USF Libraries
USF Digital Collections

Determining habitat preferences of the juvenile gopher tortoise (Gopherus polyphemus) using spatially modeled vegetation...

MISSING IMAGE

Material Information

Title:
Determining habitat preferences of the juvenile gopher tortoise (Gopherus polyphemus) using spatially modeled vegetation on a central Florida sandhill
Physical Description:
Book
Creator:
Raymond, Kristan Marie Nicole
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
GIS
HSI
K function
Resource selection function
Forage
Dissertations, Academic -- Biology -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Public and private conservation areas are becoming increasingly important to the continued survival of the gopher tortoise, making it imperative that land managers know the specific habitat requirements of juvenile gopher tortoises because recruitment is key to species persistence. Little is currently known about environmental factors that underlie hatchling and juvenile survival and recruitment in gopher tortoise populations. Because of the short duration and distance of juvenile tortoise foraging journeys, food availability, thermoregulatory conditions, and refugia near the burrow may considerably affect juvenile growth and survival. This two-year study of a central Florida sandhill examines the spatial relationship between juvenile gopher tortoise burrows and the surrounding habitat. Gopher tortoise burrow positions, activity, and width were recorded in four complete surveys of the 4-hectare study area.Coincident with three of the burrow surveys, vegetation and structural habitat characteristics, such as forb and canopy cover, were surveyed in a uniform grid design. Vegetation cover was reclassified using habitat suitability functions (HSFs) derived from qualitative literature values and combined into habitat suitability indices (HSIs) to model the relationships between habitat variables and the likelihood of juvenile gopher tortoise presence. Chi-squared tests and spatial point pattern analysis were used to validate and identify well-forming models. In general, the best performing HSI models for the juvenile gopher tortoise were those that incorporated all three gopher tortoise life requisites in a compensatory relationship (geometric mean): thermoregulation (total high canopy, bare ground, or litter), predation (oak mid-canopy), and food (forb or wiregrass).The models could be improved by using the observed relative abundance of juvenile burrows in each vegetation cover class to modify the HSFs. These methods will help identify habitat characteristics associated with active juvenile gopher tortoise burrows that can be used by public and private land managers to improve existing tortoise habitat and to identify high-quality habitat for future preserves.
Thesis:
Thesis (M.S.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Kristan Marie Nicole Raymond.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 85 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001989300
oclc - 308433785
usfldc doi - E14-SFE0002255
usfldc handle - e14.2255
System ID:
SFS0026573:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200397Ka 4500
controlfield tag 001 001989300
005 20090217112050.0
007 cr mnu|||uuuu
008 090217s2007 flu s 000 0 eng
datafield ind1 8 ind2 024
subfield code a E14-SFE0002255
035
(OCoLC)308433785
040
FHM
c FHM
049
FHMM
090
QH307.2 (Online)
1 100
Raymond, Kristan Marie Nicole.
0 245
Determining habitat preferences of the juvenile gopher tortoise (Gopherus polyphemus) using spatially modeled vegetation on a central Florida sandhill
h [electronic resource] /
by Kristan Marie Nicole Raymond.
260
[Tampa, Fla] :
b University of South Florida,
2007.
500
Title from PDF of title page.
Document formatted into pages; contains 85 pages.
502
Thesis (M.S.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
520
ABSTRACT: Public and private conservation areas are becoming increasingly important to the continued survival of the gopher tortoise, making it imperative that land managers know the specific habitat requirements of juvenile gopher tortoises because recruitment is key to species persistence. Little is currently known about environmental factors that underlie hatchling and juvenile survival and recruitment in gopher tortoise populations. Because of the short duration and distance of juvenile tortoise foraging journeys, food availability, thermoregulatory conditions, and refugia near the burrow may considerably affect juvenile growth and survival. This two-year study of a central Florida sandhill examines the spatial relationship between juvenile gopher tortoise burrows and the surrounding habitat. Gopher tortoise burrow positions, activity, and width were recorded in four complete surveys of the 4-hectare study area.Coincident with three of the burrow surveys, vegetation and structural habitat characteristics, such as forb and canopy cover, were surveyed in a uniform grid design. Vegetation cover was reclassified using habitat suitability functions (HSFs) derived from qualitative literature values and combined into habitat suitability indices (HSIs) to model the relationships between habitat variables and the likelihood of juvenile gopher tortoise presence. Chi-squared tests and spatial point pattern analysis were used to validate and identify well-forming models. In general, the best performing HSI models for the juvenile gopher tortoise were those that incorporated all three gopher tortoise life requisites in a compensatory relationship (geometric mean): thermoregulation (total high canopy, bare ground, or litter), predation (oak mid-canopy), and food (forb or wiregrass).The models could be improved by using the observed relative abundance of juvenile burrows in each vegetation cover class to modify the HSFs. These methods will help identify habitat characteristics associated with active juvenile gopher tortoise burrows that can be used by public and private land managers to improve existing tortoise habitat and to identify high-quality habitat for future preserves.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Co-advisor: Henry Mushinsky, Ph.D.
Co-advisor: Earl McCoy, Ph.D.
653
GIS
HSI
K function
Resource selection function
Forage
690
Dissertations, Academic
z USF
x Biology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2255



PAGE 1

Determining Habitat Preferences of the Juvenile Gop her Tortoise ( Gopherus polyphemus ) Using Spatially Modeled Vegetation on a Central Florida Sandhill by Kristan Marie Nicole Raymond A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Biology College of Arts and Science University of South Florida Co-Major Professor: Henry Mushinsky, Ph.D. Co-Major Professor: Earl McCoy, Ph.D. Susan Bell, Ph.D. Date of Approval: November 13, 2007 Keywords: GIS, HSI, K function, resource selection function, forage Copyright 2007, Kristan Raymond

PAGE 2

Acknowledgements I would like to thank my committee for all the supp ort and encouragement that they have offered me in my research. I would also like to thank the following biology graduate and undergraduate students for fi eld assistance: Brian Halstead, Neal Halstead, Susan Ri edl, Robin Moore, Katie Basiotis, Sarah Smiley, Nej ma Petit, Thayne Holder, and Sharyn. I would also lik e to thank Dr. Steven Reader for his assistance in spatial statistics. I would also like to thank the Univers ity of South Florida Biology Department, Graduate Professional Student Council, and Biological Resear ch Associates for the travel funding they provided to present the results of this thesis. And I especial ly thank Travis Robbins, without whose support and guidance I would never have believed in my project enough to finish it.

PAGE 3

i Table Of Contents List of Figures ii List of Tables iv Abstract v 1.0 INTRODUCTION 1 2.0 METHODS 4 2.1 Study Area 4 2.2 Sampling Procedure 4 2.2.1 Burrow Survey 5 2.2.2 Vegetation Survey 6 2.3 Data Analysis 7 3.0 RESULTS 14 3.1 Burrows 14 3.2 Vegetation 18 3.3 Spatial Modeling 19 3.4 Expected Suitability Versus Observed Preferenc es 37 4.0 DISCUSSION 39 5.0 LITERATURE CITED 43 Appendices 49 Appendix A: Plant data collection fieldsheet 50 Appendix B: HSI probability models created by combi ning individual parameter habitat suitability functions 51 Appendix C: Size distribution of gopher tortoise b urrows by season 53 Appendix D: Burrow activity and stage history 55 Appendix E: Vegetation cover in Braun-Blauquet clas ses and reclassification based on habitat suitability function 62

PAGE 4

ii List Of Figures Figure 1. Aerial photograph of sandhill portion of the University of South Florida EcoArea. 5 Figure 2. Habitat suitability functions for selecte d vegetation parameters. 11 Figure 3. Active and inactive juvenile gopher torto ise burrow locations from Fall 2002 to Spring 2004 in the USF Ecological Research Area. 15 Figure 4. Active and inactive adult gopher tortoise burrow locations from Fall 2002 to Spring 2004 in the USF Ecological Research Area. 16 Figure 5. Abandoned gopher tortoise burrow location s from Fall 2002 to Spring 2004 in the USF Ecological Research Area. 17 Figure 6. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mod el 23. 22 Figure 7. Spring 2003 cross-K function showing spat ial dependence of active and inactive juvenile goph er tortoise burrows and points distributed according t o habitat suitability model 23. 22 Figure 8. Fall 2003 Active and inactive juvenile go pher tortoise burrows and habitat suitability model 23. 23 Figure 9. Fall 2003 cross-K function showing spatia l dependence of active and inactive juvenile gopher tortoise burrows and points distributed according t o habitat suitability model 23. 23 Figure 10. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 23. 24 Figure 11. Spring 2004 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 23. 24 Figure 12. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 87. 25 Figure 13. Spring 2003 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 87. 25 Figure 14. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 87. 26 Figure 15. Fall 2003 cross-K function showing spati al dependence of active and inactive juvenile gophe r tortoise burrows and points distributed according t o habitat suitability model 87. 26 Figure 16. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 87. 27 Figure 17. Spring 2004 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 87. 27

PAGE 5

iii Figure 18. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 47. 28 Figure 19. Spring 2003 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 47. 28 Figure 20. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 49. 29 Figure 21. Fall 2003 cross-K function showing spati al dependence of active and inactive juvenile gophe r tortoise burrows and points distributed according t o habitat suitability model 49. 29 Figure 22. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 49. 30 Figure 23. Spring 2004 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 49. 30 Figure 24. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 66. 31 Figure 25. Spring 2003 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 66. 31 Figure 26. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 66. 32 Figure 27. Fall 2003 cross-K function showing spati al dependence of active and inactive juvenile gophe r tortoise burrows and points distributed according t o habitat suitability model 66. 32 Figure 28. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 66. 33 Figure 29. Spring 2004 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 66. 33 Figure 30. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 55. 34 Figure 31. Spring 2003 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 55. 34 Figure 32. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 55. 35 Figure 33. Fall 2003 cross-K function showing spati al dependence of active and inactive juvenile gophe r tortoise burrows and points distributed according t o habitat suitability model 55. 35 Figure 34. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 55. 36 Figure 35. Spring 2004 cross-K function showing spa tial dependence of active and inactive juvenile gop her tortoise burrows and points distributed according t o habitat suitability model 55. 36 Figure 36. Expected habitat suitability functions a nd standardized average relative abundance of activ e and inactive juvenile gopher tortoise burrows for each vegetation variable. 38

PAGE 6

iv List Of Tables Table 1. Vegetation survey scale used for 2003 surv eys. 7 Table 2. The life requisites and juvenile gopher to rtoise habitat suitability associated with each veg etation parameter. 10 Table 3. Number of gopher tortoise burrows of each life stage, activity, and season. 14 Table 4. Number of active and inactive juvenile gop her tortoise burrows (<=20cm) in each study plot. 1 8 Table 5. Results of Chi-squared test between the ra tio of burrows and random points in suitable versus unsuitable habitat for selected models. 21

PAGE 7

v Determining Habitat Preferences of the Juvenile Gop her Tortoise ( Gopherus polyphemus ) Using Spatially Modeled Vegetation on a Central Flo rida Sandhill Kristan Marie Nicole Raymond ABSTRACT Public and private conservation areas are becoming increasingly important to the continued survival of the gopher tortoise, making it imperati ve that land managers know the specific habitat requirements of juvenile gopher tortoises because r ecruitment is key to species persistence. Little i s currently known about environmental factors that un derlie hatchling and juvenile survival and recruitm ent in gopher tortoise populations. Because of the sho rt duration and distance of juvenile tortoise forag ing journeys, food availability, thermoregulatory condi tions, and refugia near the burrow may considerably affect juvenile growth and survival. This two-year study of a central Florida sandhill examines the s patial relationship between juvenile gopher tortoise burro ws and the surrounding habitat. Gopher tortoise b urrow positions, activity, and width were recorded in fou r complete surveys of the 4-hectare study area. Coincident with three of the burrow surveys, vegeta tion and structural habitat characteristics, such a s forb and canopy cover, were surveyed in a uniform grid d esign. Vegetation cover was reclassified using hab itat suitability functions (HSFs) derived from qualitati ve literature values and combined into habitat suit ability indices (HSIs) to model the relationships between h abitat variables and the likelihood of juvenile gop her tortoise presence. Chi-squared tests and spatial po int pattern analysis were used to validate and iden tify well-forming models. In general, the best performi ng HSI models for the juvenile gopher tortoise were those that incorporated all three gopher tortoise l ife requisites in a compensatory relationship (geom etric mean): thermoregulation (total high canopy, bare gr ound, or litter), predation (oak mid-canopy), and f ood (forb or wiregrass). The models could be improved by using the observed relative abundance of juvenil e

PAGE 8

vi burrows in each vegetation cover class to modify th e HSFs. These methods will help identify habitat characteristics associated with active juvenile gop her tortoise burrows that can be used by public and private land managers to improve existing tortoise habitat and to identify high-quality habitat for fu ture preserves.

PAGE 9

1 1.0 INTRODUCTION Gopherus polyphemus is one of the four North American species of torto ise, and like its congeners, it is in decline throughout its range (A uffenburg and Franz 1982). Although the gopher tor toise is protected to some extent in the southeastern Uni ted States, its preferred xeric habitat is also ver y appealing for human development, including agricult ure, mining, or urbanization (Butler and Sowell 199 6). Many existing tortoise populations are on protected lands, but land management is difficult because no t enough is known about tortoise life history. While some aspects of life history are fairly well studi ed, like home ranges, reproductive characters, and preferred habitat for adults (Germano 1994), information abo ut the factors that underlie hatchling and juvenile su rvival and recruitment in gopher tortoise populatio ns is lacking (Morafka 1994). Realized fecundity in most tortoise populations is very low (as low as 0.5 yo ung per adult female) (Diemer and Moore 1994), making f actors that influence survival of young more important than the effort spent thus far would indi cate (Berish 2001). Sites that are considered suit able gopher tortoise habitat may support a “healthy” pop ulation of adult tortoises, but may show little evi dence of recruitment into the population (Aresco and Guye r 1999b). To manage tortoise habitat for juvenile tortoises, research is needed to identify the speci fic habitat requirements of juvenile gopher tortois es in xeric uplands (Nagy et al. 1997, Wilson et al. 1999, Mushinsky et al. 2003). Because gopher tortoises spend the majority of their early lives within 15 m of their burrows (Wilson 1991, Diemer 1992a, Doona n and Stout 1994), my study is designed to determine how proximal environmental characteristics, such as available forage vegetation, spatially correlate wi th the placement of burrows by young gopher tortois es on a central Florida sandhill. Hatchling tortoises may stay in leaf litter, nest c avities, or adult burrows until the spring of their first year, but most construct their own burrows wi thin 15 m of the nest site (McRae et al. 1981, Butler et al. 1995, Butler and Sowell 1996, Aresco 1999). Juveni le gopher tortoises construct several (1-8) burrows often within 15 m of each other (McRae et al. 1981, Wilson 1991, Diemer 1992a, Wilson et al. 1994, Butler et al. 1995). Burrows of small tortoises are cryptic bec ause of a tendency of juvenile tortoises to

PAGE 10

2 burrow under objects that provide cover from predat ion, like dense shrubs or fallen logs (Alford 1980, Diemer 1992a, Tom 1994, Aresco 1999, Wilson et al. 1999). Hatchling (50 mm carapace length (CL)) an d juvenile (<130 – 150 mm CL) gopher tortoises are pa rticularly vulnerable to predation because of their small size and soft carapace and plastron (Iverson 1980, Wilson 1991, Diemer 1992b, Diemer and Moore 1994, Wilson et al. 1994, Butler and Hull 1996, Aresco 1999). Indeed, mortality rate estimates for young tortoises are variable, but high, ranging from 33% 100% yearly mortality (Landers et al. 1980, Alford 1980, Wright 1982, Wilson 1991, Witz et al. 1992, Diemer and Moore 1994, Germano 1994, Butler et al. 1995, Butler and Sowell 1996). Gopher tortoises forage on a variety of herbaceous plants, including grasses ( Aristida other Poaceae), legumes ( Crotalaria, Galactia, Shrankia, Chamaecristae ), asters ( Pityopsis, Liatris, Phoebanthus, Elephantopus ), and other forbs ( Richardia, Hedytotis, Eriogonum, Dyschoriste, Evolv us, Polygala, etc.) (Garner and Landers 1981, Macdonald and Mushi nsky 1988, Mushinsky et al. 2003). While grasses, especially wiregrass ( Aristida ), constitute the bulk of adult gopher tortoise die t (Wright 1982, Macdonald and Mushinsky 1988), juvenile torto ises are more likely to consume plant species that are low in crude fiber and high in nutrients (Macdonald and Mushinsky 1988). A central Florida study of gopher tortoise scats found that although grasses w ere the most common plants in all scats, young (1-7 years) gopher tortoise scats contained significantl y less grass, relatively fewer plants with external defenses ( Cnidoscolus, Rubus ), and more plants high in nitrogen and calcium (le gumes, Dyschoriste ) than adult gopher tortoise scats (Macdonald and Mushinsky 1988 ). A diet preference study in the same area found that juvenile gopher tortoises ate grasses in propo rtion to their availability, but preferentially sel ected plants with high nutrient (especially nitrogen) and low fi ber content when available. Both central Florida s tudies found that diet preferences were seasonal; juvenile gopher tortoises ate more grasses in the fall and winter when nutritious forbs were not available (Garner an d Landers 1981, Macdonald and Mushinsky 1988, Mushinsky et al. 2003). The juveniles of other tortoise species ha ve also been shown to prefer forbs to grasses (Tom 1994, Nagy et al. 1997, Wilson et al. 1999), strongly suggesting that protein and other nutrients are important to rapid early growth in ju venile tortoises (Landers et al. 1982, Mushinsky et al. 1994, Nagy et al. 1997, Aresco and Guyer 1999b). Herbaceous ground cover of nutritious forbs may be the most important indicator of habitat quality for juv enile gopher tortoises because of its influence on tortoise

PAGE 11

3 growth and survival; in poor quality habitat, juven ile tortoises must move farther from their burrows when foraging, thereby increasing predation risk (Mushin sky et al. 1994, Tom 1994, Wilson et al. 1994, Nagy et al. 1997, Aresco and Guyer 1999a, Aresco 1999, Wilson et al. 1999, Mushinsky et al. 2003). Because of the short duration and distance of juven ile tortoise foraging journeys, food availability near the burrow may considerably affect juvenile gr owth and survival (Wilson 1991, Diemer 1992a, Doonan and Stout 1994, Wilson et al. 1994, Tom 1994, Butler et al. 1995, Wilson et al. 1999). Young G. polyphemus may locate their burrows near nutritious forbs, pe rhaps relocating to another area when preferred vegetation is depleted by herbivory or ch anging seasons (Diemer 1986, Mushinsky et al. 2003). By comparing the spatial distribution of juvenile t ortoise burrows with forb and grass availability, i t may be possible to detect a correlation between burrow pla cement and preferred forage plants. Other considerations, however, may disguise the effect of forage availability on burrow placement; juvenile gopher tortoises may only forage for short periods because of predation pressure or thermoregulation n eeds (Douglass and Layne 1978, Rose and Judd 1982, Wilso n 1991, Tom 1994, Wilson et al. 1999). Therefore, a study of habitat selection for juvenile tortoise should survey other habitat characteristics relatin g to these considerations, including percent canopy cover and percent bare ground. In this study, I developed spatial habitat suitabil ity index (HSI) models of sandhill habitat that can predict the presence of juvenile gopher tortoise bu rrows based on what is known about their burrow characteristics, predation pressures, diet preferen ces, and thermoregulatory requirements. This study presents data from a two-year survey of the spatial distribution of tortoise burrows, vegetation, and other habitat characteristics in a central Florida sandhi ll habitat. The habitat requirements of juvenile gopher tortoises are twofold: 1) juvenile tortoises need n utritious food for growth and 2) predation pressure and thermoregulatory constraints prevent them from fora ging long distances in search of food in low-qualit y habitat. Vegetative characters that represent the se habitat requirements (diet, predation, and thermoregulation) were used in habitat suitability modeling. The habitat suitability models developed in this study are built from the relationships found i n the literature between habitat characteristics an d suitability for juvenile gopher tortoises.

PAGE 12

4 2.0 METHODS 2.1 Study Area The study area is located within the 200 ha Univers ity of South Florida Ecological Research Area (ERA) (82 23’ W, 28 4’ N) in Tampa, Hillsborough County, Florida (Figure 1). Four study plots were selected within the 13 ha upland portion of the ERA a sandhill with well-drained drained yellow sand (Lakeland series) and a xeric vegetative community. The vegetation includes longleaf ( Pinus palustris) and slash pine ( Pinus ellittoi) turkey ( Quercus laevis) and sand live oak ( Quercus geminata ), saw palmetto ( Serona) grasses ( Aristida, Andropogon ), and forbs ( Pityopsis, Liatris, etc.). The sandhill has been divided into plots (approximately 1 ha), each with differen t burn regimes; plots are burned in the summer ever y 1, 2, 5, or 7 years, with the exception of the control plots (Mushinsky 1985, Wilson 1991). My specific study area included four burn-plots from the USF ERA: 2E, 5E, 7E, and CE. The CE or control plot has not burned since 1965. Plot 2E was burned ten times, p lot 5E five times, and plot 7E three times from 197 9 to 2001; plots 2E and 5E were burned in winter 2003. Plots are separated by fire lanes that juvenile gop her tortoises seldom cross. I selected these burn plot s because fire interval in pine-oak sandhill affect s both canopy and ground cover (Mushinsky and Gibson 1991) and affects the density and size distribution of resident gopher tortoises (Mushinsky et al. 2006). 2.2 Sampling Procedure The four study plots were surveyed three times over two years to capture seasonal variation in plant species composition and gopher tortoise activ ity. After an initial burrow survey in Fall 2002, the plots were surveyed for burrows and vegetation in t wo-month periods: April-May 2003, SeptemberOctober 2003 (prior to Winter 2003 burns), and Apri l-May 2004. Locations of each burrow and vegetatio n quadrat were recorded with a Trimble GPS Pathfinder (max accuracy of within 1m).

PAGE 13

5 Figure 1. Aerial photograph of sandhill portion of the University of South Florida EcoArea, showing th e study plots double-outlined. The burn regime of th e study plots (clockwise from upper-left) is about every five years, no burn control, seven years, and two y ears. 2.2.1 Burrow Survey Because gopher tortoises spend only 10% of their li ves outside of their burrows, burrow counts are a more convenient method for estimating tortoise de nsity than direct tortoise counts (Breininger et al. 1991). Although the number, density, and stage of gopher tortoises can only be estimated from burrow surveys, these surveys are still valuable and less timeand resource-intensive than burrow excavation tortoise trapping or tracking, or camera surveys (C arthy et al. 2005). Gopher tortoise size may be estimated from burrow width, which is highly correlated with gopher tortoise carapace length (burrow width is slightly larger) (Alford 1980, Wilson et al. 1991, Doonan and Stout 1994). For each gopher tortoise burrow survey in this stud y, the plots were completely searched for gopher tortoise burrows during warm, sunny days, an d only burrows classified as active or inactive wer e used for spatial analysis (McCoy and Mushinsky 1992 ). Active burrows are those with signs of recent activity (plastron scrape marks or footprints), ina ctive burrows have no signs of activity but can be used without modification, and abandoned burrows must be excavated before being used by tortoises (McCoy and Mushinsky 1992, Mushinsky and McCoy 1994). Aba ndoned burrows are unlikely to be occupied (Witz et al. 1991) and were omitted from the analysis.

PAGE 14

6 The GPS coordinates and width of each gopher tortoi se burrow were used to determine the precise location and approximate size and stage of the goph er tortoise inhabitant (Alford 1980, Wilson et al. 1991, Doonan and Stout 1994). Because burrow width (BW ) is proportional to carapace length (CL), but not at a 1:1 ratio, I considered juvenile gopher tortoise (< 15 cm CL) burrows to be those with a BW < 20 cm at 30-cm depth within the burrow (or 10-cm depth in sm all burrows) (Alford 1980, Wilson 1991, Mushinsky and McCoy 1994, Doonan and Stout 1994, Mushinsky et al. 1994, Wilson et al. 1999). Burrow widths at 30 cm (or 10 cm) were measured using calipers made from two meter sticks bolted together in the center (Wilson et al. 1991). 2.2.2 Vegetation Survey For each survey, vegetation and canopy percent cove r were systematically sampled in 1 m2 quadrats in a grid with 10 m separation. Vegetati on was sampled on a grid, not randomly (Hermann et al. 1992, Stewart et al. 1993) or at burrows (Aresco and Guyer 1999a), to p revent undersampling and biased interpolation (Myers and Shelton 1980, Greenwood 19 96, Griffith 1996, Coker 2000, O’Sullivan et al. 2003). Percent cover was recorded for broad categories of vegetation (grass/forb/woody), as well as for more than 25 plant genera (Kuchler 1967, Aresco and Guyer 1999a) (Appendix A). Many of the plant genera surveyed in the study have been identified a s those that adult and juvenile gopher tortoises ma y select or avoid when foraging (Garner and Landers 1 981, MacDonald and Mushinsky 1988, Mushinsky et al. 2003). Percent cover was recorded on an ordinal s cale, the Braun-Blanquet (Kuchler 1967, Shimwell 1971, Bullock 1996). In the Braun-Blanquet scale, percent cover is combined into 6 classes of non-equ al intervals (Table 1). This method of estimating per cent cover has the advantage of speed and consisten cy (Bullock 1996), but it precludes using some statist ics on the results, including spatial statistics li ke inverse distance weighting, ordinary kriging, and spatial l ogistic regression. High (> 3m) and mid-canopy (0.5 m – 3 m) cover were estimated visually for the 10 m x 10 m area around each vegetation quadrat (Appendix A), a nd the number of species present was noted (Bullock 1996, Spearman et al. 2000, Armitage et al. 2000). Ground cover (< 0.5 m), including bare gro und, litter, and plant cover, was visually estimated for each 1 m2 quadrat. This method for estimating ground cover

PAGE 15

7 was biased toward plants with basal leaves, so the number of individuals of each genus was also record ed (Myers and Shelton 1980, Kaczor and Hartnett 1990, Bullock 1996). Table 1. Vegetation survey scale used for 2003 sur veys. Braun-Blanquet Scale Percent Cover 5 76-100% 4 51-75% 3 26-50% 2 6-25% 1 1-5% 0.5 or + <1% 0 0 2.3 Data Analysis For my data analysis, I had two objectives: 1) to e valuate potential habitat suitability for juvenile gopher tortoises and 2) to validate potential habit at suitability by correlation with observed juvenil e gopher tortoise burrow patterns. To meet these objectives I used literature of gopher tortoise ecology and behavior to create habitat suitability functions (U SFWS 1981) for the juvenile gopher tortoise. I conv erted vegetation data into individual habitat suitability functions and combined indices that were validated using a variety of methods. I reduced the number of vegetation parameters from more than eighty measures of canopy and ground cover (Appendix A) to just nine using Princi pal Components Analysis (PCA) and correlation analysis (Ludwig and Reynolds 1988; STATISTICA 7.1, StatSoft, Inc 2005). Vegetation parameters for each season that were on the first seven components (eigen value > 1) were extracted. Spearman’s nonparametric correlation analysis was used to reduce colinearity; if two vegetation parameters were significantly correlated (p < 0.05), then the param eter with the largest eigen value was selected to b e used in the next step in the model. The following param eters were selected for the model: total high canop y cover (THC), oak midcanopy cover (OMC), bare ground cover (BG), litter cover (LT), wiregrass cover (WG), forb cover (FC), forb genera richness (FG), a ster cover (AST), and legume cover (LEG). The nine vegetation parameters were then interpolat ed from quadrat points to a continuous surface over the study area using the Theissen polygon inte rpolation method and then converted to 1 m2 rasters (ArcGIS 9.x Spatial Analyst, ESRI 2001, O’Sullivan et al. 2003). Theissen, or proximity, polygons are

PAGE 16

8 formed by connecting lines that bisect the midpoint s of neighboring control points. When used for interpolation, the area of the proximity polygon is given the attribute value of its center point. Th e resulting attribute surface is not smooth, however, because adjacent polygons have clearly defined boundaries. The Theissen polygon method has the ad vantage of being appropriate to use on ordinal or nominal data, unlike the more advanced interpolatio n methods (O’Sullivan et al. 2003). The nine vegetation grids for each season were the n reclassified from percent cover to predicted habitat suitability for juvenile gopher tortoises. Reclassification was important because juvenile to rtoise preferences may not be unimodally related to percen t cover. Therefore, a more detailed prediction of habitat preference was inferred using a technique c alled habitat suitability indices (HSI) (USFWS 1981 Store and Kangas 2001) or resource selection functi ons (RSF) (Boyce and MacDonald 1999, Manly et al. 2002, Boyce et al. 2002). Habitat suitability indices (HSI) were ori ginally designed by the United States Fish and Wildlife Service to develop a numerical in dex of the quality of a habitat for a given species These indices may be formed using empirical data (a s in RSF) (Manly et al. 2002), from general inferences drawn from other studies of that organism (Kliskey et al. 1999, Felix et al 2004), or from expert opinion. HSI takes parameters that represent life requisites of a study organism and creates functions that des cribe the linear relationship between those measurable pa rameters and habitat suitability, with habitat suit ability scaled from zero to one for each parameter. The in dividual suitability functions are combined into on e HSI (USFWS 1981). Habitat suitability functions (HSF) were created fo r vegetation parameters that are related to “life requisites” of juvenile gopher tortoises. Life req uisites include food, cover, water, and reproductiv e or other special resources provided by the organism’s habitat (USFWS 1981). Some possible life requisite s for juvenile gopher tortoises based on past researc h include food (FC, WG, FG, LEG, AST), thermoregulation (THC, BG, LIT), and protection fro m predation (OMC, THC, LIT). To reclassify my vegetation parameters into habitat suitability, I r eviewed existing literature on preferred diet and h abitat characteristics of juvenile and adult gopher tortoi ses and created a HSF (Figure 2) for each parameter Using existing literature will result in a model th at is more generally applicable to gopher tortoise habitat (Larson et al. 2003); however, the “preferences” gleaned from th e literature generally will be qualitative, not quantitative. My habitat suitability functions and HSI model were developed for juvenile-stage gop her

PAGE 17

9 tortoises in Central Florida in sandhill or other p ine-dominated, fire-dependent uplands (Table 2, Fig ure 2). I used ArcView 9.x Spatial Analyst Raster Calculato r (ESRI 2001) to convert vegetation cover layers in to habitat suitability layers for each parameter. The individual habitat suitability functions were co mbined into Habitat Suitability Index (HSI) layers based on juvenile gopher tortoise life requi sites and the relationships between those parameter s (Store and Kangas 2001). The USFWS (1981, Larson et al. 2003) defines four types of relationships between life requisites for HSI: limiting, cumulati ve, compensatory, and spatial. For all types of relationships, the final HSI value should be betwee n zero and one. Instead of creating just one HSI to test for the juvenile gopher tortoise, I created a varie ty of HSI models using different combinations of va riables to test for the sensitivity of the weights and rela tionships that I assigned (Store and Kangas 2001). The variables were combined in cumulative (addition), c ompensatory (arithmetic or geometric mean), or more complicated equations that included all types of re lationships. The variables and equations used in e ach model are shown in Appendix B. For each of the ove r 90 HSI models, I used ArcView 9.x Spatial Analyst Raster Calculator (ESRI 2001) to combine vegetation raster layers for each season.

PAGE 18

10 Table 2. The life requisites and juvenile gopher to rtoise habitat suitability associated with each veg etation parameter. Vegetation parameter Life requisite Sources Habitat Suitability THC Thermoregulation, predation Breininger et al. 1994, Boglioli et al. 2000, Hermann et al. 2002, Jones and Dorr 2004 More juvenile tortoises on open/planted pineland th an closed hardwood or agricultural lands. Juveniles more den se on plot > 7 yrs since burn, subadults on plots 0-3 yrs since burn. Burrows had less canopy cover (30%) than random poi nts (70%). Burrow density highest where canopy cover wa s < 65%. OMC Predation Berry and Turner 1986, Stewart et al. 1993, Tom 1994, Breininger et al. 1994, Wilson et al. 1999, Boglioli et al. 2000 Juvenile Bolson and desert tortoises found under ca nopy of woody shrubs or Opuntia instead of in the open. Ha tchling and juvenile density highest in 14% oak shrub, and subadult density highest in 63% shrub. Adult burrows associ ated with low shrub abundance. Overall tortoise density higher in 20% and low in 83% woody shrubs. BG Thermoregulation Douglass and Layne 1978, Stewart et al. 1993, Aresco 1999 Tortoises heat faster on open sand. Juveniles asso ciated with refugia. Overall tortoise density higher in 1.3% 47% bare ground, than 0% bare ground. LIT Thermoregulation, predation Douglass and Layne 1978, Aresco 1999 Tortoises heat faster on open sand. Juveniles asso ciated with refugia. WG Food Douglass and Layne 1978, Garner and Landers 1981, MacDonald and Mushinsky 1988, Stewart et al. 1993, Breininger et al. 1994, Aresco and Guyer 1999a, Mushinsky et al. 2003 Heating rate less on grass than sand. Slightly mor e grass around burrows than around random points. Little g rass in stomachs and scat of juvenile tortoises. Density c orrelated with overall herbaceous groundcover. Juvenile dens ity high in 17.5 45% wiregrass, low in 0% wiregrass. FC, FG, LEG, AST Food Garner and Landers 1981, Lohoefener and Lohmeier 1981, MacDonald and Mushinsky 1988, Stewart et al. 1993, Breininger et al. 1994, Aresco and Guyer 1999a, Mushinsky et al. 2003, Jones and Dorr 2004 More forbs at active burrows than random points. O ften found in juvenile scat and stomachs. Tortoise dens ity high in 30-70% herbaceous cover, low in 24%. Adult and subadult tortoise density high at 71% forbs, medium at 25%, and low at 5%. Highest density found at > 35% herb aceous ground cover. More legumes at active burrows than random points. Juveniles east mostly legumes and broad le aved grasses. Legumes high in nutrients. Firebreak Spatial Tortoise cannot maintain burrow on firebreaks between plots.

PAGE 19

11 Figure 2. Habitat suitability functions for select ed vegetation parameters: a) total high canopy cove r, b) oak midcanopy cover, c) bare ground, d) litter cove r, e) wiregrass cover, f) forb cover (dot) and aster/legume cover (dash). The x-axis represents t he vegetation Braun-Blanquet index value (Appendix) and the y-axis is the reclassified percent suitabil ity value. 0 0.5 1 2 3 4 5 High = 0.99 Mid = 0.50 Low = 0.01 0 0.5 1 2 3 4 5 High = 0.99 Mid = 0.50 Low = 0.01 a b 0 0.5 1 2 3 4 5 High = 0.99 Mid = 0.50 Low = 0.01 0 0.5 1 2 3 4 5 High = 0.99 Mid = 0.50 Low = 0.01 c d 0 0.5 1 2 3 4 5 High = 0.99 Mid = 0.50 Low = 0.01 0 0.5 1 2 3 4 5 High = 0.99 Mid = 0.50 Low = 0.01 e f

PAGE 20

12 The HSI models from Appendix B were validated by co mparing habitat suitability predicted by each model to the juvenile gopher tortoise preferen ce (as indicated by active or active/inactive burro w locations) that I recorded over three seasons. I us ed two methods to choose several well-performing models: chi-squared test of the ratio of suitable t o unsuitable habitat for burrows and completely ran dom points (Basnet et al. 2002, Kolowski and Woolf 2002, Lauver et al. 2002, Manly et al. 2002) and cross Kfunction of burrows and points distributed accordin g to the habitat suitability model (Gibson 2002, O’Sullivan et al. 2003, Reader 2000). A model was considered to perf orm well at predicting suitable habitat if at least one the following were true: th e chi-squared test showed significantly more suitab le habitat at burrows and/or the cross K-function simu lation showed spatial dependence within 33 meters (three times the juvenile gopher tortoise feeding r adius (Wilson et al. 1994)) between burrows and modeled points. The ratio of used versus available habitat suitabi lity (Manly et al. 2002) has been used in resource selection studies for many years. If a HSI model i s valid, one would expect the target species to use areas with high predicted suitability more often than wou ld be predicted by random availability. To use thi s validation method on my data, I divided the suitabi lity results of each HSI model into two categories: suitable habitat (HSI >= 0.50) and unsuitable habit at (HSI < 0.50). For each Active and inactive juven ile gopher tortoise burrow location, as well as fifty r andom locations, I determined the number of burrows and random points within each HSI category for each mod el and used Chi-squared analysis to compare the distribution of used and available habitat suitabil ity (Lauver et al. 2002). My second method of validation was spatial point pa ttern analysis based on the Cross-K function. For this method, I used the HSI layers to simulate burrow patterns based on the predicted habitat suit ability. The GIS extension Hawth’s tools (Beyer 2004) was us ed to generate 50 weighted-random points for each HSI layer. The positioning of these points was wei ghted by the underlying value of the HSI model; ra ster cells with higher suitability values were more like ly to receive a random point than lower valued cell s. I then compared the simulated points from the HSI mod els with the empirical gopher tortoise burrow locations to determine the spatial dependence betwe en the two patterns. To compare spatial patterns of simulated point posi tions and empirical gopher tortoise burrow positions, I used the point pattern analysis method K function (Kaluzny et al. 1998, Reader 2000, Dixon

PAGE 21

13 2002, O’Sullivan et al. 2003, Weigand and Moloney 2004). The spatial K function uses all distances between all pairs of points in a point pattern (Rea der 2000, Dixon 2002, Gibson 2002, O’Sullivan et al. 2003). To calculate the K function, a series of rings of fixed radii are gene rated at each event, and a cumulative count of events within the circles gener ates a cumulative distribution function. The K function may be used to infer the spatial dispersion of a po int pattern, as well as the scale of clustering if the pattern is aggregated. The K function is often plotted as its derived form, the L function, because the L fun ction has a constant variance under complete spatial rand omness and is easy to interpret (Dixon 2002). To compare two point patterns, such as simulated po ints and juvenile gopher tortoise burrow locations, I used a variant of K-function analysis called the CrossK functions of spatial independence (S+ Spatial Statistics, Mathsoft 2000). The Cross -K function uses the between-event, not within-event, distances. To test for significance, 19 Monte Carl o simulations (~ 95% upper CI) were calculated base d on very small random shifts of one pattern on the othe r (Dixon 2002). If the observed CrossK function falls within the extremes of the simulation (simulation e nvelope), then the point patterns are independent. CrossK functions above the extremes indicate attraction b etween events in the two patterns, while functions below the extremes indicate repulsion (Ka luzny et al. 1998, O’Sullivan et al. 2003). If the simulated and empirical burrow patterns are spatial ly dependent at short distances (< 33.0 m, three ti mes the juvenile gopher tortoise feeding radius (Wilson et al. 1994)), then the predictive HSI model from which the simulated points were created is an adequate pr edictor of juvenile gopher tortoise burrow presence After spatial modeling was completed, I compared th e predicted habitat suitability functions from Figure 2 with the observed preferences of juvenile gopher tortoise for each season to determine if the hypothesized relationships derived from the literat ure were similar to those observed at this study si te. I calculated the abundance of juvenile gopher tortois e (A+I) burrows in each percent cover class relativ e to the percent of the study area in that vegetation cl ass. The relative abundances for each variable and percent cover were averaged across the three seasons and st andardized to fall between zero and 1 for each vari able. These standardized average relative abundances were compared to the original habitat suitability funct ions from Figure 2.

PAGE 22

14 3.0 RESULTS 3.1 Burrows Within the study plots, adult gopher tortoise burr ows (burrow width > 20 cm) were twice as abundant as juvenile gopher tortoise burrows (burro w width <= 20 cm) (Table 3). The size distribution of gopher tortoise burrows was bimodal, with peaks at 6 – 16 cm BW and 26 – 36 cm BW (Appendix C). The number of active and inactive juvenile gopher torto ise burrows varied seasonally, with the fewest numb er found in Spring 2003 (26 burrows) and the most foun d in Spring 2004 (47 burrows), after a controlled burn. Adult gopher tortoise burrows followed a sim ilar pattern (51-70 burrows). The ratio of active to inactive gopher tortoise burrows was similar among juveniles for Fall 2002 and Spring 2003, but the number of active burrows increased in Fall 2003 and Spring 2004. The adult gopher tortoise ratio of a ctive to inactive burrows stayed constant from Fall 2002 to Fall 2003, but the number of active burrows almo st doubled in Spring 2004. In general, the number of abandoned burrows found increased over time as more of the active or inactive burrows were abandoned an d new burrows excavated. Over time, many individua l burrows changed either activity status or size clas s (Figures 3 – 5, Appendix D). Table 3. Number of gopher tortoise burrows of each life stage, activity, and season. Activity Season Life stage (<=20 or >20 cm) Active Inactive Abandoned Total (A + I) Juvenile 21 13 34 Fall 2002 Adult 29 32 64 61 Juvenile 15 11 26 Spring 2003 Adult 24 27 128 51 Juvenile 32 9 41 Fall 2003 Adult 29 32 142 61 Juvenile 35 12 47 Spring 2004 Adult 42 28 158 70

PAGE 23

15 Figure 3. Active and inactive juvenile gopher torto ise burrow locations from Fall 2002 to Spring 2004 in the USF Ecological Research Area.

PAGE 24

16 Figure 4. Active and inactive adult gopher tortoise burrow locations from Fall 2002 to Spring 2004 in the USF Ecological Research Area.

PAGE 25

17 Figure 5. Abandoned gopher tortoise burrow location s from Fall 2002 to Spring 2004 in the USF Ecologic al Research Area.

PAGE 26

18 The number and activity status of juvenile gopher t ortoise burrows was influenced by the burn regime f or each experimental area (Table 4, Figure 3). When c orrected for plot size, plots 2E and 5E, which have an open canopy and dense herbaceous ground cover, had two to six times more Active and inactive juvenile burrows/ha in each season than plots 7E and CE. Pl ots 2E and 5E had around the same number of burrows/ha in Fall 2002 and Spring 2003, but plot 2 E had many more burrows/ha in Fall 2003 and Spring 2004. Plots 7E and CE had around the same number o f burrows in every season. Table 4. Number of active and inactive juvenile go pher tortoise burrows (<=20cm) in each study plot (burrows/ha in parentheses). Season Activity 2E (1.0 ha) 5E (1.1 ha) 7E (1.3 ha) CE (0.7 ha) Active 8 (8) 8 (7.3) 4 (3.1) 1 (1.4) Inactive 5 (5) 8 (7.3) 0 (0) 0 (0) Fall 2002 Total (A + I) 13 (13) 16 (14.5) 4 (3.1) 1 (1.4) Active 6 (6) 4 (3.6) 4 (3.1) 1 (1.4) Inactive 3 (3) 6 (5.5) 1 (0.8) 1 (1.4) Spring 2003 Total (A + I) 9 (9) 10 (9.1) 5 (3.8) 2 (2.9) Active 17 (17) 8 (7.3) 4 (3.1) 3 (4.3) Inactive 3 (3) 4 (3.6) 2 (1.5) 0 Fall 2003 Total (A + I) 20 (20) 12 (10.9) 6 (4.6) 3 (4.3) Active 16 (16) 14 (12.7) 2 (1.5) 3 (4.3) Inactive 5 (5) 3 (2.7) 3 (2.3) 1 (1.4) Spring 2004 Total (A + I) 21 (21) 17 (15.5) 5 (3.8) 4 (5.7) 3.2 Vegetation In all seasons, plots 7E and CE generally had high er THC and LIT cover and lower suitability than plots burned more frequently (Appendix E). Oak mid -canopy cover (OMC), BG, and WG are more evenly distributed over the study area, both in percent co ver and habitat suitability (Appendix D). Forb cov er (FC), FG, AST, and LEG were all slightly higher in plots 2E and 5E, both in cover and suitability (Appendix E). The habitat suitability functions (F igure 2) for BG, LIT, FG, and LEG predicted very li ttle highly suitable habitat for most seasons. The habi tat suitability predictions and percent cover of FG and LEG were higher for Fall 2003 than for Spring 2003. The habitat suitability predictions for THC, BG, and FG were higher for Spring 2004 than for Spring and Fall 2003, and FC and AST were higher for Spring 2004 than for Spring 2003.

PAGE 27

19 3.3 Spatial Modeling Four groups of models performed well for all three seasons: models with THC, OMC, BG, LT, WG, FC, and FG (models 22 – 24, 87 – 88), models wi th THC, OMC, BG, and WG (models 47 – 49), models with TMC, OMC, WG, and FC (66), and models w ith THC and FC (models 53 – 55). Spring 2003, with the lowest gopher tortoise burrow sample size and least amount of suitable habitat for several variables, had fewer well-performing models than ot her seasons. Fall 2003 and Spring 2004 had wellperforming models that included legume and aster co ver, unlike Spring 2003. All seasons had wellperforming models that included combinations of can opy variables (THC, OMC) and ground cover variables (FC, WG). Models in the well-performing groups predicted wide ranges of habitat suitability Model 23, the arithmetic mean of THC, OMC, BG, LIT, WG, FC, and FG, predicted that 1 % (Spring 2003) to 21 % (Spring 2004) of the study ar ea would be highly suitable (HSI >= 0.75) for juven ile tortoises, and 39 % (Spring 2003) to 52 % (Spring 2 004) of the study area would be suitable (HSI >= 0. 50). Areas of suitable and unsuitable habitat were patch y and difficult to distinguish in this model (Figur es 6, 8, and 10). This model predicted a large difference i n the percentage of juvenile burrows (63 – 85 %) wi thin suitable habitat (HSI >= 0.5) compared to random po ints (22 – 36 %) (Table 5). Burrow locations and point patterns based on this model were spatially d ependent in each season; at < 7m and 9 – 14 m in S pring 2003, at 2.5 m and 6 – 30 m in Fall 2003, and at < 9 m in Spring 2004 (Figures 7, 9, and 11). Model 87, ((THC2 LIT BG)1/4 (OMC) (WG+FC))1/3, predicted that 4 % (Fall 2003) to 10 % (Spring 2004) of the study area would be highly sui table (HSI >= 0.75) for juvenile tortoises, and 16 % (Spring 2004) to 29 % (Spring 2003) of the study ar ea would be suitable (HSI >= 0.50). Areas of suita ble and unsuitable habitat were more distinct than in m odel 23, making it easier to identify large areas o f low habitat suitability in this model (Figures 12, 14, and 16). This model predicted a large difference i n percentage of juvenile burrows (56 – 66 %) within s uitable habitat (HSI >= 0.5) compared to random poi nts (12 – 18 %), although the percentage of burrows in suitable habitat was lower than for some other mode ls (Table 5). Burrow locations and point patterns bas ed on this model were spatially dependent in each season; at 4 – 5 m and 8 – 30 m in Spring 2003, at 8 – 21 m and 23 – 30 m in Fall 2003, and from 2 – 1 0 m in Spring 2004 (Figures 13, 15, and 17).

PAGE 28

20 Models 47, 48, and 49 (the sum, arithmetic mean, an d geometric mean of THC, OMC, BG, and WG) predicted that 5 % (Fall 2003) to 18 % (Spring 2004) of the study area would be highly suitable (H SI >= 0.75) and 46 % (Fall 2003) to 68 % (Spring 2004) of the study area would be suitable (HSI >= 0.50). Areas of suitable and unsuitable habitat were not v ery distinct, like model 23 (Figures 18, 20, and 22 ). This model predicted a small difference in percentage of juvenile burrows (66 – 89 %) within suitable habit at (HSI >= 0.5) compared to random points (41 – 52 %), although the percentage of burrows in suitable habitat was higher than for some other models (Tabl e 5). Burrow locations and point patterns based on this model group were only intermittently spatially depe ndent in each season; at 7 m, 14 – 18 m and > 23 m in Spring 2003, at 2 – 7 m and 15 – 18 m in Fall 2003, and at 3 – 8 m, 14 m, and 21 – 33 m in Spring 2004 (Figures 19, 21, and 23). Model 66, (THC3 OMC2 *(FC + WG))1/6, predicted that 16 % (Fall 2003) to 32 % (Spring 2 004) of the study area would be highly suitable (HSI >= 0.75) and 48 % (Fall 2003) to 60 % (Spring 2004) of the study area would be suitable (HSI >= 0.50). Areas o f suitable and unsuitable habitat were more distinc t than in models 23 or 47 – 49, making it easy to ide ntify large areas of low habitat suitability in thi s model (Figures 24, 26, and 28). This model predicted a l arge difference in percentage of juvenile burrows ( 69 – 83 %) within suitable habitat (HSI >= 0.5) compared to random points (34 – 46 %) (Table 5). Burrow locations and point patterns based on this model we re spatially dependent in each season; at 15 – 30 m in Spring 2003, at 10 – 11 m and > 24 m in Fall 2003, and at 6 m, 10 m, 15 – 18 m, and 20 – 30 m in Sprin g 2004 (Figures 25, 27, and 29). Models 54 and 55 (arithmetic and geometric means of THC and FC) predicted that 9 % (Spring 2003) to 20 % (Spring 2004) of the study area would be highly suitable (HSI >= 0.75) and 59 % (Fall 20 03) to 69 % (Spring 2004) of the study area would be su itable (HSI >= 0.50). Areas of suitable and unsuita ble habitat were more distinct than in models 23 or 47 – 49, but not as distinct as models 66 and 87 (Figu res 30, 32, and 34). Models 54 and 55 predicted a larg e difference in percentage of juvenile burrows (77 – 94 %) within suitable habitat (HSI >= 0.5) compared to random points (50 – 58 %), although the percentage of random points that fell within suitable habitat was high compared to other models (Table 5). Burrow locations and point patterns based on this model we re spatially dependent in each season; at 2 – 23 m in Spring 2003, at 6 – 14 m in Fall 2003, and at 7 – 1 0 m in Spring 2004 (Figures 31, 33, and 35).

PAGE 29

21 Table 5. Results of Chi-squared test between the ra tio of burrows and random points in suitable (HSI > = 0.50) versus unsuitable (HSI < 0.50) habitat for se lected models (X2 crit = 3.841, df = 1). Season Juvenile A + I burrows Model 23 Model 48 Model 53 Model 66 Model 87 % burrows in suitable habitat 73 77 77 69 58 % random pts in suitable habitat 22 50 50 38 18 Spring 2003 26 X2 18.68 5.13 5.13 6.66 12.47 % burrows in suitable habitat 63 66 78 78 56 % random pts in suitable habitat 26 41 50 34 12 Fall 2003 41 X2 16.67 5.14 7.57 17.59 20.18 % burrows in suitable habitat 85 89 94 83 66 % random pts in suitable habitat 36 52 58 46 12 Spring 2004 47 X2 24.3 16.14 19.52 14.36 29.89

PAGE 30

22 Figure 6. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mod el 23 (white = most suitable, black = least suitable). Figure 7. Spring 2003 Cross K-function showing spat ial dependence of active and inactive juvenile goph er tortoise burrows and points distributed according t o habitat suitability model 23. Spatial Scale (m) Ripley’s L(r)

PAGE 31

23 Figure 8. Fall 2003 Active and inactive juvenile go pher tortoise burrows and habitat suitability model 23 (white = most suitable, black = least suitable). Figure 9. Fall 2003 Cross K-function showing spatia l dependence of active and inactive juvenile gopher tortoise burrows and points distributed according t o habitat suitability model 23. Spatial Scale (m) Ripley’s L(r)

PAGE 32

24 Figure 10. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 23 (white = most suitable, black = least suitable). Figure 11. Spring 2004 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 23. Spatial Scale (m) Ripley’s L(r)

PAGE 33

25 Figure 12. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 87 (white = most suitable, black = least suitable). Figure 13. Spring 2003 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 87. Spatial Scale (m) Ripley’s L(r)

PAGE 34

26 Figure 14. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 87 (white = most suitable, black = least suitable). Figure 15. Fall 2003 Cross K-function showing spati al dependence of active and inactive juvenile gophe r tortoise burrows and points distributed according t o habitat suitability model 87. Spatial Scale (m) Ripley’s L(r)

PAGE 35

27 Figure 16. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 87 (white = most suitable, black = least suitable). Figure 17. Spring 2004 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 87. Spatial Scale (m) Ripley’s L(r)

PAGE 36

28 Figure 18. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 47 (white = most suitable, black = least suitable). Figure 19. Spring 2003 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 47. Spatial Scale (m) Ripley’s L(r)

PAGE 37

29 Figure 20. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 49 (white = most suitable, black = least suitable). Figure 21. Fall 2003 Cross K-function showing spati al dependence of active and inactive juvenile gophe r tortoise burrows and points distributed according t o habitat suitability model 49. Spatial Scale (m) Ripley’s L(r)

PAGE 38

30 Figure 22. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 48 (white = most suitable, black = least suitable). Figure 23. Spring 2004 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 48. Spatial Scale (m) Ripley’s L(r)

PAGE 39

31 Figure 24. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 66 (white = most suitable, black = least suitable). Figure 25. Spring 2003 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 66. Spatial Scale (m) Ripley’s L(r)

PAGE 40

32 Figure 26. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 66 (white = most suitable, black = least suitable). Figure 27.Fall 2003 Cross K-function showing spatia l dependence of active and inactive juvenile gopher tortoise burrows and points distributed according t o habitat suitability model 66. Spatial Scale (m) Ripley’s L(r)

PAGE 41

33 Figure 28. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 66 (white = most suitable, black = least suitable). Figure 29.Spring 2004 Cross K-function showing spat ial dependence of active and inactive juvenile goph er tortoise burrows and points distributed according t o habitat suitability model 66. Spatial Scale (m) Ripley’s L(r)

PAGE 42

34 Figure 30. Spring 2003 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 55 (white = most suitable, black = least suitable). Figure 31.Spring 2003 Cross K-function showing spat ial dependence of active and inactive juvenile goph er tortoise burrows and points distributed according t o habitat suitability model 55. Spatial Scale (m) Ripley’s L(r)

PAGE 43

35 Figure 32. Fall 2003 Active and inactive juvenile g opher tortoise burrows and habitat suitability mode l 55 (white = most suitable, black = least suitable). Figure 33.Fall 2003 Cross K-function showing spatia l dependence of active and inactive juvenile gopher tortoise burrows and points distributed according t o habitat suitability model 55. Spatial Scale (m) Ripley’s L(r)

PAGE 44

36 Figure 34. Spring 2004 Active and inactive juvenile gopher tortoise burrows and habitat suitability mo del 55 (white = most suitable, black = least suitable). Figure 35. Spring 2004 Cross K-function showing spa tial dependence of active and inactive juvenile gopher tortoise burrows and points distributed acco rding to habitat suitability model 55. Spatial Scale (m) Ripley’s L(r)

PAGE 45

37 3.4 Expected Suitability Versus Observed Preferences For some variables, like THC, BG, LIT, and FC, the observed and expected habitat suitability were very similar over the range of percent cover ( Figure 36). Other variables had very different exp ected and observed suitability, like OMC and WG. The obs erved suitability of extreme values of THC was underestimated and mid-range values overestimated b y the THC habitat suitability function, making the actual maximum suitability from 1 – 5 % THC instead of 6 – 25 % THC. For BG, the suitability of mid-to high range values were overestimated, putting the m aximum suitability at 25 – 50 % BG, not 25 – 75 % BG. The observed and expected suitability for LIT were very similar, with only the suitability of > 5 0 % LIT cover underestimated by the habitat suitability function. The predicted and observed suitability of FC (and FG, LEG, and AST) were very similar except at the low range of cover, which was all that was available in the study area, so the habitat suitabi lity function may not have provided enough informat ion about the variation in suitability for low levels o f forb cover. For OMC, the maximum observed suitab ility was from 50 – 75 %, rather than the expected range of 6 – 25 %. The habitat suitability function for W G also underestimated the suitability of the high ran ge of percent cover (> 25 %).

PAGE 46

38 0 0.25 0.5 0.75 1 0255075100 Total High Canopy CoverSuitability 0 0.25 0.5 0.75 1 0255075100 Oak Midcanopy CoverSuitability 0 0.25 0.5 0.75 1 0255075100 Bareground CoverSuitability 0 0.25 0.5 0.75 1 0255075100 Litter CoverSuitability 0 0.25 0.5 0.75 1 0255075100 Wiregrass CoverSuitability 0 0.25 0.5 0.75 1 0255075100 Forb CoverSuitability Figure 36. Expected habitat suitability functions a nd average standardized relative abundance of juven ile gopher tortoise active/inactive burrows by percent cover class for each vegetation variable. (Solid li ne is observed; dashed line is expected.)

PAGE 47

39 4.0 DISCUSSION In general, the best performing habitat suitabilit y indices (HSI) for the juvenile gopher tortoise were those that incorporated all three gopher torto ise life requisites in a compensatory relationship (geometric mean): thermoregulation (THC, BG, or LIT ), predation (OMC), and food (FC or WG). Within each life requisite, the variables in the HSI were sometimes treated as cumulative (food) or compensat ory (thermoregulation). All of the best-performing mod els included THC, which indicates that high canopy cover is a very strong predictor of juvenile gopher tortoise habitat suitability. Models that weighte d THC more heavily showed more distinct delineations betw een suitable and unsuitable habitat. Juvenile torto ises are more likely to be found in areas with low high canopy cover, which have a high percentage of herbaceous cover for forage and sunlight for thermo regulation (Wilson et al. 1994). Most of the models also included OMC, which I related to sheltering fr om predation in juvenile gopher tortoises (Berry an d Turner 1986, Tom 1994). Juvenile tortoises were mo re abundant in areas with either very sparse or mid -tohigh concentrations of young oak trees. All well-p erforming models included the food life requisite a s either or both FC and WG. Juvenile gopher tortoise s are known to prefer nutritious forbs instead of t he adult tortoise diet staple of wiregrass (Garner and Landers 1981, MacDonald and Mushinsky 1988), but areas with a high percent cover of wiregrass may al so have conditions suitable for high forb cover dur ing a different season. The food requisite may be the mo st ephemeral, and the study may not have captured t he key vegetation conditions that led to the choice to excavate burrows in certain areas for that season. Legumes and asters, while considered to be key in t he juvenile gopher tortoise diet (Garner and Lander s 1981, MacDonald and Mushinsky 1988, Mushinsky et al. 2003), may not have been abundant enough in the study area during most seasons to affect the HS I modeling, or the HSFs were not sensitive to the differences in suitability at the lower range of co ver. The last two variables, BG and LIT, were part of some well-performing HSI models, but not all, sugge sting that the potential influence of these variabl es on juvenile gopher tortoise habitat suitability is inc luded in one of the other variables, such as percen t cover of forbs and wiregrass or total high canopy cover.

PAGE 48

40 The HSI models created in this study performed rea sonably well at predicting areas where juvenile gopher tortoises burrowed, but the predictions were less than 90% accurate. Using the observed relationships between juvenile gopher tortoise burr ow locations and vegetation in this study could imp rove the accuracy of the HSI models, but these improveme nts have to be tested in a separate location becaus e the new models would be biased to this study area. The individual HSFs could be improved by adjusting the m to reflect the observed suitability (as represented by the ratio of used to available habitat) for eac h variable (Figure 36). Adjusting the HSF of the few key vari ables (THC, OMC, WG, and FC) could drastically change the suitability prediction of the overall HS I models; if the vegetation classes with the most differences between observed and expected were chan ged, 10 % of the THC HSF, 30 % of the WG HSF, 50 % of the OMC HSF, and 90 % of the FC HSF would c hange suitability values. Some of the differences between observed and expected habitat suitability m ay be because the expected suitability was constrai ned to three values (1, 50, 99) for simplicity, and bec ause the habitat suitability functions covered a wi der range of percent cover than what was available at the stu dy site for some variables. Other potential model improvements are methodologi cal, and would not be biased toward the specific study area. The applications of geostatis tics in ecology are advancing in leaps and bounds. One method of kriging, called indicator kriging, may be able to handle the ordinal index of vegetation dat a and convert it into a smoothed map of percent cover for each variable (Goovaerts 1997). Kriging would res ult in a much more continuous surface for HSF and HSI m odeling than the sharp edges caused by the Theissen polygon method and would improve the prediction suc cess for habitat suitability for burrows that lie o n the edge of the Theissen polygons (about 10 burrows per season). The spatial point pattern validation method could also be improved by using the HSI models to create many simulations of heterogeneous Poisson po int processes and compare the K function of the burrow pattern with the outer limits of the Poisson processes. If the model is at least one adequate explanation of the burrow pattern process, then the burrow K function would be within the simulation envelope of the heterogeneous Poisson process (Weig and and Moloney 2004). In my current method, each model is used to create one simulation of a heterog eneous point process and the simulation envelope is created by toroidal shifts of that pattern. The al ternate method would create many simulations of the

PAGE 49

41 heterogeneous point process for each model, and the extreme K-functions of those simulations would be used as the simulation envelope. The data collection effort for this study was very timeand labor-intensive. The process could be simplified if one could find a way to convert exist ing remote-sensing data into a form that mimics the variables of interest in this study. Unfortunately while THC could easily be estimated by canopy clo sure from satellite imagery, other strata like OMC and g round cover (FC and WG) cannot. It is possible tha t LIDAR (light detection and ranging) imagery, which can provide a 3D representation of forest structure (Lefsky et al. 2001, Lefsky et al. 2002), could help to estimate the vegetation inten sity in each strata using a program like NASA’a SLICER. Unfortunately, LIDAR technology is expensive at this time, and it is unlikely that the species composition could be iden tified from the LIDAR waveforms. Field collection of vegetation characteristics may be the only viable w ay to identify the small-scale variation in the var iables that affect juvenile gopher tortoise. The models developed in this study are limited to u se on xeric communities that have already been identified as potential gopher tortoise habitat and also may not be suited for use in habitats at the edges of the gopher tortoise range where soils and vegetatio n communities are very different. Existing GIS mod els that have identified potential gopher tortoise habi tat in Florida using broad land use and soils infor mation (Cox et al. 1994, McCoy et al. 2002) could be used to select appropriate test sit es for these HSI models. Potential test sites should have existing populatio ns of gopher tortoises with some evidence of juveni le recruitment. Selected xeric areas on the site shou ld be completely surveyed for juvenile gopher torto ise burrows. Field data collection of vegetation varia bles can be simplified by sampling vegetation in a stratified sampling design that samples more intens ely in heterogeneous areas (i.e. forest) than homogeneous areas (i.e. pasture), instead of a syst ematic grid. The number of sampling points needed would depend on the scale at which you wish to dist inguish suitable habitat and the range of variabili ty in the study area. Vegetation variables identified in the test case may need to be modified from the var iables identified in these models, especially if the domin ant shrub species does not have a similar growth pa ttern as oak shrub. With the new gopher tortoise habitat management gu idelines (FWC 2007), more money and time is being spent in finding and managing private miti gation banks. These private landowners, who have

PAGE 50

42 more potential to manage their lands actively, must be provided with the tools and guidance they need to manage their lands for healthy, growing gopher tort oise populations. Therefore, it is important that the mitigation banks include areas that are suitable fo r juvenile gopher tortoises, as well as adults. Th e HSI models from this study, especially those that showe d distinct differences between suitable and unsuita ble habitats, can be used to predict where existing juv enile gopher tortoise burrows may be located or whe re relocated individuals should be released. They can also evaluate the amount of suitable habitat that is available for juvenile gopher tortoises on the site and to identify which portion of sites need addit ional management. The individual HSFs can be used to det ermine which management strategies would be most effective; areas with low percent ground cover may need to be burned or rollerchopped, while those are as with very high total canopy cover may need to be se lectively logged (Berish 2001). The long-term monitoring of gopher tortoise recipient sites propo sed in the new management plan (FWC 2007) calls for field site assessments of the mitigation areas ever y three years. With a reasonable amount of data collection effort on the part of the FWC personnel, private landowners, or their consultants, the meth ods proposed in this study could be used to evaluate th e gopher tortoise recipient areas as part of that t hree-year site assessment. The effort required to collect ha bitat information at these sites would be very rewa rding if it can lead to better management strategies for juv enile gopher tortoise recruitment.

PAGE 51

43 5.0 LITERATURE CITED Alford, R. A. 1980. Population structure of Gopherus polyphemus in Northern Florida. Journal of Herpetology. 14 (2): 177-182. Aresco, M. J. 1999. Habitat structures associated with juvenile gopher tortoise burrows on pine plantations in Alabama. Chelonian Conservation and Biology 3 (3): 507-509. Aresco, M. J. and C. Guyer. 1999a. Burrow abandon ment by gopher tortoises in slash pine plantations of the Conecuh National Forest. Journal of Wildlife Management 63 (1): 26-35. Aresco, M. J. and C. Guyer. 1999b. Growth of the t ortoise Gopherus polyphemus in slash pine plantations of southcentral Alabama. Herpetologica. 55 (4): 499-506. Armitage, R. P., R. E. Weaver, and M. Kent. 2000. Remote sensing of semi-natural upland vegetation: the relationship between species composition and spectr al response. Pages 83-103 in R. Alexander and A. C. Millington (eds), Vegetation Mapping. Jo hn Wiley & Sons, LTD, Chichester, England. Auffenburg, W. and R. Franz. 1982. The status and distribution of the gopher tortoise ( Gopherus polyphemus ). Pages 95-126 in R. B. Bury (ed.), North American Tortoises: Conser vation and Ecology,. U.S. Fish and Wildlife Service, Wildlife Research Report 12. Basnet, B. B., A. A. Apan, and S. R. Raine. 2002. Geographic information system based manure application plan. Journal of Environmental Management 64 : 99-113. Berish, J. E. 2001. Management considerations for t he gopher tortoise in Florida. Florida Fish and Wi ldlife Conservation Commision Final Report, Tallahassee, F lorida. Berry, K. H., and F. B. Turner. 1986. Spring activi ties and habits of juvenile desert tortoises, Gopherus agassizii in California. Copeia 1986 (4):1010-1012. Beyer, H. L. 2004. Hawth's Analysis Tools for ArcGI S. Available at http://www.spatialecology.com/htools Boglioli, M. D., W. K. Michener, and C. Guyer. 200 0. Habitat selection and modification by the gophe r tortoise, Gopherus polyphemus in Georgia longleaf pine forest. Chelonian Conservation and Biology 3 (4): 699-705. Boyce, M. S., and L. L. McDonald. 1999. Relating populations to habitats using resource selection functions. TREE 14 (7): 268-272. Boyce, M. S., P. R. Vernier, S. E. Nielsen, and F. K. A. Schiegelow. 2002. Evaluating resource selec tion functions. Ecological Modelling 157 (2002): 281-300. Breininger, D. R., P. A. Schmalzer, and C. R. Hinkl e. 1991. Estimating occupancy of gopher tortoise ( Gopherus polyphemus ) burrows in coastal scrub and slash pine flatwoods Journal of Herpetology. 25 (3): 317-321.

PAGE 52

44 Breininger, D. R., P. A. Schmalzer, and C. R. Hinkl e. 1994. Gopher tortoise ( Gopherus polyphemus ) densities in coastal scrub and slash pine flatwoods in Florida. Journal of Herpetology 28:60–65. Bullock, J. 1996. Plants. Pages 111-138 in W. J. Sutherland (ed.), Ecological Census Techniqu es: A Handbook. Cambridge University Press. Cambridge, England. Butler, J. A., R. D. Bowman, T. W. Hull, and S. Sow ell. 1995. Movements and home range of hatchling and yearling gopher tortoises, Gopherus polyphemus Chelonian Conservation and Biology 1 (3): 173-180. Butler, J. A. and T. W. Hull. 1996. Reproduction of the tortoise, Gopherus polyphemus in northeastern Florida. Journal of Herpetology. 30 (1): 14-18. Butler, J. A, and S. Sowell. 1996. Survivorship a nd predation of hatchling and yearling gopher torto ises, Gopherus polyphemus Journal of Herpetology. 30 (3): 455-458. Carthy, R. R., M. K. Oli, J. B. Wooding, J. E. Beri sh, and W. D. Meyer. 2005. Analysis of gopher tor toise population estimation techniques. Final Report. P repared for U.S. Army Corps of Engineers. Construction Engineering Research Laboratory. Champ aign, Illinois. Coker, P. D. 2000. Vegetation analysis, mapping a nd environmental relationship at a local scale, Jotunheimen, southern Norway. Pages 135-159 in R. Alexander and A. C. Millington (eds), Vegetation Mapping. John Wiley & Sons, LTD, Chiche ster, England. Cox, J., R. Kautz, M. MacLaughlin, and T. Gilbert. 1994. Closing the gaps in Florida’s wildlife habi tat conservation system. Tallahassee, Florida, Florida Game and Freshwater Fish Commission. 239 pp. Diemer, J. E. 1986. The ecology and management of the gopher tortoise in the southeastern United Sta tes. Herpetologica. 42 (1): 125-133. Diemer, J. E. 1992a. Home range and movements of the tortoise Gopherus polyphemus in northern Florida. Journal of Herpetology. 26 (2): 158-165. Diemer, J. E. 1992b. Demography of the tortoise Gopherus polyphemus in Northern Florida. Journal of Herpetology. 26 (3):281-289. Diemer, J. E. and C. T. Moore. 1994. Reproduction of tortoises in North-central Florida. Pages 129-1 38 in R. B. Bury and D. J. Germano, editors. Biology of North American Tortoises. U. S. Fish an d Wildlife Service, Wildlife Research Report 13. Dixon, P. M. 2002. Ripley’s K function. Encyclopedia of Envirometrics 3 : 1796-1803. Doonan, T. J. and I. J. Stout. 1994. Effects of g opher tortoise ( Gopherus polyphemus ) body size on burrow structure. American Midland Naturalist 131 : 272-280. Douglass, J. F. 1978. Refugia of juvenile gopher tortoises, Gopherus polyphemus (Reptilia, Testudines, Testudinidae). Journal of Herpetology. 12 (3): 412-413. Douglass, J. F. and J. N. Layne. 1978. Activity a nd thermoregulation of the gopher tortoise ( Gopherus polyphemus ) in southern Florida. Herpetologica. 34 (4): 359-374. ESRI. 2001. ArcGIS Statistical Analyst: Statistic al Tools for Data Exploration, Modeling, and Advanc ed Surface Generation. ESRI. Redlands, California. 23 pp.

PAGE 53

45 Felix, A. B., H. Campa III, K. F. Millenbah, S. R. Winterstein, and W. E. Moritz. 2004. Development of landscape-scale habitat-potential models for forest wildlife planning and management. Wildlife Society Bulletin 32 (3): 795-806. Florida Fish and Wildlife Conservation Commision. 2007. Gopher tortoise management plan: Gopherus polyphemus Florida Fish and Wildlife Conservation Commission Tallahassee, Florida. 143 pp. Garner, J. A. and J. L. Landers. 1981. Foods and habitat of the gopher tortoise in southwestern Geo rgia. Proceedings of the Annual Conference Southeast Asso ciation Fish Wildlife Agencies. 35:120134. Germano, D.J. 1994. Comparative life histories of North American Tortoises. Pages 175-185 in R. B. Bury and D. J. Germano, editors. Biology of North American Tortoises. U. S. Fish an d Wildlife Service, Wildlife Research Report 13. Goovaerts, P. 1997. Geostatistics for Natural Reso urces Evaluation. Oxford University Press. New Yo rk, New York. Gibson, D. J. 2002. Methods in comparative plant population ecology. Oxford University Press. Oxfo rd, England. Greenwood, J. J. D. 1996. Basic Techniques. Page s 11-110 in W. J. Sutherland (ed.), Ecological Census Techniques: A Handbook. Cambridge University Press Cambridge, England. Griffith, D. A. 1996. Introduction: the need for spatial statistics. Pages 1-16 in S. L. Arlinghaus (ed), Practical Handbook of Spatial Statistics. CRC Pres s, Boca Raton, Florida. Hermann, S. M., C. Guyer, J. H. Waddle, M. G. Nelms 2002. Sampling on private property to evaluate population status and effects of land use practices on the gopher tortoise, Gopherus polyphemus. Biological Conservation 108 : 289-298. Iverson, J. B. 1980. The reproductive biology of Gopherus polyphemus (Chelonia: Testudinidae). American Midland Naturalist 103 (2): 353-359. Jones, J. C. and Brian Dorr. 2004. Habitat associ ations of gopher tortoise burrows on industrial timberlands. Wildlife Society Bulletin. 32 (2): 456-464. Kaczor, S. A. and D. C. Hartnett. 1990. Gopher to rtoise ( Gopherus polyphemus ) effects on soils and vegetation in a Florida sandhill community. American Midland Naturalist 123 : 100-111. Kaluzny, S. P., S. C. Vega, T. P. Cardoso, and A. A Shelly. 1998. S+ SpatialStats: User’s Manual fo r Windows and UNIX. Mathsoft, INC. Springer, New Yo rk. Kliskey, A. D., E. C. Lofroth, W. A Thompson, S. Br own, and H. Schreier. 1999. Simulating and evaluating alternative resource-sue strategies usin g GIS-based habitat suitability indices. Landscape and Urban Planning 45 (1999): 163-175. Kolowski, J. M. and A. Woolf. 2002. Microhabitat use by bobcats in southern Illinois. Journal of Wildlife Management 66 (3): 822-832. Kuchler, A. W. 1967. Vegetation Mapping. The Ron ald Press Company, New York. Landers, J. L., W. A. McRae, and J. A. Garner. 198 2. Growth and Maturity of the gopher tortoise in southwestern Georgia. Bulletin of the Florida State Museum, Biological Sc iences. 27 (2): 81-110.

PAGE 54

46 Landers, J. L. and D. W. Speake. 1980. Management needs of sandhill reptilies in southern Georgia. Proceedings of Annual Conference of the Southeast A ssociation of Fish and Wildlife Agencies. 34 : 515-529. Larson, M. A., W. D. Dijak, F. R. Thompson, II, and J. J. Millspaugh. 2003. Landscape-level habitat suitability models for twelve wildlife species in s outhern Missouri. Gen. Tech. Rep. NC-233. St. Paul, MN: U. S. Department of Agriculture, Forest Service, North Central Research Station. 51 p. Lauver, C. L W. H. Busby, and J. L. Whistler. 20 02. Testing a GIS model of habitat suitability for a declining grassland bird. Environmental Management 30 (1): 88-97. Lefsky, M. A., W. B. Cohen, D. J. Harding, G. G. Pa rker, S. A. Acker, and S. T. Gower. 2001. Lidar remote sensing of aboveground biomass in three biom es. International Archives of Photogrammetry and Remote Sensing 34 (3/W4): 155-160. Lefsky, M. A., W. B. Cohen, G. G. Parker, and D. J. Harding. 2002. Lidar remote sensing for ecosyste m studies. BioScience 52 (1): 19-30. Lohoefener, R., and L. Lohmeier. 1981. Comparison o f gopher tortoise ( Gopherus polyphemus ) habitats in young slash pine and old longleaf pine areas in sou thern Mississippi. Journal of Herpetology 15:239–242. Ludwig, J. A. and J. F. Reynolds. 1988. Statistic al Ecology: a Primer on Methods and Computing. Joh n Wiley & Sons, New York. Macdonald, L. A. and H. R. Mushinsky. 1988. Foragi ng ecology of the gopher tortoise, Gopherus polyphemus in a sandhill habitat. Herpetologica 44 : 345-353. Manly, B. F. J., L. L. McDonald, D. L. Thomas, T. L McDonald, and W. P. Erickson. 2002. Resource selection by Animals: Statistical Design and Analys is for Field Studies. 2nd ed. Kluwer Academic Publishers. Boston, Massachusetts. 221p. Mathsoft. 2000. S+ Spatial Stats Version 1.5 Supp lement. Data Analysis Products Division. MathSoft Inc., Seattle, Washington. 84 pp. McCoy, E.D. and H. R. Mushinsky. 1992. Studying a species in decline: gopher tortoises and the dilem ma of “correction factors.” Herpetologica. 48 (4): 402-407. McCoy, E. D., B. Stys, and H. R. Mushinsky. 2002. A comparison of GIS and survey estimates of gopher tortoise habitat and numbers of individuals in Flor ida. Chelonian Conservation and Biology 4 (2): 472-478. McRae, W. A., J. L. Landers, and J. A. Garner. 198 1. Movement patterns and home range of the gopher tortoise. American Midland Naturalist 106 (1): 165-179. Morafka, D. J. 1994. Neonates: Missing links in t he life histories of North American tortoises. Pag es 161173 in R. B. Bury and D. J. Germano, editors. Biology of North American Tortoises. U. S. Fish and Wildlife Service, Wildlife Research Report 13. Mushinsky, H. R. 1985. Fire and the Florida sandh ill herpetofaunal community: with special attention to responses of Cnemidophorus sexlineatus Herpetologica. 41 (3): 333-342. Mushinsky, H. R. and D. J. Gibson. 1991. The infl uence of fire periodicity on habitat structure. Pa ges 237-259 in S. S. Bell, E. D. McCoy, and H. R. Mushinsky, edit ors. Habitat structure: the physical arrangement of objects in space. Chapman & Hall, London, UK.

PAGE 55

47 Mushinsky, H. R. and E. D. McCoy. 1994. Compariso n of gopher tortoise populations on islands and on the mainland in Florida. Pages 39-48 in R. B. Bury and D. J. Germano, editors. Biology of North American Tortoises. U. S. Fish and Wildlife Servic e, Wildlife Research Report 13. Mushinsky, H. R., T. A. Stilson, and E. D. McCoy. 2003. Diet and dietary preference of the juvenile gopher tortoise. Herpetologica 59 : 477-485. Mushinsky, H. R., E. D. McCoy, J. E. Berish, R. E. Ashton, Jr., and D. S. Wilson. 2006. Gopherus polyphemus – gopher tortoise. Chelonian Research Monographs. 3 : 350-375. Myers, W. L. and R. L. Shelton. 1980. Survey Meth ods for Ecosystem Management. John Wiley & Sons, INC, New York. Nagy, K. A., D. J. Morafka, and R. A. Yates. 1997. Young desert tortoise survival: energy, water, an d food requirements in the field. Chelonian Conservation and Biology 2 (3): 396-404. O’Sullivan D. and D. J. Unwin. 2003. Geographic I nformation Analysis. John Wiley & Sons, INC. New Jersey. Reader, S. 2000. Using survival analysis to study spatial point patterns in geographical epidemiolog y. Social Science & Medicine 50 (2000): 985-1000. Rose, F. L. and F. W. Judd. 1982. Biology and Sta tus of Berlandier’s Tortoise ( Gopherus berlandieri ). Pages 57-70 in R. B. Bury (ed.), North American Tortoises: Conser vation and Ecology,. U.S. Fish and Wildlife Service, Wildlife Research Report 12. Shimwell, D. W. 1971. The Description and Classif ication of Vegetation. University of Washington Press, Seattle, Washington. Spearman, P. J., R. W. Alexander, and I. D. S. Brod ie. 2000. The use of microscale field mapping in a study of the Ox-eye Daisy ( Leucanthemum vulgare L.) as a component of wild flower meadows. Pages 41-52 in R. Alexander and A. C. Millington (eds), Vegetatio n Mapping. John Wiley & Sons, LTD, Chichester, England. StatSoft, Inc. 2005. STATISTICA (data analysis sof tware system), version 7.1. www.statsoft.com. Stewart, M.C., D. F. Austin, and G.R. Bourne. 1993 Habitat structure and the dispersion of gopher tortoises on a nature preserve. Florida Scientist. 56 (2): 70-81. Store, R. and J. Kangas. 2001. Intergrating spati al multi-criteria evalution and expert knowledge fo r GISbased habitat suitability modeling. Landscape and Urban Planning 55 (2001): 79-93. Tom, J. 1994. Microhabitats and use of burrow of Bolson tortoise hatchlings. Pages 139-146 in R. B. Bury and D. J. Germano, editors. Biology of North A merican Tortoises. U. S. Fish and Wildlife Service, Wildlife Research Report 13. US Fish and Wildlife Service. 1981. Standards for the Development of Habitat Suitability Index Model s (103 ESM). Department of the Interior, Washington, D.C. Wiegand, T. and K. A. Moloney. 2004. Rings, circl es, and null-models for point pattern analysis in ecology. Oikos 104: 209-229. Wilson, D. S. 1991. Estimates of survival for juv enile gopher tortoises, Gopherus polyphemus Journal of Herpetology. 25 (3): 376-379.

PAGE 56

48 Wilson, D. S., H. R. Mushinsky, and E. D. McCoy. 1 991. Relationship between gopher tortoise body siz e and burrow width. Herpetological Review 22 (4): 122-124. Wilson, D. S., H. R. Mushinsky, and E. D. McCoy. 1 994. Home range, activity, and use of burrows of juvenile gopher tortoises in central Florida. Page s 147-160 in R. B. Bury and D. J. Germano, editors. Biology of North American Tortoises. U. S Fish and Wildlife Service, Wildlife Research Report 13. Wilson, D. S., C. R. Tracy, K. A. Nagy, and D. J. M orafka. 1999. Physical and microhabitat characteristics of burrows used by juvenile desert tortoises ( Gopherus agassizii ). Chelonian Conservation and Biology 3 (3): 448-453. Witz, B. S., D. S. Wilson, and M. D. Palmer. 1991. Distribution of Gopherus polyphemus and its vertebrate symbionts in three burrow categories. American Midland Naturalist 126 : 152-158. Witz, B. S., D. S. Wilson, and M. D. Palmer. 1992. Estimating population size and hatchling mo rtality of Gopherus polyphemus Florida Scientist. 55 (1): 14-19. Wright, S. 1982. The distribution and population biology of the gopher tortoise ( Gopherus polyphemus ) in South Carolina. Unpubl. M.S. Thesis, Clemson Unive rsity, Clemson, South Carolina. 74 pp.

PAGE 57

49 Appendices

PAGE 58

50 Appendix A: Plant data collection fieldsheet Vegetation Data Sheets Braun-Blanquet Cover/# genera Date/Time Burn Plot Sample No. WITHIN 10M2 High Canopy Pine Oak Other Mid-story Pine Oak Other WITHIN 1M2 Bare Ground Litter Grass Wiregrass Woody Oak seedling Pine seedling Other seedling Forb Balduina (yellow buttons) Berlandiera (greeen eyes) Carphephorus (paintbrush) Chamaecristae (legume) Cnidoscolus (Tread-softly) Crotalaria (legume) Croton (Silver croton) Dyschoriste (twinflower) Elephantopus Eriogonum (buckwheat) Eupatorium (Dog Fennel) Galactia (legume) Hedyotis (Innocence) Helianthemum (Rock-rose) Hieracium (Hawkweed) Hypericum (Pineweed) Liatris (blazing star) Lupinus (Lupine) Opuntia Pityopsis (Goldenaster) Polygala (milkwort) Ruellia (wild petunia) Scutellaria (Skull-cap) Tephrosia (legume) Vaccinium (Blueberry) Other Legumes Other Asters Braun-Blanquet Coverage 5 76-100% 2 6-25% 4 51-75% 1 1-5% 3 26-50% + <1%

PAGE 59

51 Appendix B: HSI probability models created by combi ning individual parameter habitat suitability functions Model # Combination type THC OMC BG LT WG FC FG LEG AST 13 Addition X X X 14 Arithmetic mean X X X 15 Geometric mean X X X 22 Addition X X X X X X X 23 Arithmetic mean X X X X X X X 24 Geometric mean X X X X X X X 25 Addition X X X X X X 26 Arithmetic mean X X X X X X 27 Geometric mean X X X X X X 28 Addition X X X X 29 Arithmetic mean X X X X 30 Geometric mean X X X X 31 LEG X 32 AST X 33 Addition X X 34 Arithmetic mean X X 35 Geometric mean X X 36 Geometric mean X X X X X X X X X 37 Addition X X X X X X X X 38 Arithmetic mean X X X X X X X X 39 Addition X X X X X X X X X 40 Arithmetic mean X X X X X X X X X 41 Geometric mean X X X X X X X X X 42 Addition X X X X X X X X 43 Geometric mean X X X X X X X X 44 Addition X X 45 Arithmetic mean X X 46 Geometric mean X X 47 Addition X X X X 48 Arithmetic mean X X X X 49 Geometric mean X X X X 50 Addition X X 51 Arithmetic mean X X 52 Geometric mean X X 53 Addition X X 54 Arithmetic mean X X 55 Geometric mean X X 56 FC X 57 THC X 58 FC+WG*BB X X 59 FC+WG+FG*BB X X X 60 WG+LEG+AST*BB X X X 61 FC+FG*BB X X 62 LEG+AST*BB X X 63 OMC*BB X 64 THC*BB X 65 (THC3*LT)1/4*BB X X 66 ((FC+WG)*THC3*OMC2)1/6*BB X X X X 67 ((LEG+AST)*THC3*OMC2)1/6*BB X X X X 68 ((FC+WG)*THC*OMC)1/3*BB X X X X 69 ((LEG+AST)*THC*OMC)1/3*BB X X X X 70 ((FC+WG)*((THC3*LT)1/4)3 *OMC2)1/6*BB X X X X X 71 ((LEG+AST)*((THC3*LT)1/4)3 *OMC2)1/6*BB X X X X X 72 ((FC+WG)*(THC3*LT)1/4 *OMC)1/3*BB X X X X X 73 ((LEG+AST)*(THC3*LT)1/4 *OMC)1/3*BB X X X X X

PAGE 60

52 Model # Combination type THC OMC BG LT WG FC FG LEG AST 74 ((WG+LEG+AST)*THC3 *OMC2)1/6*BB X X X X 75 ((WG+LEG+AST)*((THC3*LT)1/4)3 *OMC2)1/6*BB X X X X X X 76 ((WG+LEG+AST)*THC *OMC)1/3*BB X X X X X 77 ((WG+LEG+AST)*(THC3*LT)1/4 *OMC)1/3*BB X X X X X X 78 ((WG+FC+FG)*THC3 *OMC2)1/6*BB X X X X X 79 ((WG+ FC+FG)*((THC3*LT)1/4)3 *OMC2)1/6*BB X X X X X X 80 ((WG+ FC+FG)*THC *OMC)1/3*BB X X X X X 81 ((WG+ FC+FG)*(THC3*LT)1/4 *OMC)1/3*BB X X X X X X 82 (THC2*LT*BG)1/4*BB X X X 83 ((FC+WG)*((THC2*LT*BG)1/4)3 *OMC2)1/6*BB X X X X X X X 84 ((FC+WG+FG)*((THC2*LT*BG)1/4)3 *OMC2)1/6*BB X X X X X X X 85 ((WG+LEG+AST)*((THC2*LT*BG)1/4) *OMC2)1/6*BB X X X X X X X 86 ((LEG+AST)*((THC2*LT*BG)1/4)3 *OMC2)1/6*BB X X X X X X 87 ((FC+WG)*(THC2*LT*BG)1/4 *OMC)1/3*BB X X X X X X 88 ((FC+WG+FG)*(THC2*LT*BG)1/4 *OMC)1/3*BB X X X X X X X 89 ((WG+LEG+AST)*(THC2*LT*BG)1/4 *OMC)1/3*BB X X X X X X X 90 ((LEG+AST)*(THC2*LT*BG)1/4 *OMC)1/3*BB X X X X X X

PAGE 61

53 Appendix C: Size distribution of gopher tortoise bu rrows by season Fall 2002 0 5 10 15 20 25 0-66-1111-1616-2121-2626-3131-3636-4141-46 Burrow Width (cm)Number of Burrows Active Inactive Figure C-1. Size distribution of active and inactiv e gopher tortoise burrows in Fall 2002. Spring 2003 0 5 10 15 20 25 0-66-1111-1616-2121-2626-3131-3636-4141-46 Burrow Width (cm)Number of Burrows Active Inactive Figure C-2. Size distribution of active and inactiv e gopher tortoise burrows in Spring 2003.

PAGE 62

54 Appendix C: (Continued) Fall 2003 0 5 10 15 20 25 0-66-1111-1616-2121-2626-3131-3636-4141-46 Burrow Width (cm)Number of Burrows Active Inactive Figure C-3. Size distribution of active and inactiv e gopher tortoise burrows in Fall 2003. Spring 2004 0 5 10 15 20 25 0-66-1111-1616-2121-2626-3131-3636-4141-46 Burrow Width (cm)Number of Burrows Active Inactive Figure C-4. Size distribution of active and inactiv e gopher tortoise burrows in Spring 2004.

PAGE 63

55 Appendix D: Burrow activity and stage history Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 2001 Adult Inactive -Abandoned Adult Active Adult Inactive 2002 Adult Inactive Adult Inactive Adult Active -Abandoned 2003 Juvenile Inactive -Abandoned -Abandoned -Abandoned 2004 Adult Active Adult Inactive Adult Active -Ab andoned 2005 Adult Active Adult Active Adult Active Adult A ctive 2006 Adult Inactive -Abandoned Juvenile Inactive Adult Active 2007 -Abandoned -Abandoned Adult Inactive Adult Active 2008 Adult Active Adult Active Juvenile Active Adul t Active 2009 Adult Inactive -Abandoned -Abandoned Adult Active 2010 Adult Inactive -Abandoned -Abandoned Adult Active 2011 Juvenile Active -Abandoned -Abandoned -A bandoned 2012 Adult Inactive -Abandoned -Abandoned -Ab andoned 2013 -Abandoned -Abandoned -Abandoned -Aban doned 2014 Adult Inactive -Abandoned Adult Inactive -Abandoned 2015 Juvenile Active Juvenile Active Juvenile Activ e Juvenile Active 2016 -Abandoned -Abandoned -Abandoned -Aban doned 2017 -Abandoned -Abandoned Juvenile Active Adul t Active 2018 Juvenile Active Juvenile Active Juvenile Activ e Juvenile Active 2019 Adult Active Adult Inactive -Abandoned -Ab andoned 2020 -Abandoned -Abandoned -Abandoned -Aban doned 2021 Adult Active -Abandoned Adult Inactive -Ab andoned 2022 -Abandoned -Abandoned -Abandoned -Aban doned 2023 Adult Inactive -Abandoned -Abandoned -Ab andoned 2024 Juvenile Active Juvenile Active -Abandoned Abandoned 2025 -Abandoned -Abandoned -Abandoned -Aban doned 2026 -Abandoned -Abandoned -Abandoned -Aban doned 2027 Adult Inactive Adult Active Adult Inactive Adu lt Active 2028 Juvenile Active Juvenile Active Juvenile Activ e Juvenile Active 2029 Juvenile Inactive Juvenile Inactive Juvenile A ctive -Abandoned 2030 Juvenile Inactive -Abandoned -Abandoned -Abandoned 2031 Juvenile Inactive -Abandoned -Abandoned -Abandoned 2032 -Abandoned -Abandoned -Abandoned -Aban doned 2033 Adult Inactive Adult Active Adult Inactive -Abandoned 2034 -Abandoned Adult Inactive -Abandoned Adult Active 2035 Juvenile Active Adult Active Adult Inactive -Abandoned 2036 -Abandoned -Abandoned Adult Inactive Adult Inactive 2037 -Abandoned -Abandoned -Abandoned -Aban doned 2038 -Abandoned -Abandoned -Abandoned -Aban doned 2039 -Abandoned -Abandoned -Abandoned -Aban doned 2040 Juvenile Active Juvenile Active -Abandoned Abandoned 2041 Adult Active Adult Inactive Adult Active Adult Active 2042 -Abandoned -Abandoned -Abandoned -Aban doned 2043 Adult Active Adult Inactive Adult Active Adult Active

PAGE 64

56 Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 2044 -Abandoned -Abandoned -Abandoned -Aban doned 2045 -Abandoned -Abandoned Adult Inactive Adult Active 2046 -Abandoned -Abandoned -Abandoned -Aban doned 2047 -Abandoned -Abandoned Adult Active Adult A ctive 2048 -Abandoned -Abandoned Adult Inactive Adult Active 2049 -Abandoned Adult Active Adult Inactive Adult Active 2050 Adult Inactive Adult Active Adult Active Adult Inactive 2051 Juvenile Active Juvenile Inactive -Abandoned Juvenile Inactive 2052 Adult Active Adult Active Adult Active Adult A ctive 2053 -Abandoned -Abandoned Juvenile Inactive -Abandoned 2054 -Abandoned -Abandoned -Abandoned Adult A ctive 2055 -Abandoned -Abandoned Adult Inactive Adult Active 2056 -Abandoned -Abandoned -Abandoned -Aban doned 2057 -Abandoned -Abandoned -Abandoned -Aban doned 2058 Juvenile Inactive -Abandoned -Abandoned -Abandoned 2059 Adult Active -Abandoned -Abandoned -Aban doned 2060 Adult Active -Abandoned Adult Inactive -Ab andoned 2061 Adult Inactive Adult Active -Abandoned 2062 -Abandoned -Abandoned -Abandoned 2063 -Abandoned -Abandoned -Abandoned 2064 Juvenile Active Juvenile Active Juvenile Ina ctive 2065 Juvenile Inactive -Abandoned -Abandoned 2066 -Abandoned -Abandoned -Abandoned 2067 -Abandoned -Abandoned -Abandoned 2068 -Abandoned Juvenile Active Juvenile Active 2069 -Abandoned -Abandoned -Abandoned 2070 -Abandoned -Abandoned -Abandoned 2071 -Abandoned -Abandoned -Abandoned 2072 Adult Active Adult Active 2073 Juvenile Active -Abandoned 2074 Juvenile Active Juvenile Active 2075 Juvenile Active Juvenile Active 2076 Juvenile Active Juvenile Active 2077 Juvenile Active Juvenile Active 2078 Juvenile Active Juvenile Active 2079 Juvenile Active Juvenile Active 2080 Juvenile Active Juvenile Active 2081 Juvenile Inactive Juvenile Inactive 2082 -Abandoned Adult Inactive 2083 -Abandoned -Abandoned 2084 -Abandoned -Abandoned 2085 -Abandoned Adult Active 2086 -Abandoned -Abandoned 2087 -Abandoned -Abandoned 2088 Juvenile Active -Abandoned

PAGE 65

57 Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 2089 -Abandoned 2090 -Abandoned 2091 Juvenile Active 2092 Juvenile Active 2093 Juvenile Active 2094 Juvenile Inactive 2095 Juvenile Inactive 2096 -Abandoned 2097 Juvenile Active 2098 Juvenile Active 2099 Adult Active 2100 Adult Inactive 5001 Juvenile Inactive Juvenile Inactive Juvenile A ctive Juvenile Active 5002 Adult Inactive -Abandoned -Abandoned -Ab andoned 5003 Juvenile Active Adult Active Adult Active Juve nile Active 5004 Adult Active -Abandoned Adult Inactive Adult Active 5005 Adult Active Adult Inactive -Abandoned -Ab andoned 5006 -Abandoned -Abandoned -Abandoned -Aban doned 5007 -Abandoned -Abandoned -Abandoned -Aban doned 5008 Adult Active -Abandoned -Abandoned -Aban doned 5009 Adult Inactive Adult Inactive Adult Active Adu lt Active 5010 Adult Active Adult Active Adult Active Adult I nactive 5011 -Abandoned -Abandoned -Abandoned -Aban doned 5012 -Abandoned -Abandoned -Abandoned -Aban doned 5013 -Abandoned -Abandoned -Abandoned -Aban doned 5014 Juvenile Active -Abandoned -Abandoned -A bandoned 5015 Juvenile Active Juvenile Active Juvenile Activ e Juvenile Inactive 5016 Adult Active Adult Inactive -Abandoned Adult Inactive 5017 Juvenile Inactive -Abandoned -Abandoned -Abandoned 5018 Adult Active Adult Active -Abandoned -Aban doned 5019 Juvenile Active -Abandoned -Abandoned -A bandoned 5020 Adult Active Adult Active Adult Active Adult A ctive 5021 Juvenile Inactive Juvenile Inactive -Abandon ed -Abandoned 5022 -Abandoned -Abandoned -Abandoned -Aban doned 5023 Adult Active Adult Active Adult Active Adult A ctive 5024 Adult Inactive Adult Inactive Adult Inactive Abandoned 5025 Juvenile Inactive -Abandoned -Abandoned -Abandoned 5026 Adult Inactive -Abandoned -Abandoned Adult Inactive 5027 Juvenile Inactive Juvenile Inactive Juvenile I nactive -Abandoned 5028 Adult Inactive -Abandoned -Abandoned -Ab andoned 5029 -Abandoned -Abandoned -Abandoned -Aban doned 5030 Juvenile Inactive -Abandoned -Abandoned -Abandoned 5031 -Abandoned -Abandoned Adult Inactive Adult Inactive 5032 -Abandoned -Abandoned -Abandoned -Aban doned 5033 Adult Inactive Adult Active Adult Active Adult Inactive

PAGE 66

58 Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 5034 Adult Active -Abandoned -Abandoned -Aban doned 5035 Adult Active Adult Inactive -Abandoned Adult Inactive 5036 -Abandoned -Abandoned -Abandoned -Aban doned 5037 Adult Active -Abandoned Adult Active Adult A ctive 5038 Adult Active Adult Inactive Adult Active Adult Inactive 5039 Adult Active Adult Active Adult Active Adult A ctive 5040 -Abandoned Adult Active Adult Inactive Adult Inactive 5041 Adult Active -Abandoned -Abandoned -Aban doned 5042 -Abandoned -Abandoned -Abandoned -Aban doned 5043 -Abandoned -Abandoned Juvenile Inactive -Abandoned 5044 Juvenile Inactive -Abandoned -Abandoned -Abandoned 5045 Adult Active -Abandoned -Abandoned -Aban doned 5046 -Abandoned -Abandoned -Abandoned -Aban doned 5047 -Abandoned -Abandoned -Abandoned -Aban doned 5048 Juvenile Active Juvenile Active Adult Active A dult Active 5049 -Abandoned -Abandoned -Abandoned -Aban doned 5050 Juvenile Active Juvenile Inactive Juvenile Ina ctive Juvenile Inactive 5051 Juvenile Active Adult Inactive Adult Inactive Adult Inactive 5052 -Abandoned -Abandoned -Abandoned -Aban doned 5053 Juvenile Active Juvenile Active Juvenile Ina ctive 5054 Adult Active Adult Inactive Adult Inactive Adu lt Inactive 5055 Juvenile Inactive -Abandoned Adult Active -Abandoned 5056 Adult Active -Abandoned -Abandoned -Aban doned 5057 Adult Active Adult Inactive -Abandoned -Ab andoned 5058 Adult Inactive Adult Inactive -Abandoned -Abandoned 5059 -Abandoned -Abandoned -Abandoned 5060 Juvenile Active Juvenile Inactive -Abandoned -Abandoned 5061 -Abandoned Adult Active Adult Inactive 5062 -Abandoned -Abandoned -Abandoned 5063 -Abandoned -Abandoned -Abandoned 5064 -Abandoned -Abandoned -Abandoned 5065 Adult Active Adult Inactive Adult Active 5066 Adult Inactive -Abandoned -Abandoned 5067 Adult Inactive Adult Inactive -Abandoned 5068 -Abandoned -Abandoned -Abandoned 5069 -Abandoned -Abandoned Juvenile Active 5071 Juvenile Active -Abandoned Adult Active 5072 -Abandoned -Abandoned -Abandoned 5073 -Abandoned -Abandoned -Abandoned 5074 -Abandoned -Abandoned -Abandoned 5075 Juvenile Inactive Juvenile Inactive -Aband oned 5076 -Abandoned -Abandoned -Abandoned 5077 Juvenile Active Adult Active 5078 Juvenile Active Juvenile Active 5079 Juvenile Active -Abandoned

PAGE 67

59 Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 5080 Juvenile Active -Abandoned 5081 Juvenile Active Juvenile Active 5082 Adult Active Adult Inactive Adult Inactive 5083 -Abandoned -Abandoned -Abandoned 5084 Adult Active Adult Active 5085 Juvenile Active 5086 -Abandoned 5087 Adult Active 5088 Adult Inactive 5089 Juvenile Active 5090 Juvenile Active 5091 Juvenile Active 5092 Juvenile Active 5093 Juvenile Active 5094 -Abandoned 5095 Juvenile Active 5096 Juvenile Active 5097 Juvenile Active 7001 Adult Inactive Adult Inactive Adult Inactive A dult Inactive 7002 Juvenile Active Juvenile Inactive -Abandoned -Abandoned 7003 -Abandoned Adult Inactive -Abandoned -Ab andoned 7004 Adult Inactive Adult Inactive Adult Inactive A dult Inactive 7005 Juvenile Active Juvenile Active -Abandoned Abandoned 7006 Adult Inactive Adult Inactive -Abandoned -Abandoned 7007 -Abandoned -Abandoned -Abandoned -Aban doned 7008 -Abandoned -Abandoned -Abandoned -Aban doned 7009 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7010 -Abandoned -Abandoned -Abandoned -Aban doned 7011 Adult Inactive Adult Inactive Adult Active Adu lt Active 7012 -Abandoned -Abandoned -Abandoned -Aban doned 7013 -Abandoned -Abandoned Adult Active Adult A ctive 7014 -Abandoned Adult Inactive -Abandoned Adult Inactive 7015 Adult Inactive Adult Inactive Adult Inactive A dult Inactive 7016 Juvenile Active Juvenile Active Juvenile Inact ive -Abandoned 7017 Adult Inactive Adult Inactive Adult Inactive A dult Inactive 7018 Juvenile Active Juvenile Active Juvenile Activ e Juvenile Active 7019 -Abandoned -Abandoned -Abandoned -Aban doned 7020 -Abandoned -Abandoned -Abandoned -Aban doned

PAGE 68

60 Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 7021 Adult Active Adult Active Juvenile Active Juve nile Inactive 7022 -Abandoned Juvenile Active -Abandoned -A bandoned 7023 -Abandoned -Abandoned -Abandoned -Aban doned 7024 -Abandoned -Abandoned -Abandoned -Aban doned 7025 -Abandoned -Abandoned -Abandoned 7026 -Abandoned -Abandoned -Abandoned 7027 -Abandoned -Abandoned -Abandoned 7028 -Abandoned -Abandoned -Abandoned 7029 -Abandoned -Abandoned 7030 -Abandoned -Abandoned -Abandoned 7031 -Abandoned -Abandoned -Abandoned 7032 -Abandoned -Abandoned -Abandoned 7033 -Abandoned -Abandoned 7034 -Abandoned -Abandoned 7035 -Abandoned -Abandoned -Abandoned 7036 -Abandoned -Abandoned -Abandoned 7037 -Abandoned -Abandoned -Abandoned 7039 Juvenile Active Juvenile Inactive 7040 Juvenile Active Juvenile Active 7041 -Abandoned -Abandoned 7042 Juvenile Inactive Juvenile Inactive 7043 -Abandoned -Abandoned -Abandoned 7044 -Abandoned -Abandoned -Abandoned 7045 -Abandoned -Abandoned 7046 -Abandoned -Abandoned 7047 -Abandoned -Abandoned 7048 -Abandoned -Abandoned 7049 -Abandoned -Abandoned 7050 Adult Active Adult Active Adult Active 7051 Adult Active 7052 Adult Active 9001 -Abandoned -Abandoned -Abandoned -Aban doned 9002 Adult Inactive Adult Active Adult Inactive Adu lt Active 9003 -Abandoned -Abandoned -Abandoned -Aban doned 9004 -Abandoned -Abandoned Adult Inactive -Ab andoned 9005 Adult Inactive Adult Active Adult Active Adult Active 9006 Adult Inactive -Abandoned Adult Inactive Adu lt Active 9007 Adult Inactive Adult Active Adult Inactive Adu lt Inactive 9008 Juvenile Active Juvenile Active Juvenile Activ e Juvenile Inactive 9009 -Abandoned -Abandoned -Abandoned -Aban doned 9010 Adult Inactive -Abandoned Adult Inactive Adu lt Inactive 9011 -Abandoned -Abandoned Adult Active Adult I nactive 9012 -Abandoned -Abandoned -Abandoned Adult A ctive 9013 Adult Inactive -Abandoned Adult Inactive Adu lt Inactive 9014 Adult Inactive -Abandoned -Abandoned -Ab andoned

PAGE 69

61 Burrow Number Fall 2002 Spring 2003 Fall 2003 Spring 2004 Stage Activity Stage Activity Stage Activity Stage Activity 9015 Adult Inactive -Abandoned Adult Inactive Adu lt Inactive 9016 -Abandoned -Abandoned -Abandoned -Aban doned 9017 -Abandoned Adult Active -Abandoned -Aban doned 9018 Juvenile Inactive Juvenile Active Juvenile A ctive 9019 -Abandoned -Abandoned -Abandoned 9020 -Abandoned -Abandoned 9021 Juvenile Active Juvenile Active 9022 -Abandoned -Abandoned 9023 -Abandoned -Abandoned 9024 -Abandoned 9025 -Abandoned 9026 -Abandoned 9027 Juvenile Active

PAGE 70

62 Appendix E: Vegetation cover in Braun-Blauquet clas ses and reclassification based on habitat suitabili ty function Figure E-1. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for total high canopy cover in Spring 2003. (Cover: white = low, black = high; Suitability: black = low, white = high)

PAGE 71

63 Appendix E: (Continued) Figure E-2. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for oak midcanopy cov er in Spring 2003. (Cover: white = low, black = hig h; Suitability: black = low, white = high)

PAGE 72

64 Appendix E: (Continued) Figure E-3. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for bare ground in Sp ring 2003. (Cover: white = low, black = high; Suita bility: black = low, white = high)

PAGE 73

65 Appendix E: (Continued) Figure E-4. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for litter cover in S pring 2003. (Cover: white = low, black = high; Suit ability: black = low, white = high)

PAGE 74

66 Appendix E: (Continued) Figure E-5. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for wiregrass cover i n Spring 2003. (Cover: white = low, black = high; Suitability: black = low, white = high)

PAGE 75

67 Appendix E: (Continued) Figure E-6. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for forb cover in Spr ing 2003. (Cover: white = low, black = high; Suitab ility: black = low, white = high)

PAGE 76

68 Appendix E: (Continued) Figure E-7. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for forb genera in Sp ring 2003. (Cover: white = low, black = high; Suita bility: black = low, white = high)

PAGE 77

69 Appendix E: (Continued) Figure E-8. Vegetation cover after reclassificatio n based on habitat suitability function for aster c over (top) and legume cover (bottom) in Spring 2003. (Cover: w hite = low, black = high; Suitability: black = low, white = high)

PAGE 78

70 Appendix E: (Continued) Figure E-9. Vegetation cover in Braun-Blauquet cla sses (top) and reclassification based on habitat suitability function (bottom) for total high canopy cover in Fall 2003. (Cover: white = low, black = h igh; Suitability: black = low, white = high)

PAGE 79

71 Appendix E: (Continued) Figure E-10. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for oak midcanopy cov er in Fall 2003. (Cover: white = low, black = high; Suitability: black = low, white = high)

PAGE 80

72 Appendix E: (Continued) Figure E-11. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for bare ground in Fa ll 2003. (Cover: white = low, black = high; Suitabi lity: black = low, white = high)

PAGE 81

73 Appendix E: (Continued) Figure E-12. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for litter cover in F all 2003. (Cover: white = low, black = high; Suitab ility: black = low, white = high)

PAGE 82

74 Appendix E: (Continued) Figure E-13. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for wiregrass cover i n Fall 2003. (Cover: white = low, black = high; Suitability: black = low, white = high)

PAGE 83

75 Appendix E: (Continued) Figure E-14. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for forb cover in Fal l 2003. (Cover: white = low, black = high; Suitabil ity: black = low, white = high)

PAGE 84

76 Appendix E: (Continued) Figure E-15. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for forb genera in Fa ll 2003. (Cover: white = low, black = high; Suitabi lity: black = low, white = high)

PAGE 85

77 Appendix E: (Continued) Figure E-16. Vegetation cover after reclassificati on based on habitat suitability function for aster cover (top) and legume cover (bottom) in Fall 2003. (Cove r: white = low, black = high; Suitability: black = low, white = high)

PAGE 86

78 Appendix E: (Continued) Figure E-17. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for total high canopy cover in Spring 2004. (Cover: white = low, black = high; Suitability: black = low, white = high)

PAGE 87

79 Appendix E: (Continued) Figure E-18. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for oak midcanopy cov er in Spring 2004. (Cover: white = low, black = hig h; Suitability: black = low, white = high)

PAGE 88

80 Appendix E: (Continued) Figure E-19. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for bare ground in Sp ring 2004. (Cover: white = low, black = high; Suita bility: black = low, white = high)

PAGE 89

81 Appendix E: (Continued) Figure E-20. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for litter cover in S pring 2004. (Cover: white = low, black = high; Suit ability: black = low, white = high)

PAGE 90

82 Appendix E: (Continued) Figure E-21. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for wiregrass cover i n Spring 2004. (Cover: white = low, black = high; Suitability: black = low, white = high)

PAGE 91

83 Appendix E: (Continued) Figure E-22. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for forb cover in Spr ing 2004. (Cover: white = low, black = high; Suitab ility: black = low, white = high)

PAGE 92

84 Appendix E: (Continued) Figure E-23. Vegetation cover in Braun-Blauquet cl asses (top) and reclassification based on habitat suitability function (bottom) for forb genera in Sp ring 2004. (Cover: white = low, black = high; Suita bility: black = low, white = high)

PAGE 93

85 Appendix E: (Continued) Figure E-24. Vegetation cover after reclassificati on based on habitat suitability function for aster cover (top) and legume cover (bottom) in Spring 2004. (Co ver: white = low, black = high; Suitability: black = low, white = high)