092 ≤ P ≤ 0.138 80-89.9* 0.032 0.023 ≤ P ≤ 0.041 0.066 0.048 ≤ P ≤ 0.083 90-100* 0.015 0.009 ≤ P ≤ 0.022 0.042 0.028 ≤ P ≤ 0.056 Shrub Cover (overall χ2 = 89.821) 0-9.9* 0.546 0.521 ≤ P ≤ 0.572 0.742 0.710 ≤ P ≤ 0.773 10-19.9* 0.237 0.215 ≤ P ≤ 0.258 0.145 0.119 ≤ P ≤ 0.170 20-29.9* 0.122 0.105 ≤ P ≤ 0.139 0.067 0.049 ≤ P ≤ 0.085 30-39.9 0.044 0.033 ≤ P ≤ 0.054 0.025 0.014 ≤ P ≤ 0.036 40-49.9 0.028 0.020 ≤ P ≤ 0.037 0.012 0.004 ≤ P ≤ 0.020 50-59.9 0.010 0.005 ≤ P ≤ 0.016 0.002 0.000 ≤ P ≤ 0.006 60-69.9 0.007 0.003 ≤ P ≤ 0.012 0.002 0.000 ≤ P ≤ 0.006 >70 0.005 0.001 ≤ P ≤ 0.008 0.005 0.000 ≤ P ≤ 0.010 Bare Ground (overall χ2 = 42.489) 0-9.9* 0.396 0.372 ≤ P ≤ 0.421 0.462 0.426 ≤ P ≤ 0.498 10-19.9* 0.306 0.283 ≤ P ≤ 0.330 0.216 0.187 ≤ P ≤ 0.246 20-29.9 0.155 0.137 ≤ P ≤ 0.174 0.132 0.108 ≤ P ≤ 0.157 30-39.9 0.078 0.064 ≤ P ≤ 0.091 0.093 0.072 ≤ P ≤ 0.114 40-49.9 0.035 0.025 ≤ P ≤ 0.044 0.044 0.030 ≤ P ≤ 0.059 50-59.9 0.017 0.011 ≤ P ≤ 0.024 0.016 0.007 ≤ P ≤ 0.025 60-69.9 0.007 0.003 ≤ P ≤ 0.012 0.020 0.010 ≤ P ≤ 0.030 70-79.9 0.003 0.000 ≤ P ≤ 0.006 0.006 0.001 ≤ P ≤ 0.012 >80 0.002 0.000 ≤ P ≤ 0.004 0.010 0.003 ≤ P ≤ 0.017

123

Table A.6. continued. Upland Riparian Class (%) Proportion Bonferroni interval Proportion Bonferroni interval Forb Cover (overall χ2 = 27.818) 0-9.9* 0.350 0.325 ≤ P ≤ 0.374 0.446 0.410 ≤ P ≤ 0.482 10-19.9 0.254 0.232 ≤ P ≤ 0.276 0.232 0.202 ≤ P ≤ 0.263 20-29.9 0.173 0.153 ≤ P ≤ 0.192 0.143 0.118 ≤ P ≤ 0.169 30-39.9 0.108 0.092 ≤ P ≤ 0.124 0.083 0.063 ≤ P ≤ 0.103 40-49.9 0.060 0.048 ≤ P ≤ 0.072 0.044 0.030 ≤ P ≤ 0.059 50-59.9 0.030 0.021 ≤ P ≤ 0.039 0.022 0.012 ≤ P ≤ 0.033 60-69.9 0.018 0.012 ≤ P ≤ 0.025 0.014 0.005 ≤ P ≤ 0.022 70-79.9 0.004 0.001 ≤ P ≤ 0.008 0.009 0.002 ≤ P ≤ 0.015 >80 0.003 0.000 ≤ P ≤ 0.006 0.006 0.001 ≤ P ≤ 0.012 Litter (overall χ2 = 42.085) 0-9.9 0.491 0.466 ≤ P ≤ 0.517 0.533 0.497 ≤ P ≤ 0.569 10-19.9 0.256 0.234 ≤ P ≤ 0.279 0.211 0.182 ≤ P ≤ 0.241 20-29.9* 0.144 0.126 ≤ P ≤ 0.162 0.099 0.077 ≤ P ≤ 0.120 30-39.9 0.060 0.048 ≤ P ≤ 0.072 0.067 0.049 ≤ P ≤ 0.085 40-49.9 0.031 0.022 ≤ P ≤ 0.040 0.040 0.026 ≤ P ≤ 0.054 50-59.9 0.010 0.005 ≤ P ≤ 0.015 0.022 0.012 ≤ P ≤ 0.033 60-69.9 0.004 0.001 ≤ P ≤ 0.008 0.012 0.004 ≤ P ≤ 0.020 70-79.9 0.001 0.000 ≤ P ≤ 0.002 0.010 0.003 ≤ P ≤ 0.017 >80 0.002 0.000 ≤ P ≤ 0.005 0.006 0.001 ≤ P ≤ 0.012 * Individual class comparison is different between upland and riparian

124

PERMISSION TO COPY

In presenting this thesis in partial fulfillment of the requirements for a master’s

degree at Texas Tech University or Texas Tech University Health Sciences Center, I

agree that the Library and my major department shall make it freely available for research

purposes. Permission to copy this thesis for scholarly purposes may be granted by the

Director of the Library or my major professor. It is understood that any copying or

publication of this thesis for financial gain shall not be allowed without my further

written permission and that any user may be liable for copyright infringement.

Agree (Permission is granted.)

Galon Hall 3 May 2005

________________________________________________ ________________ Student Signature Date Disagree (Permission is not granted.) _______________________________________________ _________________ Student Signature Date