USF Libraries
USF Digital Collections

Geographic variation in life history tactics, adaptive growth rates, and habitat-specific adaptations in phylogeneticall...

MISSING IMAGE

Material Information

Title:
Geographic variation in life history tactics, adaptive growth rates, and habitat-specific adaptations in phylogenetically similar species : the eastern fence lizard,_sceloporus undulatus undulatus_, and the florida scrub lizard, _sceloporus woodi_
Physical Description:
Book
Language:
English
Creator:
Robbins, Travis
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Intrinsic survival
Bergmann's rule
Intrinsic growth
Squamates
Local adaptation
Dissertations, Academic -- Biology-Integrative -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: To understand the evolutionary and ecological significance of geographic variation in life history traits, we must understand whether the patterns are induced through plastic or adaptive responses. The Eastern Fence Lizard, Sceloporus undulatus, exhibits countergradient variation (larger body sizes, et cetera, in northern, cooler environments; presumed adaptive) in life history traits across its large geographic range. However, cogradient variation (the expected result from a plastic response, although not necessarily inconsistent with adaptation) has been suggested as a null hypothesis, especially on fine geographic scales because of relatively small environmental changes. Here we focus on life history variation on a fine geographic scale to test whether cogradient variation is exhibited even though countergradient variation is exhibited at larger scales, and if so, what mechanisms are involved in the switch. We examined north and south populations (~2deg latitude between) of the S. undulatus, and the Florida Scrub Lizard, S. woodi, by measuring adult body sizes, reproduction, and hatchling body sizes over a two year period and conducting reciprocal transplants of juvenile lizards each year. Our results indicate cogradient variation (larger body size in the southern population experiencing a warmer environment) in life history traits of S. undulatus and countergradient variation, a lack of variation in adult body size, in S. woodi along the Florida peninsula. Thus, S. undulatus exhibits cogradient variation at fine geographic scales and countergradient variation at larger scales. Reciprocal transplants revealed that the larger adult body sizes in the southern population of S. undulatus could be explained by longer growth periods allowed by greater intrinsic survival. In S. woodi, the larger than expected adult body sizes in the north could be explained by faster intrinsic and extrinsic juvenile growth rates in the northern population. Because S. undulatus and S. woodi remain distinct species associated with distinct, though adjacent, habitats, we also looked for habitat-specific adaptations. The second reciprocal transplant (between species and habitats) revealed habitat-specific adaptations in juvenile growth rates, but not juvenile survival. Each native species grew faster and had a higher average probability of reaching size at maturity in their native environment than did the foreign species.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
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 Travis Robbins.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X 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:
usfldc doi - E14-SFE0004651
usfldc handle - e14.4651
System ID:
SFS0027966:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Geographic Variation in Life History Tactics, Adapt ive Growth Rates, and Habitatspecific Adaptations in Phylogenetically Similar Sp ecies: The Eastern Fence Lizard, Sceloporus undulatus undulatus, and the Florida Scrub Lizard, Sceloporus woodi by Travis R. Robbins A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Integrative Biology College of Arts and Sciences University of South Florida Co-major Professor: Henry R. Mushinsky, Ph.D. Co-major Professor: Earl D. McCoy, Ph.D. Gordon A. Fox, Ph.D. Gary R. Huxel, Ph.D. Date of Approval: July 14, 2010 Keywords: Intrinsic survival, Local adaptation, Squ amates, Age at maturity, Plasticity Copyright 2010, Travis R. Robbins

PAGE 2

Dedication I dedicate this dissertation to my beautiful wife, Kristan, who has supported me emotionally and financially through the late nights the occasional taxi services, and the nervous breakdowns. You have shown extreme understanding, and I love you with all my being. I am so glad I found you! I could not have done this witho ut you. I would also like to thank my parents for giving m e the opportunities for college, and college, and more college. I have now lived at uni versities almost as long as I lived at home. Both parts of my life were truly wonderful and have shaped me into the multiple personalities we are now. I want to thank my major professors, Henry and Earl for their perseverance, patience, understanding, knowledge, teaching, money, and beer You are both amazing, and I could not imagine a better situation in which to prosper. I w ould also like to thank the amazing faculty of Columbia University’s Biosphere 2 Center, who nurtu red my ecological youth through immersion and experience. Thank you Tony, Annie, Frank, Mich ael, and Rick. And there is one other, who was instrumental in regard to my dissertation resea rch and the army of undergraduates necessary to complete it. Lorelei, thank you for y our dedication to these projects, and more importantly, for being my muse.

PAGE 3

Acknowledgments Many thanks go to the undergraduate army that helpe d me collect these data and care for the lizards in the laboratory, especially Lorel ei Straub. I had many graduate students help me in the field as well, building the enclosures and/o r collecting data in the middle of the Florida summer. For your invaluable assistance, I thank yo u all: Brian Halstead, Neal Halstead, Alison Myers, Dave Karlen, Nick Osman, Jen Rhora, and ever yone else. I thank Tom Raffel for statistical help and Ray Martinez for invaluable en gineering of specialized tools. I also thank my committee members, Gary Huxel and Gordon Fox, for t heir nurturing along the way, and reviewing this manuscript, for it has been much imp roved. This research was partially funded by Sigma Xi Grants-in-Aid of research to TRR. Lizards were collected under collection permit WX05107 issued by the State of Florida Fish and Wil dlife Conservation Commission. All protocols were reviewed and accepted by the USF Ins titutional Animal Care and Use Committee, IACUC file #2778.

PAGE 4

i Table of Contents List of Tables .................................................. ................................................... .......................... iii List of Figures .................................................. ................................................... .......................... v Abstract .................................................. ................................................... ......................... vii Chapter 1: Introducing the Lizards, Life Histories, and Research ............................................ 1 Introduction ...................................... ................................................... ................................ 1 Study Species .................................... ................................................... .............................. 1 Sources of Geographic Variation in Sceloporus Life History Tactics ............................. .... 3 Food availability ............................... ................................................... ................... 4 Thermal environments ............................ ................................................... ............ 5 Mortality rates ................................. ................................................... ..................... 5 Research Outline ................................. ................................................... ............................ 6 References ....................................... ................................................... ................................ 8 Chapter 2: Variation in Life History Tactics on a F ine Geographic Scale and Along a Temperature Gradient Elucidates the Cogradient to C ountergradient Switch in Sceloporus Lizards .................................................. ................................................... .......... 12 Abstract .......................................... ................................................... ................................ 12 Key words ....................................... ................................................... .................. 13 Introduction ..................................... ................................................... ............................... 13 Materials and Methods ............................ ................................................... ....................... 15 Study species ................................... ................................................... ................. 15 Collection and housing of adult female lizards .. .................................................. 16 Reproduction .................................... ................................................... ................. 17 Hatchlings ...................................... ................................................... ................... 18 Environmental variables ......................... ................................................... ........... 18 Results ......................................... ................................................... ................................. 19 Adult female lizards ............................ ................................................... ............... 19 Reproduction .................................... ................................................... ................. 20 Hatchlings ...................................... ................................................... ................... 20 Environmental variables ......................... ................................................... ........... 21 Discussion ....................................... ................................................... ............................... 21 Conclusions ..................................... ................................................... .................. 24 Acknowledgments .................................. ................................................... ........................ 25 References ....................................... ................................................... .............................. 26 Chapter 3: On Intrinsic Growth and Juvenile Surviva l of Lizard Populations Along a Fine Scale Temperature Gradient: a Reciprocal Trans plant Approach ........................... 45 Abstract .......................................... ................................................... ................................ 45 Key words ....................................... ................................................... .................. 46 Introduction ..................................... ................................................... ............................... 46 Materials and Methods ............................ ................................................... ....................... 50 Study species ................................... ................................................... ................. 50

PAGE 5

ii Collection and housing of female lizards ........ ................................................... .. 51 Egg incubation and hatchling husbandry .......... ................................................... 52 Reciprocal transplants .......................... ................................................... ............ 52 Environmental covariates ........................ ................................................... .......... 53 Data analyses ................................... ................................................... ................ 55 Results .......................................... ................................................... ................................ 58 Sceloporus undulatus .............................. ................................................... ......... 58 Sceloporus woodi ................................ ................................................... .............. 60 Summary of overall trends ......................... ................................................... ....... 62 Discussion ....................................... ................................................... ............................... 62 Reaction norms and effects of environmental varia bles in Sceloporus undulatus ......................................... ................................................... ........... 63 Reaction norms and effects of environmental variabl es in Sceloporus woodi ............................................. ................................................... ............. 65 Large scale versus small scale trends in adult body sizes .................................. 66 Conclusions ....................................... ................................................... ................ 67 Acknowledgements ................................. ................................................... ....................... 68 References ....................................... ................................................... .............................. 69 Chapter 4: Habitat-specific Adaptations in Growth R ates Play a Role in Species Distribution of Sceloporus Lizards in Florida .................................................. .................. 90 Abstract .......................................... ................................................... ................................ 90 Key words ....................................... ................................................... .................. 91 Introduction ..................................... ................................................... ............................... 91 Methods .......................................... ................................................... ............................... 93 Collection and housing of female lizards ........ ................................................... .. 93 Egg incubation and hatchling husbandry .......... ................................................... 94 Reciprocal transplants .......................... ................................................... ............ 94 Environmental covariates ........................ ................................................... .......... 96 Data analyses ................................... ................................................... ................ 98 Results ......................................... ................................................... ............................... 100 Discussion ....................................... ................................................... ............................. 103 Acknowledgements ................................. ................................................... ..................... 105 References ....................................... ................................................... ............................ 106 About the Author .................................................. ................................................... ......... End Page

PAGE 6

iii List of Tables Table 2.1. Effects of species, latitude, and year on adult fe male body sizes ................................ 40 Table 2.2. Effects of species, latitude, and year on lizard r eproduction ....................................... 41 Table 2.3. Effects of species, latitude, and year on hatchlin g body sizes ..................................... 42 Table 2.4. Results of correlation analyses between average en vironmental temperatures and average trait values .......................... ................................................... .................. 43 Table 2.5. Mean trait values for Sceloporus undulatus and S. woodi from the north and south populations ................................. ................................................... ..................... 44 Table 3.1. Effects of source population on morphological trai ts at time of release into the field ............................................ ................................................... .............................. 79 Table 3.2 ANOVA results for growth rates among resident pop ulations and reciprocally transplanted hatchlings of Sceloporus undulatus .............................. .......................... 80 Table 3.3. Results of post-hoc ANOVAs on population-specific (North and South) growth rates among growth environments for Sceloporus undulatus .............................. ........ 81 Table 3.4. Relationships between environmental variables, gro wth, and survival across and within species ................................ ................................................... ..................... 82 Table 3.5. Survival models including source population (pop) and growth environment (env) for resident and reciprocally transplanted ha tchlings of Sceloporus undulatus in warmer, southern and cooler, northern habitats ..................................... 83 Table 3.6. Survival models including only covariates of sourc e population and growth environment for resident and reciprocally transplan ted hatchlings of Sceloporus undulatus in warmer, southern and cooler, northern habitats ................. 84 Table 3.7. Relationships among population-environment treatme nts with regard to environmental variables ........................... ................................................... ................. 85 Table 3.8. Results of ANOVAs for growth rates among resident populations and reciprocally transplanted hatchlings of Sceloporus woodi .................................. ......... 86 Table 3.9. Results of post-hoc ANOVAs on population-specific (North and South) growth rates among growth environments for Sceloporus woodi .................................. .......... 87 Table 3.10. Survival models including source population (pop) and growth environment (env) for resident and reciprocally transplanted ha tchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats ......................................... 88

PAGE 7

iv Table 3.11. Survival models including only covariates of source population and growth environment for resident and reciprocally transplan ted hatchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats ...................... 89 Table 4.1. Results of ANOVA for juvenile growth rates of resi dent and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ............. ................................................... .......... 121 Table 4.2. Relationships between environmental variables, gro wth, and survival across all treatments and within species ................. ................................................... ........... 122 Table 4.3. Survival models including only factors of resident and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ............................... ................................................... ................... 123 Table 4.4. Results of model averaging across the candidate mo del set of the factor only survival models ................................... ................................................... .................... 124 Table 4.5. Survival models including only covariates of resid ent and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ............. ................................................... .......... 125 Table 4.6. Results of model averaging across the candidate mo del set of the covariate only survival models .............................. ................................................... .................. 126 Table 4.7. Population specific life history data showing the numbers used to calculate the habitat-specific probabilities of reaching size at maturity for Sceloporus undulatus and S. woodi .......................................... ................................................... 127 Table 4.8. Environmental variables among latitude-habitat tre atments ...................................... 128

PAGE 8

v List of Figures Figure 2.1 Map of study site locations ..................... ................................................... .................. 34 Figure 2.2. Species x latitude interaction in adult female SVL .................................................. .... 35 Figure 2.3. Species x latitude interaction in total clutch mas s ................................................. ..... 36 Figure 2.4. Species x latitude interaction in hatchling tail l ength ............................................. ..... 37 Figure 2.5 Food availability for north and south populations of each lizard species ................... 38 Figure 2.6. Cogradient variation in snout-vent length of adult female Sceloporus undulatus along the Florida peninsula ...................... .................................................. 39 Figure 3.1. Map of reciprocal transplant locations ........... ................................................... .......... 74 Figure 3.2. Growth rates of resident and reciprocally transpla nted hatchlings of S. undulatus in warmer, southern and cooler, northern habitats ................................... 75 Figure 3.3. Survivorship of resident and reciprocally transpla nted hatchlings of Sceloporus undulatus in warmer, southern and cooler, northern habitats ................ 76 Figure 3.4. Growth rates of resident and reciprocally transpla nted hatchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats ....................... 77 Figure 3.5. Survivorship of resident and reciprocally transpla nted hatchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats ....................... 78 Figure 4.1. Map of reciprocal transplant locations for habitat -specific adaptations study .......... 115 Figure 4.2. Daily growth rates of resident and reciprocally tr ansplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ......................................... ................................................... ........................ 116 Figure 4.3. Monthly juvenile survivorship of resident and reci procally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ............................... ................................................... ................. 117 Figure 4.4. Time in months until maturity is reached for resid ent and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ............. ................................................... ........ 118 Figure 4.5. Probability of reaching size at maturity of reside nt and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes ............. ................................................... ........ 119

PAGE 9

vi Figure 4.6. Average probability of reaching size at maturity o f resident and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between sandhill and scrub habitats ....................... ................................................... ............. 120

PAGE 10

vii Geographic Variation in Life History Tactics, Adapt ive Growth Rates, and Habitatspecific Adaptations in Phylogenetically Similar Sp ecies: The Eastern Fence Lizard, Sceloporus undulatus undulatus, and the Florida Scrub Lizard, Sceloporus woodi Travis R. Robbins Abstract To understand the evolutionary and ecological signi ficance of geographic variation in life history traits, we must understand whether the patt erns are induced through plastic or adaptive responses. The Eastern Fence Lizard, Sceloporus undulatus exhibits countergradient variation (larger body sizes, et cetera, in northern, cooler environments; presumed adaptive) in life history traits across its large geographic range. However, cogradient variation (the expected result from a plastic response, although not necessarily incons istent with adaptation) has been suggested as a null hypothesis, especially on fine geographic sc ales because of relatively small environmental changes. Here we focus on life history variation o n a fine geographic scale to test whether cogradient variation is exhibited even though count ergradient variation is exhibited at larger scales, and if so, what mechanisms are involved in the switch. We examined north and south populations (~2 latitude between) of the S. undulatus and the Florida Scrub Lizard, S. woodi by measuring adult body sizes, reproduction, and hatch ling body sizes over a two year period and conducting reciprocal transplants of juvenile lizar ds each year. Our results indicate cogradient variation (larger body size in the southern populat ion experiencing a warmer environment) in life history traits of S. undulatus and countergradient variation, a lack of variation in adult body size, in S. woodi along the Florida peninsula. Thus, S. undulatus exhibits cogradient variation at fine geographic scales and countergradient variation at larger scales. Reciprocal transplants revealed that the larger adult body sizes in the southern po pulation of S. undulatus could be explained by

PAGE 11

viii longer growth periods allowed by greater intrinsic survival. In S. woodi, the larger than expected adult body sizes in the north could be explained by faster intrinsic and extrinsic juvenile growth rates in the northern population. Because S. undulatus and S. woodi remain distinct species associated with distinct, though adjacent, habitats we also looked for habitat-specific adaptations. The second reciprocal transplant (bet ween species and habitats) revealed habitatspecific adaptations in juvenile growth rates, but not juvenile survival. Each native species grew faster and had a higher average probability of reac hing size at maturity in their native environment than did the foreign species.

PAGE 12

1 Chapter 1: Introducing the Lizards, Life Histories, and Resear ch Introduction My dissertation research examines observed geograph ic variation in life history tactics of two lizard species, the Eastern Fence Lizard, Sceloporus undulatus and the Florida Scrub Lizard, S. woodi on a relatively fine geographic scale. Two popul ations of each species were compared. The populations within each species are separated b y approximately 2 latitude, which corresponds with a 1 C difference between the nort h and south environmental temperatures experienced. Two years of observed life history da ta were collected. The first year, reciprocal transplants of juvenile lizards were also conducted within species to differentiate between plastic and adaptive influence on juvenile growth rates and survival. The second year, reciprocal transplants of juvenile lizards were conducted betw een species to examine species specific plasticity in growth rates and differences in intri nsic juvenile survival with regard to species specific habitats. Environmental variables were al so measured to examine relationships between habitats and juvenile growth rates and habitats and juvenile survival. In this context I can separate population-specific reasons for intrinsic and extrinsic juvenile growth rates and survival and examine their relationships with other populati on-specific life history tactics. Study Species Sceloporus is well suited for studies of sources of variation in life history traits because members of this genus are relatively abundant where found, easy to care for in the laboratory, and easy to mark and re-capture in the field. A la rge body of knowledge exists for this genus, which makes a strong comparative base for results a nd experimental techniques. Although they are different species, S. woodi and S. undulatus are model organisms for studying plastic and

PAGE 13

2 adaptive sources of variation in life history tacti cs because they have important similarities in higher order factors that significantly affect life histories, such as mode of reproduction (ovipary vs. vivipary) and foraging mode (sit-and-wait vs. a ctive foraging) (Dunham and Miles 1985; Huey and Pianka 1981). Both species are oviparous and s it-and-wait predators. Sceloporus woodi is a phrynosomatid spiny lizard that is precinctiv e to the sand-pine scrub habitat in Florida. It ranges from Ocala Nat ional Forest in the north to Highlands Co. near Archbold Biological Station in the south, restricte d mostly to the central ridges of Florida. Populations also exist on the central and southern Atlantic coast and the southwestern Gulf coast (Jackson 1973a). Genetic variation among S. woodi populations is quite high. In fact, when mitochondrial DNA of S. woodi was analyzed for 135 samples from 16 patches on 5 major ridges in Florida, analysis of molecular variance (AMOVA) showed an estimated 10.4% total genetic variation within patches, 17.5% among patches (with in ridges), and 72.1% among ridges. These data suggest that some populations of S. woodi have persisted in isolation for time frames in excess of 1 Myr (Clark et al. 1999). Sceloporus undulatus also a phrynosomatid spiny lizard, has been exten sively studied because of its commonness and life history variatio n throughout its large range. S. undulatus is found from Lat 24 N to 40 N and spans from the ea st coast of the United States westward to southwest Utah (Tinkle and Ballinger 1972). There are currently four species groups throughout this range (Leache 2009) with extreme variation in life history tactics, making generalizations of this species difficult. The habitats in which indi viduals of the S. undulatus complex can be found vary tremendously and include abandoned buildings o n old farms, cleared forest areas, mixed deciduous forests, the pine barrens of New Jersey, mesquite and juniper grasslands, and the sage and juniper canyonlands of the Colorado Platea u (Tinkle and Ballinger 1972). The species group in Florida is Sceloporus undulatus (previously Sceloporus undulatus undulatus ; Wiens and Reeder 1997) and is found mostly in the longleaf-pi ne/turkey oak habitats ranging from the panhandle in the north down to central Florida, nor th of the Everglades. Little is known about S. undulatus in Florida because of the paucity of studies.

PAGE 14

3 In one recently constructed phylogeny of the genus Sceloporus, using molecular and morphological evidence, S. woodi was placed closer in relation to S. undulatus undulatus than five of the six other undulatus subspecies (Wiens and Reeder 1997). Allozyme data on the Sceloporus complex supports the findings of Wiens and Reeder (1997) and sheds more light on the phylogentic relationships in the southeast U. S The allozyme data suggest a closer relationship between S. undulatus undulatus in South Carolina and S. woodi than S. u. undulatus in Florida and S. woodi (Miles et al. 2002). In another more recent phylo geny, S. woodi was placed in the eastern clade (one of the four previo usly mentioned species groups) with S. undulatus (Leache 2009; Leache and Reeder 2002). The life h istory and genetic variation within the S. undulatus complex, and the previously discussed genetic varia tion within S. woodi, appear to be as much or more than the variation between th e two species. Although data suggest that Sceloporus undulatus and S. woodi have been distinct for a million years, where the sand-pine scrub and dry pi ne/oak forests are adjacent these two species hybridize (Jackson 1972; Robbins et al. 2010). Gut analyses also show that prey selection of each species is similar in taxa and proportion of e ach taxon (Jackson 1973b). The most conspicuous factor separating the species is a habi tat preference. Sources of Geographic Variation in Sceloporus Life History Tactics Geographic variation in life history tactics of Sceloporus species has been examined for trends in rand K-selection, bet-hedging (Stearns 1976; Tinkle and Dunham 1986), habitat type, and phylogenetics, all of which explain some level of variation, but ultimately emphasize the importance of local environmental conditions (Fergu son and Talent 1993; Jones and Ballinger 1987; Niewiarowski and Roosenburg 1993; Tinkle and Ballinger 1972). Sceloporus undulatus is a model organism for studying life history variatio n because the suites of life history tactics vary greatly throughout its large geographic range (Ferg uson et al. 1980; Ferguson and Talent 1993; Niewiarowski and Roosenburg 1993; Smith 1998; Stear ns 1976; Tinkle and Ballinger 1972; Tinkle and Dunham 1986). In general, life history tactics follow the climatic gradient adjusted for survivorships at the scale of the entire geographic range. In colder, northern habitats, adult

PAGE 15

4 female body size is larger, instrinsic growth rates are faster but growth still occurs slower over longer lifespans, and clutch mass is greater per bo ut when compared to populations from warmer habitats. In warmer, southern habitats, adult fema le body size is smaller, intrinsic growth rates are slower but growth still occurs faster and more consistently over the lifespan, and clutch mass is less per bout but more bouts occur compared to p opulations from cooler habitats (Angilletta et al. 2004a; Ferguson and Talent 1993; Niewiarowski a nd Roosenburg 1993; Sears and Angilletta 2004; Tinkle and Ballinger 1972). Life history tactics are generally constrained by n et assimilated energy. The plasticity that occurs in life history traits among different environments can also be governed by populationspecific energy allocation rules (Sinervo and Adolp h 1994). For example, lizard populations from environments with relatively short daily activity p eriods may not allocate more energy to growth even when experiencing longer potential activity pe riods (Sinervo and Adolph 1994). These rules allocate the net assimilated energy into growth, st orage, maintenance, and/or reproduction (Congdon et al. 1982; Dunham et al. 1989). The env ironment influences net assimilated energy through factors like food availability, and rates o f energy use through factors like operative temperatures. Thus, food is energy input and time spent at certain temperatures determines the rates of energy use and amounts of energy used. Su rvival is also a major component of fitness. Food availability. Food availability, which can constrain the amount o f energy assimilated, can directly influence growth, storage and reproduction. Growth rates often increase as food availability increases (Ballinger 1977; Ballinger and Congdon 1980; Dunham 1978; Smith 1998). For example, a 50% reduction in food caused a reduction in growth rates equal to that of a 50% reduction in daily activity times (Sinervo and Adolph 1994). Supplemental feeding experiments, however, have shown no effect on growth rate (Jones et al. 1987a; Niewiarowski 1995), but the experiments were done i n the field where the saturation point in either growth rate or assimilation efficiency may h ave been reached prior to the food supplementation. Greater food availability also in creases lipid storage and the size of the first clutch of the season. Lipids stored are an importa nt energy source for production of the first

PAGE 16

5 clutch and have also been linked with survival thro ugh overwintering success (Ballinger 1977; Derickson 1976). Thermal environments. Periods of daily activity and seasonal activity are generally determined for ectotherms by environmental temperat ures, therefore constraining metabolism through rates and durations. For instance, the len gth of the activity season correlates with number of clutches per season in Sceloporus species allowing them to be reproductive opportunists with variable numbers of clutches (Jon es and Ballinger 1987; Jones et al. 1987b). The activity season and the daily activity times ha ve a positive relationship with growth rates as well (Ballinger et al. 1981; Dunham 1978; Grant and Dunham 1990; Niewiarowski and Roosenburg 1993; Smith 1998; Tinkle 1972; Tinkle an d Ballinger 1972). Furthermore, in S. graciosus and S. occidentalis, potential growth rates (norms of reaction) plateaue d when the experimental daily activity time approached the nat ural activity time specific to each population (Sinervo and Adolph 1994). Thermal regimes have also been modelled to predict growth rates and subsequently age and size at maturity. Generally, annual activity t imes are negatively correlated with annual survival rate and positively correlated with annual reproduction. Longer activity seasons, presumably, increase predation risk but allow great er numbers of clutches (Adolph and Porter 1993). The consistency between experimental manipu lations and ecological modelling suggests a strong relationship between thermal physical envi ronments and the expression of life history traits. Mortality rates. Mortality regimes play a strong role in life histor y theory. Juvenile mortality rates have been shown to positively corre late with growth rate (Jones and Ballinger 1987; Vinegar 1975), and the norm of reaction for a ge and size at maturity has been successfully modelled using the differences between juvenile and adult mortality regimes (Stearns and Koella 1986). Life span and age at maturity, which are ma jor components of lifetime fitness by

PAGE 17

6 constraining time allocated to reproduction, have b een shown to positively correlate when other environmental variables are similar (Tinkle and Bal linger 1972). Research Outline One cannot examine all of the complexities involved in the manifestation of phenotypic variation. However, examining populations on a rel atively fine geographic scale by measuring environmental variables known to influence life his tory tactics, observing realized variation in life history tactics, and differentiating between intrin sic and extrinsic causes of some of these traits, will elucidate some mechanisms through which phenot ypic variation manifests. My dissertation begins with comparing life history tactics between Florida populations of Sceloporus lizards, S. undulatus and S. woodi and ends with reciprocal transplants to compare p lasticity found within species to plasticity, caused by each other’s habit ats, found between species. This dissertation research is the first to directly compare plasticit y in growth within and between habitat specific, allopatric species. The observed life history diff erences among these populations in their respective habitats and latitudes will give a compa rative baseline for the results of the reciprocal transplant experiments. These studies provide the first life history data on a S. undulatus population this far south in Florida. The reciproc al transplants will allow me to differentiate between population-specific and environmental sourc es of the observed differences in life history tactics. For instance, if differences exist betwee n the species, can food availability, thermal environments, and mortality rates explain them? Wh at are the relative intrinsic and extrinsic contributions to the observed life history differen ces? Will S. undulatus in Florida have life history tactics that are more similar to S. woodi than tactics exhibited by S. undulatus in other northern populations because of the large-scale similarities in their Florida habitats? Subsequent chapters will answer some of these questions. Chapter 2 exa mines the observed, realized life history tactics among populations of S. undulatus and S. woodi along a temperature gradient. Chapter 3 examines the relative population-specific and envir onmental influence on juvenile growth rates and survival through reciprocal transplants between the north and south populations within species. And finally, chapter 4 examines the relat ive population-specific and environmental

PAGE 18

7 influence on juvenile growth rates and survival thr ough reciprocal transplants between speciesspecific habitats.

PAGE 19

8 References Adolph, S. C., and W. P. Porter. 1993. Temperature, activity, and lizard life histories. American Naturalist 142:273-295. Angilletta, M. J., P. H. Niewiarowski, A. E. Dunham A. D. Leache, and W. P. Porter. 2004. Bergmann's clines in ectotherms: Illustrating a lif e-history perspective with sceloporine lizards. American Naturalist 164:E168-E183. Ballinger, R. E. 1977. Reproductive strategies fo od availability as a source of proximal variation in a lizard. Ecology 58:628-635. Ballinger, R. E., and J. D. Congdon. 1980. Food res ource limitation of body growth rates in Sceloporus scalaris (Sauria, Iguanidae). Copeia 198 0:921-923. Ballinger, R. E., D. L. Droge, and S. M. Jones. 198 1. Reproduction in a Nebraska sandhills population of the northern prarie lizard Scelopours undulatus garmani American Midland Naturalist 106:157-164. Clark, A. M., B. W. Bowen, and L. C. Branch. 1999. Effects of natural habitat fragmentation on an endemic scrub lizard ( Sceloporus woodi ): an historical perspective based on a mitochondrial DNA gene genealogy. Molecular Ecology 8:1093-1104. Congdon, J. D., A. E. Dunham, and D. W. Tinkle. 198 2. Energy budgets and life histories of reptiles. Biology of Reptilia 13:233-271. Derickson, W. K. 1976. Ecological and physiological aspects of reproductive strategies in 2 lizards. Ecology 57:445-458. Dunham, A. E. 1978. Food availability as a proximat e factor influencing individual growth rates in Iguanid lizard Sceloporus merriami Ecology 59:770-778. Dunham, A. E., B. W. Grant, and K. L. Overall. 1989 Interfaces between biophysical and physiological ecology and the population ecology of terrestrial vertebrate ectotherms. Physiological Zoology 62:335-355. Dunham, A. E., and D. B. Miles. 1985. Patterns of c ovariation in life-history traits of squamate reptiles the effects of size and phylogeny recons idered. American Naturalist 126:231257.

PAGE 20

9 Ferguson, G. W., C. H. Bohlen, and H. P. Woolley. 1 980. Sceloporus undulatus comparative life-history and regulation of a kansas populations Ecology 61:313-322. Ferguson, G. W., and L. G. Talent. 1993. Life histo ry traits of the lizard Sceloporus undulatus from 2 populations raised in a common laboratory en vironment. Oecologia 93:88-94. Grant, B. W., and A. E. Dunham. 1990. Elevational c ovariation in environmental constraints and life histories of the desert lizard Sceloporus merriami Ecology 71:1765-1776. Huey, R. B., and E. R. Pianka. 1981. Ecological con sequences of foraging mode. Ecology 62:991-999. Jackson, J. F. 1972. The population phenetics and b ehavioral ecology of the Florida scrub lizard, Sceloporus woodi Dissertation thesis, University of Florida, Gains eville. —. 1973a. Distribution and population phenetics of Florida scrub lizard, Sceloporus woodi Copeia:746-761. —. 1973b. Phenetics and ecology of a narrow hybrid zone. Evolution 27:58-68. Jones, S. M., and R. E. Ballinger. 1987. Comparativ e life histories of Holbrookia maculata and Sceloporus undulatus in western Nebraska. Ecology 68:1828-1838. Jones, S. M., R. E. Ballinger, and W. P. Porter. 19 87a. Physiological and environmental sources of variation in reproduction prarie lizards in a food rich environment. Oikos 48:325-335. Jones, S. M., S. R. Waldschmidt, and M. A. Potvin. 1987b. An experimental manipulation of food and water growth and time-space utilization of ha tchling lizards ( Sceloporus undulatus ). Oecologia 73:53-59. Leache, A. D. 2009. Species Tree Discordance Traces to Phylogeographic Clade Boundaries in North American Fence Lizards ( Sceloporus ). Systematic Biology 58:547-559. Leache, A. D., and T. W. Reeder. 2002. Molecular sy stematics of the Eastern Fence Lizard ( Sceloporus undulatus ): A comparison of parsimony, likelihood, and Bayes ian approaches. Systematic Biology 51:44-68. Miles, D. B., R. Noecker, W. M. Roosenburg, and M. M. White. 2002. Genetic relationships among populations of Sceloporus undulatus fail to s upport present subspecific designations. Herpetologica 58:277-292.

PAGE 21

10 Niewiarowski, P. H. 1995. Effects of supplemental f eeding and thermal environment on growth rates of eastern fence lizards, Sceloporus undulatus Herpetologica 51:487-496. Niewiarowski, P. H., and W. Roosenburg. 1993. Recip rocal transplant reveals sources of variation in growth rates of the lizard Sceloporus undulatus Ecology 74:1992-2002. Robbins, T. R., J. N. Pruitt, L. E. Straub, E. D. M cCoy, and H. R. Mushinsky. 2010. Transgressive aggression in Sceloporus hybrids confers fitness through advantages in male agonistic encounters. Journal of Animal Ecology 79:137-147. Rose, B. 1981. Factors affecting activity in Sceloporus virgatus Ecology 62:706-716. Sears, M. W., and M. J. Angilletta. 2004. Body size clines in Sceloporus lizards: Proximate mechanisms and demographic constraints, Pages 433-4 42. Sinervo, B., and S. C. Adolph. 1994. Growth plastic ity and thermal opportunity in Sceloporus lizards. Ecology 75:776-790. Smith, G. R. 1998. Habitat-associated life history variation within a population of the striped plateau lizard, Sceloporus virgatus. Acta Oecologic a-International Journal of Ecology 19:167-173. Stearns, S. C. 1976. Life history tactics a revie w of the ideas. Quarterly Review of Biology 51:347. Stearns, S. C., and J. C. Koella. 1986. The evoluti on of phenotypic plasticity in life-history traits predictions of reaction norms for age and size at m aturity. Evolution 40:893-913. Tinkle, D. W. 1972. The dynamics of a Utah populati on of Sceloporus undulatus Herpetologica 28:351-359. Tinkle, D. W., and R. E. Ballinger. 1972. Sceloporus undulatus study of intraspecific comparative demography of a lizard. Ecology 53:570&. Tinkle, D. W., and A. E. Dunham. 1986. Comparative life histories of two syntopic Sceloporine lizards. Copeia 1986:1-18. Vinegar, M. B. 1975. Life-history phenomena in 2 po pulations of lizard Sceloporus undulatus in southwestern New Mexico. American Midland Naturalis t 93:388-402.

PAGE 22

11 Wiens, J. J., and T. W. Reeder. 1997. Phylogeny of the spiny lizards ( Sceloporus ) based on molecular and morphological evidence. Herpetologica l Monographs 11:1-101.

PAGE 23

12 Chapter 2: Variation in Life History Tactics on a Fine Geograp hic Scale and Along a Temperature Gradient Elucidates the Cogradient to Countergradie nt Switch in Sceloporus Lizards Abstract Suites of life history traits observed on large geo graphic scales often follow environmental variation, such as that found in temp erature. On large geographic scales differences between populations often have an adapt ive component, and result in countergradient variation. Cogradient variation, which occurs when trait differences along an environmental gradient follow patterns that are consistent with p lastic trait responses, is often observed over fine scale environmental gradients. Most squamate speci es follow a cogradient pattern because their activity periods are constrained by the thermal env ironment: longer activity periods in warmer and/or southern climates allow for more energy acqu isition through increased foraging activity and faster assimilation rates. However, the Easter n Fence Lizard, Sceloporus undulatus is one species that shows countergradient variation in lif e history traits across its range. The mechanisms by which cogradient variation, observed at fine scales, switches to countergradient variation, observed at large scales, must occur on regional scales. We compared Sceloporus populations with only ~2 latitude between them, wh ich corresponds with a 1 C difference in average monthly temperatures. We examined north an d south populations of the Eastern Fence Lizard, S. undulatus and the Florida Scrub Lizard, S. woodi by measuring adult body sizes, reproduction, and hatchling body sizes over a two y ear period. Our results indicate cogradient variation in life history traits of S. undulatus and countergradient variation, at least in adult b ody size, in S. woodi along the Florida peninsula. Thus, S. undulatus exhibits cogradient variation at fine geographic scales although at larger geographi c scales it exhibits countergradient variation.

PAGE 24

13 The cogradient variation in adult female SVL observ ed between the populations of S. undulatus in this study still holds when examined with data from other Florida populations. Key words. Sceloporus undulatus Squamates, lizards, body size, clutch size Introduction Geographic variation in life history tactics occurs through plastic and adaptive responses to the environment (Roff 2002). Plastic responses often follow environmental variation, such as that found in temperature, especially in ectotherms in which physiological processes are directly influenced by temperature (reviewed in Congdon 1989 ; Huey 1991). When changes in trait values along an environmental gradient follow patte rns that are consistent with plastic trait responses, it is referred to as cogradient variatio n (Conover and Schultz 1995). Most squamate species follow a cogradient pattern ( Adolph and Porter 1993; Adolph and Porter 1996; Angilletta et al. 2004b; Ashton and Fe ldman 2003; Sinervo and Adolph 1989; Sinervo and Adolph 1994). One reason squamates, whi ch are ectotherms, exhibit cogradient variation is because their activity periods are con strained by the thermal environment: longer activity periods in warmer and/or southern climates allow for more energy acquisition through increased foraging activity and faster assimilation rates. Such conditions often result in more growth and larger adult body sizes (Sears and Angil letta 2004; Sinervo and Adolph 1994). Also, larger adult body size often correlates with greate r clutch mass per bout through more and/or larger eggs (Bell 1977; Gadgil and Bossert 1970; Sc haffer and Elson 1975; Stearns and Crandall 1981; Tinkle 1969; Tinkle et al. 1970; Wiley 1974). Differences between populations often also have an adaptive component (Ferguson and Talent 1993; Niewiarowski and Roosenburg 1993), esp ecially on large geographic scales, which can result in countergradient variation (Angilletta et al. 2004b; Sears and Angilletta 2004). Countergradient variation occurs when the phenotypi c differences along a gradient are not what would be predicted to occur from a purely plastic r esponse (Conover and Schultz 1995). Although most squamates exhibit cogradient variatio n in life history tactics, some groups within

PAGE 25

14 squamates exhibit countergradient variation at larg e geographic scales. Bergmann clines (Angilletta et al. 2004a; Bergmann 1847), where lar ger body sizes are observed at higher latitudes and colder climates, are examples of coun tergradient variation that are observed in some squamate species. Some Sceloporus lizard species, for example, follow Bergmann’s rul e. Among Sceloporus lizard species in North America, in fact, 4 specie s follow Bergman’s rule and 6 species follow the opposite pattern (Angilletta et al. 2004a; Ashton and Feldman 2003). The Eastern Fence Lizard, Sceloporus undulatus is one species that has been studied extensively across its North American range and exh ibits countergradient variation in adult body size and other life history traits. In colder, nor thern habitats, adult female body size is larger, intrinsic growth rates are faster but growth still occurs slower over longer lifespans, and clutch mass is greater per bout when compared to populatio ns from warmer habitats. In warmer, southern habitats, adult female body size is smalle r, intrinsic growth rates are slower but growth still occurs faster and more consistently over the lifespan, and clutch mass is less per bout but more bouts occur compared to populations from coole r habitats (Angilletta et al. 2004a; Ferguson and Talent 1993; Niewiarowski and Roosenburg 1993; Sears and Angilletta 2004; Tinkle and Ballinger 1972). The eastern S. undulatus group has three distinct clades which include one clade on either side of the Appalachians and the S. woodi clade (Leache and Reeder 2002). It is unknown whether S. woodi exhibits coor countergradient variation in life history traits along its latitudinal range, but countergradient variation was found to h ave evolved in parallel in both of the other clades within the eastern S. undulatus group, at least in regard to intrinsic embryonic g rowth (Oufiero and Angilletta 2006). Cogradient variation in life history traits has bee n suggested for use as a null model (Sears and Angilletta 2004) because on finer geogra phic scales populations of the same species are likely to be less genetically distinct, and the refore more likely to exhibit cogradient variation through plastic responses to environmental gradient s. Overall, the hypothesis is cogradient variation at fine geographic scales, because popula tions experience relatively small environmental changes, and countergradient variatio n at large geographic scales, because

PAGE 26

15 populations experience environmental changes great enough to require adaptive responses. The mechanisms by which cogradient variation switches t o countergradient variation must occur at the finer, regional scales. Comparative life histo ry studies between populations with small though marked environmental differences may elucidate the mechanisms that result in the large-scale trends. To begin examining life history differences at smal ler scales, we compared Sceloporus populations with only ~2 latitude (approximately 1 00 miles) between them. This geographic distance corresponds with a 1 C difference in aver age monthly temperatures (see results) and presumably longer activity periods experienced by s outhern populations. We included two Eastern Fence Lizard, S. undulatus populations and two Florida Scrub Lizard, S. woodi populations over the same latitudinal distance (Fig 2.1), and measured adult body sizes, reproduction, and hatchling body sizes over a two y ear period. Over this short latitudinal distance the southern, warmer populations that experience lo nger activity periods should exhibit larger adult body size and greater total clutch mass per b out, which is consistent with cogradient variation (Adolph and Porter 1993; Conover and Schu ltz 1995; Sinervo and Adolph 1994). We measured food availability for each population in o rder to examine its relationship, along with environmental temperatures, with population specifi c life history traits. We also ran a common laboratory incubation experiment to determine wheth er there was an intrinsic component influencing incubation period, a proxy for rate of embryonic development. Under the null hypothesis of observed geographic variation being c aused by plastic responses, the intrinsic rate of embryonic development should be similar between north and south populations within species. These data are examined along with what is known ab out other populations along the Florida peninsula and trends are discussed. Materials and Methods Study species. Sceloporus woodi lives in open scrub habitats on remnant Pliocene a nd Pleistocene sand ridges in central Florida. Open sc rub habitats consist of sparse sand pines, oak shrubs, and extensive bare ground. Sceloporus woodi occurs in disjunct, genetically divergent

PAGE 27

16 populations along the Florida ridge (Clark et al. 1 999; McCoy et al. 2004) and is a rare species, although locally abundant (McCoy and Mushinsky 1992 ). Sceloporus woodi resembles the most southern populations of S. undulatus in life histories (Tinkle and Ballinger 1972), wit h relatively small body sizes, small clutch sizes, and short lif e spans (Demarco 1989; McCoy et al. 2004). Sceloporus undulatus is common in the southeastern United States (Conan t 1975) and abundant in sandhill habitats of central Florida. Sandhill habitats consist of long-leaf pines, turkey oaks, and ground cover of wiregrass and fallen pine needl es (Myers and Ewel 1990b). Sceloporus woodi precinctive to Florida scrub, presumably diverged from S. undulatus when rising waters of the Pleistocene isolated several populations on the sandy ridges of Florida (Clark et al. 1999). These Florida populations are closely related, with both species in the eastern S. undulatus group according the most recent phylogeny (Leache 2009; L eache and Reeder 2002). Isolated hybridization events also occur where the species a nd habitats are adjacent (Jackson 1973b; Robbins et al. 2010) Collection and housing of adult female lizards. Female Sceloporus lizards ( N =278) were collected from four populations in Florida, on e northern and one southern population of each species. Collecting occurred from March to Septemb er in 2004 and 2005. The northern populations were collected from the Ocala National Forest, Marion County. Each species was collected from their respective habitats, which inc luded a S. undulatus population ( N =75) from N 2902’18”, W 8133’35”and a S. woodi population ( N =69) from N 2906’29”, W 8148’34”(see map; Fig. 2.1). The southern populations of S. undulatus ( N =64) and S. woodi ( N =60) were collected from Balm Boyette Preserve, Hillsborough County, N 27 45 60 W 82 15 07 and Avon Park Air Force Range, Highlands County, N 27 37’ 07”, W 81 15’20”, respectively. Lizards were captured using a noosing technique, gi ven a unique toe clip for identification (Waichman 1992), contained individually in a cotton bag or plastic-ware, and then collectively transported in a cooler kept at 20-30 C back to th e campus of the University of South Florida. Each lizard was housed individually in the laborat ory, labelled by their toe clip, species, capture date, and site of origin, and provided fres h water and crickets daily. Containers (30 x 17

PAGE 28

17 x 12 cm) included a sand substrate, water dish, and plastic cover object for basking and refuge. Heat lamps maintained temperature gradients within containers that averaged 31 C during the daytime portion of a 12/12 hr day/night cycle. Mea surements for each lizard were taken immediately upon return to the laboratory and inclu ded snout-vent-length (SVL) and tail-length (TL) obtained by a ruler to the nearest 0.5 mm and mass to the nearest 0.01 g obtained using an electronic balance. Mass was measured twice a week while the female lizards were in the laboratory and all measurements were re-taken befor e lizards were released to their respective habitat. Mass after oviposition was used for analy sis to minimize inconsistencies that would have been caused by different stages of gravidity when c aptured. Overall trends in the traits of interest were analy zed using analyses of variance (ANOVA), or multivariate analyses (MANOVA) where appropriate with species ( undulatus or woodi ), latitude (north or south), and year (2004 or 2005) as factors. The species x latitude interaction was also included in the models. Separate two-facto r ANOVAs were used for species specific post-hoc comparisons when species x latitude intera ctions were significant. To ensure that female lizards had reached age at maturity and were therefore adults, only individuals that oviposited were included in adult female body size analyses. Snout-vent length data was not benefited by transformations, but mass and TL were log-transformed to meet normality and homoscedasticity assumptions. Snout-vent length wa s used as covariate when mass and TL were analyzed and individuals with broken tails wer e not used when analyzing TL. Reproduction Each lizard and housing was checked daily for ov iposition. After oviposition, eggs were collected, counted, and weig hed to the nearest 0.0001 g using an electronic balance and the post-oviposition mass of the female was obtained to the nearest 0.01 g. Each clutch ( N =175) of eggs ( N =1017) was placed in a glass jar (120 ml) and burie d completely in vermiculite that was premixed to a wa ter potential of –450 kPa. Water potential for vermiculite was determined by Packard et al (1987). All vermiculite was oven-dried at 100 C for at least 4 hours prior to mixing with distilled wat er. Each jar was covered with plastic kitchen

PAGE 29

18 wrap, sealed with a rubber band and placed in an in cubator set at a constant 28 C. Vermiculite was replaced for each clutch after 25 days of incub ation. Clutch size and egg mass were analyzed with SVL of dams as a covariate. Egg masses were analysed as clutch means and excluded any eggs that were found on top of the sand and desiccated. Incubation periods were also analysed a s clutch means, which included only individuals that hatched. Average egg mass and clu tch size were multiplied for each lizard and analyzed as total clutch mass. Dam SVL and body co ndition (residuals from a regression of SVL on mass within species) were used as covariates whe n analyzing total clutch mass. Hatchlings. Eggs in the incubator were checked daily for hatchl ings ( N =567) and hatchling measurements were taken immediately after hatching. For each individual, hatching date and hatchling sex was recorded, and their SVL, TL, and mass (to the nearest 0.0001 g) measured. Male and female hatchlings did not diffe r in any phenotype (all P >0.08), thus they were combined for all analyses. Each hatchling was marked with a unique combinatio n of toeclips (Waichman 1992) and housed in a 38-liter (10 gallon) terrarium in the laboratory until their release in the field. We provided water and cricke ts (dusted with vitamin/mineral mix) daily for the hatchlings. Hatchling SVL, mass, and TL were analyzed as clutch means with dam SVL as covariate and species, latitude, and year as factors. The sp ecies x latitude interaction was also included in the model. Hatchling SVL was also used as covariate for hatchling mass and TL. Environmental variables. Population-specific environmental temperatures and food availability were examined for differences among sp ecies, latitudes, and years where possible, and then examined for overall correlations with pop ulation-specific traits. Environmental temperatures were examined between latitudes over t hat last 50 years and over the particular years of the study. For the long term, monthly tem peratures averaged over the last 50 years were used as the dependent variables in a mixed mod el analysis with time (month) as a random factor and latitude as a fixed factor. For the par ticular years of the study, average monthly

PAGE 30

19 temperatures were used as the dependent variables i n a mixed model analysis with time (month) as a random factor, and latitude, and year as fixed factors. Average monthly temperatures were from weather station data collected from stations o f north and south latitudes similar to the north and south populations of each species (long term fr om Bartow and short term from Wauchula, Florida in the south and Ocala, Florida, in the nor th; from Weatherbase at www.weatherbase.com for long term averages and the National Climate Dat a Center at www.ncdc.noaa.gov/oa/climate/climatedata.html for 2004 and 2005). Food availability was measured for each population using an array of pitf all traps (14 traps at each of the 4 sites) that were opened approximately once a month for five tra pping periods per season between August and February 2004-2006. Each trapping period laste d 3-7 days. Food availability was examined using repeated measures ANOVA with species, latitud e, and year as factors and total biomass per trap per day as the dependent variable at each time period. The species x latitude interaction was also included in the model. The index of total biomass per trap per day was estimated by summing the lengths of the individual arthropods ca ught in each trap during each time period and dividing by the number of days open. The biomass i ndex data was log-transformed to meet assumptions of the ANOVA. Arthropods greater than 5 mm in length were considered too large for consumption and not included in the analysis (J ackson 1973b). Within species, correlations between the means (fro m both years of the study) of population specific traits, food availability, and environmental temperatures were run to examine environmental influence. Estimated marginal means from the previously run ANOVAs were used in these correlations to account for influential co variates. Results Adult female lizards. In the southern populations, adult female SVL, body condition, and TL were greater overall, however, there was a s pecies x latitude interaction in SVL (Table 2.1). Sceloporus undulatus had longer SVL in the southern population, but no difference was found between the north and south populations of S. woodi (Fig. 2.2). The latitudinal difference in SVL, which is the fundamental body size measurement found in S. undulatus is consistent with

PAGE 31

20 cogradient variation, but in S. woodi there was no difference between north and south populations. Between species, adult female S. undulatus had larger body size than S. woodi. Particularly, the species differed in that S. undulatus had longer SVL and greater mass but TL was similar among species (Table 2.1). Dam SVL inf luenced both mass and TL (Table 2.1). Reproduction. Latitudinal trade-offs occurred between egg mass an d clutch size in both species. Average egg mass was larger and clutch si zes were smaller (after accounting for SVL) in the south. Total clutch mass was not different between latitudes, likely because of the tradeoffs between egg mass and clutch size, however, the re was a species x latitude interaction (Table 2.2). Sceloporus undulatus had marginally greater total clutch mass in the so uth while S. woodi had similar total clutch mass in both populations ( Fig. 2.3). Incubation periods were not different between latitudes (P=0.64) for either species. Between species, faster development and greater tot al clutch mass were found in S. undulatus when compared to S. woodi. Sceloporus undulatus had incubation periods of 55.5 days and S. woodi of 62.3 days ( P <0.001); SVL did not explain the difference. Sceloporus undulatus had larger clutch size after accounting for the ef fects of SVL and larger average egg mass that was not explained by SVL (Table 2.2). In deed, total clutch mass was greater in S. undulatus (Table 2.2). Hatchlings. Hatchling body sizes were larger in the south, howe ver, SVL was only marginally significant (Table 2.3). There was also a species x latitude interaction in hatchling TL with longer TL in the south in Sceloporus woodi but not S. undulatus (Fig. 2.4). Between species, hatchling body sizes were generally larger in S. undulatus. The longer SVL of S. undulatus hatchlings was not affected by dam SVL (Table 2.3.) but the greater mass was affected by dam SVL as well as hatchling SVL, and still significant (Table 2.3). Hatchling tail length was only marginally different between species and was not af fected by dam SVL, but it was influenced by hatchling SVL (Table 2.3).

PAGE 32

21 Environmental variables. Environmental temperatures were significantly diffe rent between north (39.1 C) and south (40.1 C) latitud es ( P <0.001) with the southern latitude having a 1 C higher average temperature over the last 50 years. Over 2004 and 2005 the southern latitude was only 0.5 C warmer ( P =0.009) and not different between years ( P =0.633). Arthropod abundance was marginally greater overall in the sou thern latitude ( P =0.057) and there was also more food available in 2004 than 2005 ( P <0.001). However, a species x latitude interaction ( P =0.02) was found (Fig. 2.5). Species specific post -hoc tests confirmed that S. undulatus had greater food availability in the south ( P =0.002) and that latitude did not affect food avail ability for S. woodi ( P =0.772). Food availability did not correlate with average environmental temperatures or with any trait within species (all P >0.18). Environmental temperatures were correlated with some traits that differed between species (Table 2. 4). Temperatures correlated with adult female SVL in S. undulatus and adult TL in S. woodi (Table 2.4). Discussion Our results indicate cogradient variation in life h istory traits of Sceloporus undulatus and countergradient variation, at least in adult body s ize, in S. woodi along the Florida peninsula. Thus, S. undulatus exhibits cogradient variation, specifically in adu lt body size, at fine geographic scales although at larger geographic scales it exhi bits countergradient variation (Angilletta et al. 2004a; Sears and Angilletta 2004). Because S. woodi the third clade in the eastern S. undulatus group, exhibits countergradient varation in adult b ody size, it appears that all three clades have evolved countergradient variation, at least in some life history traits (Oufiero and Angilletta 2006). At the fine geographic scale of this study, however S. woodi did not exhibit shorter intrinsic incubation periods in the northern population, whic h is inconsistent with what Oufiero and Angilletta found (2006) in the other two Appalachia n clades of the eastern S. undulatus group. The similar incubation periods between the north an d south populations of S. woodi may be a result of the finer geographic scale of our study, compared to the scale that Oufiero and Angilletta examined (approximately 8 latitude).

PAGE 33

22 Latitude and the associated differences in environm ental temperatures have influenced some traits in both species similarly and some diff erently. In the southern populations of both species, adult body condition and TL were greater. The positive relationships between environmental temperature and adult TL were also ob served among populations and years, although only significant in S. woodi (Table 2.4). Egg mass was greater and clutch size smaller in the southern population of both species, which resu lted in similar total clutch mass (although marginally greater in the southern population of S. undulatus ) between the north and south populations. Hatchling SVL and body condition were also greater in the southern populations of both species. Mass was, therefore, greater in the southern populations of both species at all three life history stages, that of adult, egg, and hatchling, which is consistent with what would be expected through a plastic response to temperature and cogradient variation. There were three traits – adult SVL, total clutch m ass, and hatchling TL – that showed species x latitude interactions. Sceloporus undulatus had greater adult SVL and total clutch mass in the southern population with similar hatchling T L between populations. Sceloporus woodi had similar adult SVL and total clutch mass between pop ulations with greater hatchling TL in the southern population. The greater adult SVL and tot al clutch mass in S. undulatus follows what would be expected by a plastic response to increase d temperatures, likely through greater activity periods. The positive relationship between adult f emale SVL and environmental temperature was also observed among populations and years of the st udy (Table 2.4). The increased potential activity period in the south may be used by S. undulatus for more growth, but the lack of difference in SVL between S. woodi populations suggests that S. woodi in the south does not have an increased activity period and/or uses inact ivity opportunistically (Rose 1981). One reason may be that they are behaviourally less acti ve to avoid greater predation pressure, and another that the increase in temperature is high en ough to actually decrease the activity period so that they experience warmer temperatures but a shor ter activity period. Because thermoregulation is accomplished through behavioura l movement, it may also reflect similar proportions of suitable microhabitat even though su itable microhabitats shift spatially relative to the low level vegetation. Further study is require d to parse the potential mechanisms.

PAGE 34

23 The cogradient variation in adult female SVL observ ed between the populations of Sceloporus undulatus in this study still holds when examined with data from other Florida populations. There are no publications with popula tion specific average SVL of adult female S. woodi (at least at different latitudes), but one such pu blication did exist for S. undulatus in Florida. Mobley (1998) studied two populations of S. undulatus that were intermediate in latitude to the populations in this study. Although the population s in Mobley (1998) were geographically closer than the two populations in this study, the latitud inal difference in SVL was significant and consistent with this study, with greater adult fema le SVL in the southern population. Collectively, these four populations exhibit latitudinal cogradie nt variation in adult female SVL along the Florida peninsula (Fig. 2.6). The results of this comparat ive study are consistent with the paradigm of using cogradient variation as the null hypothesis ( Sears and Angilletta 2004) with the understanding that when countergradient variation i s observed between populations some environmental threshold has been reached and some a daptive response mechanism has occurred. More studies focusing on populations at the boundaries of the cogradientcountergradient switch in phenotypic variation will provide critical insights into the mechanisms involved and the sequence of phenotypes under selec tion. Food availability and operative temperatures are tw o environmental factors that influence life history traits because they deal with energy i nput and rates of energy throughput and may very well constrain growth and adult body size simi larly. For example, a 50% reduction in food caused a reduction in growth rates equal to that of a 50% reduction in daily activity times in Sceloporus undulatus (Sinervo and Adolph 1994). We measured food avail ability at each site for both years but could not determine any overall rela tionships with life history traits. It should be noted, however, that the species x latitude interac tions found in adult female SVL and total clutch mass were similar in direction to the species x lat itude interaction found in food availability (Figs 2.2 & 2.5). It seems that environmental temperatur es had stronger relationships with life history traits than food availability but more population l evel samples would be needed to separate the effects of temperature and food.

PAGE 35

24 Although it was not the main focus of our study, ou r results are also consistent with countergradient variation between the species. In the cooler sandhill habitats of S. undulatus compared to the warmer scrub habitats of S. woodi (2.5 C warmer on average; unpublished data), adult female SVL and body condition were gre ater, intrinsic incubation periods were shorter, and total clutch mass was greater per bout with greater egg mass and larger clutch sizes (Tables 2.1, 2.2, and 2.5), all of which is consist ent with the literature (Andrews et al. 2000; Crenshaw 1955; Demarco 1992; Demarco 1989; Jackson and Telford 1974; Mobley 1998) Conclusions. Our study supports the hypothesis of cogradient var iation at fine geographic scales even when countergradient variati on is observed on larger geographic scales. With regard to adult body size, the countergradient variation observed across the range of Sceloporus undulatus has been explained through juvenile survival rates (Angilletta et al. 2004a; Sears and Angilletta 2004). The larger adult body sizes in the cooler, northern populations are associated with greater juvenile survival because e xtrinsic growth rates are slower (Angilletta et al. 2004a; Sears and Angilletta 2004). At the fine r geographic scale of our study, greater adult body sizes could be a result of greater juvenile su rvival or longer activity periods and therefore faster growth rates. Intrinsic and extrinsic survi val and growth rates need to be studied in these specific populations to better understand how they may be linked to the cogradient variation in adult body size observed in S. undulatus and the countergradient variation in adult body si ze observed in S. woodi The difference in observed geographic variation in adult body size between Sceloporus undulatus and S. woodi is interesting, and may be a result of an already constrained suite of life history tactics in S. woodi Because S. woodi exists in a relatively warmer habitat, it resemble s the most southern populations of S. undulatus in life histories (Tinkle and Ballinger 1972), wit h relatively small body sizes, small clutch sizes, an d short life spans (Demarco 1989; McCoy et al. 2004). Smaller adult body sizes may not be a viabl e option. Studies examining the influence of plastic and adaptive responses to these populationspecific environments using reciprocal transplants and common environment experiments are currently being conducted and should

PAGE 36

25 provide critical insights into the mechanisms that have resulted in the geographic variation observed in this study. More studies of life histo ry tactics on fine geographic scales are necessary to begin examining which traits are under strong selection and/or susceptible to plastic responses, and to begin examining if general sequen ces of phenotypic change occur. Acknowledgments Many thanks go to the undergraduate army that helpe d us collect these data and care for the lizards in the laboratory, especially Lorelei S traub. We thank Tom Raffel for statistical help. Lizards were collected under collection permit WX05 107 issued by the State of Florida Fish and Wildlife Conservation Commission. All protocols we re reviewed and accepted by the USF Institutional Animal Care and Use Committee, IACUC file #2778.

PAGE 37

26 References Adolph, S. C., and W. P. Porter. 1993. Temperature, activity, and lizard life histories. American Naturalist 142:273-295. —. 1996. Growth, seasonality, and lizard life histo ries: Age and size at maturity. Oikos 77:267278. Andrews, R. M., T. Mathies, and D. A. Warner. 2000. Effect of incubation temperature on morphology, growth, and survival of juvenile Scelop orus undulatus. Herpetological Monographs:420-431. Angert, A. L., and D. W. Schemske. 2005. The evolut ion of species' distributions: Reciprocal transplants across the elevation ranges of Mimulus cardinalis and M. lewisii. Evolution 59:1671-1684. Angilletta, M. J., P. H. Niewiarowski, A. E. Dunham A. D. Leache, and W. P. Porter. 2004a. Bergmann's clines in ectotherms: Illustrating a lif e-history perspective with sceloporine lizards. American Naturalist 164:E168-E183. Angilletta, M. J., T. D. Steury, and M. W. Sears. 2 004b. Temperature, growth rate, and body size in ectotherms: Fitting pieces of a life-history puz zle. Integrative and comparative biology 44:498-509. Ashton, K. G., and C. R. Feldman. 2003. Bergmann's rule in nonavian reptiles: Turtles follow it, lizards and snakes reverse it. Evolution 57:1151-11 63. Ballinger, R. E. 1977. Reproductive strategies fo od availability as a source of proximal variation in a lizard. Ecology 58:628-635. Ballinger, R. E., and J. D. Congdon. 1980. Food res ource limitation of body growth rates in Sceloporus scalaris (Sauria, Iguanidae). Copeia 198 0:921-923. Ballinger, R. E., D. L. Droge, and S. M. Jones. 198 1. Reproduction in a Nebraska sandhills population of the northern prarie lizard Scelopours undulatus garmani American Midland Naturalist 106:157-164. Bell, G. 1977. The life of the smooth newt (Trituru s vulgaris) after metamorphosis. Ecological Monographs 47:279-299.

PAGE 38

27 Bergmann, C. 1847. Uber die Verhaltnisse der Warmeo konomie der Thiere zu ihrer Grosse. Gottinger Studien 3:595-708. Brooks, H. K. 1972. The Geology of the Ocala Nation al Forest in Snedaker, ed., The Ecology of the Ocala National Forest. Tallahassee, U.S. Forest Service, U.S.D.A. Bullock, J. M., R. J. Edwards, P. D. Carey, and R. J. Rose. 2000. Geographical separation of two Ulex species at three spatial scales: does competit ion limit species' ranges? Ecography 23:257-271. Burnham, K. P., and D. R. Anderson. 2002, Model sel ection and multimodal inference: a practical information-theoretic approach. New York, NY, USA, Springer-Verlag. Clark, A. M., B. W. Bowen, and L. C. Branch. 1999. Effects of natural habitat fragmentation on an endemic scrub lizard ( Sceloporus woodi ): an historical perspective based on a mitochondrial DNA gene genealogy. Molecular Ecology 8:1093-1104. Conant, R. 1975, A field guide to the reptiles and and amphibians of eastern and central North America Boston, Houghton Mifflin. Congdon, J. D. 1989. Proximate and evolutionary con straints on energy relations of reptiles. Physiological Zoology 62:356-373. Congdon, J. D., A. E. Dunham, and D. W. Tinkle. 198 2. Energy budgets and life histories of reptiles. Biology of Reptilia 13:233-271. Conover, D. O., and E. T. Schultz. 1995. Phenotypic similarity and the evolutionary significance of countergradient variation. Trends in Ecology & Evol ution 10:248-252. Coyne, J. A., and H. A. Orr. 2004, Speciation. Sund erland, MA, Sinauer Associates, Inc. Crenshaw, J., W, Jr. 1955. The life history of the southern spiny lizard, Sceloporus undulatus undulatus Latreille. American Midland Naturalist 54 :257-298. Cumming, G. S. 2002. Comparing climate and vegetati on as limiting factors for species ranges of African ticks. Ecology 83:255-268. Demarco, V. 1992. Embryonic development times and e gg retention in 4 species of Sceloporine lizards. Functional Ecology 6:436-444.

PAGE 39

28 Demarco, V. G. 1989. Annual variation in the season al shift in egg size and clutch size in Sceloporus woodi. Oecologia 80:525-532. Derickson, W. K. 1976. Ecological and physiological aspects of reproductive strategies in 2 lizards. Ecology 57:445-458. Dunham, A. E. 1978. Food availability as a proximat e factor influencing individual growth rates in Iguanid lizard Sceloporus merriami Ecology 59:770-778. Dunham, A. E., B. W. Grant, and K. L. Overall. 1989 Interfaces between biophysical and physiological ecology and the population ecology of terrestrial vertebrate ectotherms. Physiological Zoology 62:335-355. Dunham, A. E., and D. B. Miles. 1985. Patterns of c ovariation in life-history traits of squamate reptiles the effects of size and phylogeny recons idered. American Naturalist 126:231257. Ferguson, G. W., C. H. Bohlen, and H. P. Woolley. 1 980. Sceloporus undulatus comparative life-history and regulation of a kansas populations Ecology 61:313-322. Ferguson, G. W., and L. G. Talent. 1993. Life histo ry traits of the lizard Sceloporus undulatus from 2 populations raised in a common laboratory en vironment. Oecologia 93:88-94. Gadgil, M., and W. Bossert. 1970. Life historical c onsequences of natural selection. The American Naturalist 104:1-24. Galloway, L. F., and C. B. Fenster. 2000. Populatio n differentiation in an annual legume: Local adaptation. Evolution 54:1173-1181. Grant, B. W., and A. E. Dunham. 1988. Thermally imp osed time constraints on the activity of the desert lizard Sceloporus merriami Ecology 69:167-176. —. 1990. Elevational covariation in environmental c onstraints and life histories of the desert lizard Sceloporus merriami Ecology 71:1765-1776. Hall, M. C., and J. H. Willis. 2006. Divergent sele ction on flowering time contributes to local adaptation in Mimulus guttatus populations. Evoluti on 60:2466-2477. Hendry, A. P., P. Nosil, and L. H. Rieseberg. 2007. The speed of ecological speciation. Functional Ecology 21:455-464.

PAGE 40

29 Hereford, J. 2009. A Quantitative Survey of Local A daptation and Fitness Trade-Offs. American Naturalist 173:579-588. Hereford, J., and A. A. Winn. 2008. Limits to local adaptation in six populations of the annual plant Diodia teres. New Phytologist 178:888-896. Huey, R. B. 1991. Physiological consequences of hab itat selection. American Naturalist 137:91115. Huey, R. B., and E. R. Pianka. 1981. Ecological con sequences of foraging mode. Ecology 62:991-999. Iverson, J. B. 1991. Patterns of survivorship in tu rtles (Order Testudines). Canadian Journal of Zoology-Revue Canadienne De Zoologie 69:385-391. Jackson, J. F. 1972. The population phenetics and b ehavioral ecology of the Florida scrub lizard, Sceloporus woodi Dissertation thesis, University of Florida, Gains eville. —. 1973a. Distribution and population phenetics of Florida scrub lizard, Sceloporus woodi Copeia:746-761. —. 1973b. Phenetics and ecology of a narrow hybrid zone. Evolution 27:58-68. Jackson, J. F., and S. R. Telford. 1974. Reproducti ve ecology of the Florida scrub lizard, Sceloporus woodi. Copeia:689-694. Jones, S. M., and R. E. Ballinger. 1987. Comparativ e life histories of Holbrookia maculata and Sceloporus undulatus in western Nebraska. Ecology 68:1828-1838. Jones, S. M., R. E. Ballinger, and W. P. Porter. 19 87a. Physiological and environmental sources of variation in reproduction prarie lizards in a food rich environment. Oikos 48:325-335. Jones, S. M., S. R. Waldschmidt, and M. A. Potvin. 1987b. An experimental manipulation of food and water growth and time-space utilization of ha tchling lizards ( Sceloporus undulatus ). Oecologia 73:53-59. Kawecki, T. J., and D. Ebert. 2004. Conceptual issu es in local adaptation. Ecology Letters 7:1225-1241. Leache, A. D. 2009. Species Tree Discordance Traces to Phylogeographic Clade Boundaries in North American Fence Lizards ( Sceloporus ). Systematic Biology 58:547-559.

PAGE 41

30 Leache, A. D., and T. W. Reeder. 2002. Molecular sy stematics of the Eastern Fence Lizard (Sceloporus undulatus): A comparison of parsimony, likelihood, and Bayesian approaches. Systematic Biology 51:44-68. Lebreton, J. D., K. P. Burnham, J. Clobert, and D. R. Anderson. 1992. Modeling survival and testing biological hypotheses using marked animals a unified approach with case studies Ecological Monographs 62:67-118. Linhart, Y. B., and M. C. Grant. 1996. Evolutionary significance of local genetic differentiation in plants. Annual Review of Ecology and Systematics 27 :237-277. Lowry, D. B., R. C. Rockwood, and J. H. Willis. 200 8. Ecological reproductive isolation of coast and inland races of Mimulus guttatus. Evolution 62: 2196-2214. McCoy, E. D., P. P. Hartmann, and H. R. Mushinsky. 2004. Population biology of the rare Florida scrub lizard in fragmented habitat. Herpetologica 6 0:54-61. McCoy, E. D., and H. R. Mushinsky. 1992. Rarity of organisms in the sand pine scrub habitat of Florida. Conservation Biology 6:537-548. McKinnon, J. S., S. Mori, B. K. Blackman, L. David, D. M. Kingsley, L. Jamieson, J. Chou et al. 2004. Evidence for ecology's role in speciation. Na ture 429:294-298. Miles, D. B., R. Noecker, W. M. Roosenburg, and M. M. White. 2002. Genetic relationships among populations of Sceloporus undulatus fail to s upport present subspecific designations. Herpetologica 58:277-292. Mobley, E. R. 1998. A base line population study of the southern fence lizard, Sceloporus undulatus undulatus in central Florida., University of Central Florid a, Orlando, Florida. Myers, R. L., and J. L. Ewel. 1990a. Ecosystems of Florida. Ecosystems of Florida.:i-xviii, 1-765. —. 1990b. Scrub and high pine. Ecosystems of Florid a.:150-193. Nagy, E. S., and K. J. Rice. 1997. Local adaptation in two subspecies of an annual plant: Implications for migration and gene flow. Evolution 51:1079-1089. Niewiarowski, P. H. 1994. Understanding geographic life-history variation in lizards. Lizard ecology: Historical and experimental perspectives:3 1-49.

PAGE 42

31 —. 1995. Effects of supplemental feeding and therma l environment on growth rates of eastern fence lizards, Sceloporus undulatus Herpetologica 51:487-496. Niewiarowski, P. H., and W. Roosenburg. 1993. Recip rocal transplant reveals sources of variation in growth rates of the lizard Sceloporus undulatus Ecology 74:1992-2002. Nosil, P. 2007. Divergent host plant adaptation and reproductive isolation between ecotypes of Timema cristinae walking sticks. American Naturalis t 169:151-162. Nudds, T. D. 1977. Quantifying the vegetative struc ture of wildlife cover. Wildlife Society Bulletin 5:113-117. Oufiero, C. E., and M. J. Angilletta. 2006. Converg ent evolution of embryonic growth and development in the eastern fence lizard (Sceloporus undulatus). Evolution 60:1066-1075. Randall, G. M. 1982. The dynamics of an insect popu lation throughout its altitudinal distribution: Coleophora alticolella (Lepidoptera) in northern En gland. Journal of Animal Ecology 51:993-1016. Rice, K. J., and R. N. Mack. 1991. Ecological genet ics of Bromus tectorum. III. The demography of reciprocally sown populations. Oecologia 88:91-1 01. Rieseberg, L. H., and J. H. Willis. 2007. Plant spe ciation. Science 317:910-914. Robbins, T. R., J. N. Pruitt, L. E. Straub, E. D. M cCoy, and H. R. Mushinsky. 2010. Transgressive aggression in Sceloporus hybrids confers fitness through advantages in male agonistic encounters. Journal of Animal Ecology 79:137-147. Roff, D. A. 2002, Life History Evolution. Sunderlan d, MA, Sinauer Associates, Inc. Root, T. 1988. Energy constraints on avian distribu tions and abundances. Ecology 69:330-339. Rose, B. 1981. Factors affecting activity in Sceloporus virgatus Ecology 62:706-716. Rundle, H. D. 2002. A test of ecologically dependen t postmating isolation between sympatric sticklebacks. Evolution 56:322-329. Rundle, H. D., and P. Nosil. 2005. Ecological speci ation. Ecology Letters 8:336-352. Schaffer, W. M., and P. F. Elson. 1975. Adaptive si gnificance of variations in life-history among local populations of alantic salmon in North Americ a. Ecology 56:577-590. Schluter, D. 2001. Ecology and the origin of specie s. Trends in Ecology & Evolution 16:372-380.

PAGE 43

32 Sears, M. W., and M. J. Angilletta. 2004. Body size clines in Sceloporus lizards: Proximate mechanisms and demographic constraints. Integrative and Comparative Biology 44:433442. Sinervo, B., and S. C. Adolph. 1989. Thermal sensit ivity of growth rate in hatchling Sceloporus lizards environmental, behavioral, and genetic as pects. Oecologia 78:411-419. —. 1994. Growth plasticity and thermal opportunity in Sceloporus lizards. Ecology 75:776-790. Smith, G. R. 1998. Habitat-associated life history variation within a population of the striped plateau lizard, Sceloporus virgatus. Acta Oecologic a-International Journal of Ecology 19:167-173. Stearns, S. C. 1976. Life history tactics a revie w of the ideas. Quarterly Review of Biology 51:347. Stearns, S. C., and R. E. Crandall. 1981. Quantitat ive predictions of delayed maturity. Evolution 35:455-463. Stearns, S. C., and J. C. Koella. 1986. The evoluti on of phenotypic plasticity in life-history traits predictions of reaction norms for age and size at m aturity. Evolution 40:893-913. Taniguchi, Y., and S. Nakano. 2000. Condition-speci fic competition: implications for the altitudinal distribution of stream fishes. Ecology 81:2027-2039 Terborgh, J., and J. S. Weske. 1975. The role of co mpetition in distribution of Andean birds. Ecology 56:562-576. Tinkle, D. W. 1967, The life and demography of the side-blotched lizard, Uta stansburiana : Miscellaneous Publications, v. No. 132. Ann Arbor MI, Museum of Zoology, University of Michigan. —. 1969. Concept of reproductive effort and its rel ation to evolution of life histories of lizards. American Naturalist 103:501-&. —. 1972. The dynamics of a Utah population of Sceloporus undulatus Herpetologica 28:351-359. Tinkle, D. W., and R. E. Ballinger. 1972. Sceloporus undulatus study of intraspecific comparative demography of a lizard. Ecology 53:570&.

PAGE 44

33 Tinkle, D. W., and A. E. Dunham. 1986. Comparative life histories of two syntopic Sceloporine lizards. Copeia 1986:1-18. Tinkle, D. W., H. M. Wilbur, and S. G. Tilley. 1970 Evolutionary strategies in lizard reproduction. Evolution 24:55-&. Via, S., A. C. Bouck, and S. Skillman. 2000. Reprod uctive isolation between divergent races of pea aphids on two hosts. II. Selection against migr ants and hybrids in the parental environments. Evolution 54:1626-1637. Vinegar, M. B. 1975. Life-history phenomena in 2 po pulations of lizard Sceloporus undulatus in southwestern New Mexico. American Midland Naturalis t 93:388-402. Waichman, A. V. 1992. An alphanumeric code for toe clipping amphibians and reptiles. Herpetological Review 23:19-21. Wang, H., E. D. McArthur, S. C. Sanderson, J. H. Gr aham, and D. C. Freeman. 1997. Narrow hybrid zone between two subspecies of big sagebrush (Artemisia tridentata: Asteraceae) .4. Reciprocal transplant experiments. Evolution 51 :95-102. Warner, D. A., and R. Shine. 2005. The adaptive sig nificance of temperature-dependent sex determination: Experimental tests with a short-live d lizard. Evolution 59:2209-2221. White, G. C., and K. P. Burnham. 1999. Program MARK : survival estimation from populations of marked animals. Bird Study 46:120-138. Wiens, J. J., and T. W. Reeder. 1997. Phylogeny of the spiny lizards ( Sceloporus ) based on molecular and morphological evidence. Herpetologica l Monographs 11:1-101. Wiley, R. H. 1974. Evolution of social-organization and life-history patterns among grouse. Quarterly Review of Biology 49:201-227.

PAGE 45

34 nrn nr Figure 2.1 Map of study site locations. Depictied are the n orth and south study populations of Sceloporus undulatus and S. woodi are in Florida.

PAGE 46

35 50 52 54 56 58 60 62 64 66 N SSnout-vent length (mm)Latitude Figure 2.2. Species x latitude interaction in adult female SV L. Error bars represent 1 standard error. P-values are for the effect of latitude fro m post-hoc species-specific ANOVAs. Solid circles ( ) represent S. undulatus and open circles ( ) represent S. woodi

PAGE 47

36 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 N SLatitude 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 N STotal clutch mass (g)Latitude Figure 2.3. Species x latitude interaction in total clutch mas s. Error bars represent 1 standard error. P-values are for the effect of latitude fro m post-hoc species-specific ANOVAs. Solid circles ( ) represent S. undulatus and open circles ( ) represent S. woodi

PAGE 48

37 26 27 28 29 30 31 32 N SHatchling tail length (mm)Latitude 26 27 28 29 30 31 32 N SHatchling tail length (mm)Latitude Figure 2.4. Species x latitude interaction in hatchling tail length. Error bars represent 1 standard error. P-values are for the effect of latitude fro m post-hoc species-specific ANOVAs. Solid circles ( ) represent S. undulatus and open circles ( ) represent S. woodi

PAGE 49

38 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 N SFood availability (log mm/day)Latitude 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 N SFood availability (log mm/day)Latitude Figure 2.5 Food availability for north and south populations of each lizard species. Points are estimated marginal means from the three factor repe ated measures ANOVA. Error bars represent 2 standard error. Solid circles ( ) r epresent S. undulatus and open circles ( ) represent S. woodi

PAGE 50

39 54 56 58 60 62 64 27.52828.52929.5Snout-vent length (mm)Latitude (decimal degrees) Figure 2.6. Cogradient variation in snout-vent length of adul t female Sceloporus undulatus along the Florida peninsula. Points represent population means from this study and from Mobley (1994).

PAGE 51

40 Table 2.1. Effects of species, latitude, and year on adult fe male body sizes. SVL (mm) Body condition TL (mm) Source F P F P F P SVL (mm) 151.494 <0.001 27.829 <0.001 Species 115.270 <0.001 26.261 <0.001 2.129 0.147 Latitude 43.640 <0.001 4.811 0.030 10.442 0.002 Year 0.542 0.462 2.493 0.116 0.440 0.508 Species x Latitude 34.224 <0.001 0.871 0.352 0.185 0.667 Body condition is mass relative to SVL and TL is tail-length relative to SVL. Significant probabilities are denoted in bold.

PAGE 52

41 Table 2.2. Effects of species, latitude, and year on lizard r eproduction. Clutch size (eggs) Egg mass (g) Total clutch mass (g) Source F P F P F P SVL (mm) 39.587 <0.001 0.242 0.624 33.728 <0.001 Body condition 8.798 0.004 Species 4.126 0.044 27.524 <0.001 20.331 <0.001 Latitude 5.662 0.042 21.590 <0.001 2.101 0.150 Year 1.045 0.308 0.935 0.336 0.353 0.553 Species x Latitude 2.065 0.153 2.091 0.151 6.745 0.011 SVL refers of dam SVL and was used as a covariate in all analyses. Body condition is mass relative to SVL. Significant probabilities are den oted in bold.

PAGE 53

42 Table 2.3. Effects of species, latitude, and year on hatchlin g body sizes. SVL (mm) Body condition TL (mm) Source F P F P F P Dam SVL (mm) 1.341 0.250 7.957 0.006 1.061 0.306 Hatchling SVL (mm) 206.482 <0.001 92.220 <0.001 Species 23.091 <0.001 9.942 0.002 3.774 0.055 Latitude 3.280 0.073 13.037 <0.001 6.526 0.012 Year 7.042 0.009 5.324 0.023 0.614 0.435 Species x Latitude 0.004 0.948 1.738 0.191 8.447 0.005 Body condition is mass relative to SVL and TL is tail-length relative to SVL. Dam SVL was used as a covariate in all analyses. Significant probab ilities are denoted in bold.

PAGE 54

43 Table 2.4. Results of correlation analyses between average e nvironmental temperatures and average trait values. Among populations and years ( N =4) S. undulatus S. woodi Trait r P r P Adult SVL (mm) 0.98 0.018 0.13 0.876 Body condition 0.56 0.441 0.93 0.075 Adult TL (mm) 0.87 0.135 0.97 0.035 Hatchling SVL (mm) 0.68 0.324 0.32 0.682 Hatchling mass (g) 0.77 0.232 0.61 0.391 Hatchling TL (mm) -0.24 0.761 0.83 0.172 Incubation period (days) -0.31 0.688 -0.38 0.616 Egg mass (g) 0.92 0.079 0.70 0.300 Clutch size (eggs) -0.22 0.779 -0.54 0.465 Total clutch mass (g) 0.90 0.101 -0.73 0.269 Significant probabilities are denoted in bold. T raits that are bolded are those with significant species x latitude interactions in the ANOVAs.

PAGE 55

44 Table 2.5. Mean trait values for Sceloporus undulatus and S. woodi from the north and south populations. Trait Latitude S. undulatus S. woodi SVL (mm) N 56.6 0.5 53.9 0.8 S 63.6 0.4 54.3 0.5 Mass (g) N 6.795 0.228 5.085 0.192 S 9.445 0.200 5.507 0.203 Tail length (mm) N 72.8 1.3 72.7 1.0 S 82.4 1.5 76.4 1.0 Incubation period (d) N 55.4 0.2 62.2 0.6 S 55.0 0.3 62.2 0.7 Egg mass (g) N 0.342 0.007 0.301 0.007 S 0.394 0.009 0.329 0.007 Clutch size (eggs) N 5.4 0.2 4.7 0.2 S 6.4 0.2 4.1 0.2 Total clutch mass (g/clutch) N 1.832 0.075 1.483 0.081 S 2.603 0.078 1.341 0.067 Hatchling SVL (mm) N 23.2 0.1 21.7 0.3 S 23.9 0.1 22.2 0.3 Hatchling mass (g) N 0.460 0.009 0.378 0.015 S 0.512 0.008 0.420 0.012 Hatchling tail length (mm) N 28.1 0.4 26.1 0.5 S 29.3 0.5 29.2 0.5 Values are shown as mean 1 standard error. Mea n values are from the raw data, not estimated marginal means.

PAGE 56

45 Chapter 3: On Intrinsic Growth and Juvenile Survival of Lizard Populations Along a Fine Scale Temperature Gradient: a Reciprocal Transplant Appro ach Abstract To understand the evolutionary and ecological signi ficance of geographic variation in life history traits, we must understand whether the patt erns are induced through plastic, extrinsic or adaptive, intrinsic responses. In lizards, particu larly in the genus Sceloporus geographic patterns have been studied extensively, giving us m any potential life history patterns in need of proximate explanations. Bergmann’s cline, which fi rst described the pattern of increasing body size in endotherms as environmental temperature dec reased, is one such pattern. In the lizard genus Sceloporus some species do and some species do not exhibit Be rgmann’s cline across their geographic range. Moreover, one species in p articular, S. undulatus, exhibits a reverse Bergmann’s cline at fine geographic scales and Berg mann’s cline at larger geographic scales. To begin examining how, and at what scale, life histor y tactics change from exhibiting plastic (null model) responses to that of adaptive responses, sma ll scale reciprocal transplant experiments must be conducted. We used reciprocal transplant e xperiments to examine the relative plastic and adaptive responses from populations that experi ence a 1 C difference in their monthly average temperatures. We specifically measured pre cipitation, ground cover heterogeneity, food availability, and potential activity periods of eac h population-environment treatment and examined their relationships with juvenile growth rates and survival. Two separate reciprocal transplant experiments were conducted along the latitudinal/en vironmental temperature gradient of the Florida peninsula. One experiment used populations of the Eastern Fence Lizard ( Sceloporus undulatus ), exhibiting cogradient variation in body size, an d the other used populations of the Florida Scrub Lizard ( S. woodi ), exhibiting countergradient variation in body siz e. In S.

PAGE 57

46 undulatus larger adult body sizes in the southern populatio n were not a result of faster extrinsic juvenile growth rates, although potential activity periods were greater in the southern environment. In fact, the extrinsic growth rates w ere similar between the north and south resident populations. Furthermore, we found population-spec ific, intrinsic, differences in juvenile growth rates and survival. The null hypotheses of cogradi ent variation in extrinsic juvenile growth rates and survival, therefore, were not supported. The a daptive differences in juvenile growth rates between populations were masked by plastic response s to the environment. The larger adult body sizes in the southern population can be explai ned by greater intrinsic survival that translated into greater extrinsic survival. In S. woodi similar adult body sizes between the north and so uth populations could be explained by faster intrinsic and extrinsic juvenile growth rates observed in the northern population when in their native enviro nment. We did not observe greater potential activity periods in the southern environment. Juve nile survival was not different between populations. The hypothesis of countergradient var iation in extrinsic growth rates, through intrinsic differences, therefore, was supported. In S. woodi the similarity in adult body sizes between populations is likely a result of adaptive responses. Even on fine geographic scales there appears to be complex relationships among env ironmental temperatures and trade-offs among life history traits of Sceloporus lizards. Key words Bergmann’s cline, Adult body size, squamates, Sceloporus intrinsic survival. Introduction To understand the evolutionary and ecological signi ficance of geographic variation in life history traits, we must understand whether the patt erns are induced through plastic, extrinsic, or adaptive, intrinsic responses. In lizards, particu larly in the genus Sceloporus geographic patterns have been studied extensively, giving us m any potential life history patterns in need of proximate explanations (Angilletta and Dunham 2003; Angilletta et al. 2004a; Angilletta et al. 2004b; Ashton and Feldman 2003; Ferguson et al. 198 0; Niewiarowski 1994; Niewiarowski and Angilletta 2008; Sears and Angilletta 2004; Tinkle and Dunham 1986). Large scale geographic

PAGE 58

47 patterns in life history tactics have been tested f or relationships with latitude and habitat types, f or example (Angilletta and Dunham 2003; Ashton and Fel dman 2003; Ferguson et al. 1980; Niewiarowski 1994; Tinkle and Dunham 1986), but the large scale relationships often fail to explain observed geographic variation adequately be cause many factors are involved simultaneously. For example, growth rates can be i nfluenced by food availability (Ballinger 1977; Ballinger and Congdon 1980; Dunham 1978; but see ; Sinervo and Adolph 1994; Smith 1998; but see Jones 1987 and Niewiarowski 1995) and temperatu re (Niewiarowski and Roosenburg 1993; Sinervo and Adolph 1994) as well as through intrins ically coded physiology (Angilletta 2001) and energy allocation rules (Berven 1982; Dunham et al. 1989; Ferguson and Talent 1993; Niewiarowski 2001; Niewiarowski and Roosenburg 1993 ). Because of the many environmental and intrinsic factors involved, even patterns that do explain observed geographic variation do not elucidate necessarily the mechanisms through which they occurred. Bergmann’s cline, which first described the pattern of increasing body size in endotherms as environmental temperature decreased, is one such pattern (Bergmann 1847; translation by James 1970). Bergmann’s cline generally holds for endotherms (Ashton 2002; Ashton et al. 2000; James 1970), but in some ectotherms, such as lizards, a reverse Bergmann’s cline is more common, although this pattern does not always occur (Ashton and Feldman 2003). In fact, in the lizard genus Sceloporus, some species do, and some species do not exhibit B ergmann’s cline across their geographic ranges (Angilletta et al. 2 004a; Ashton and Feldman 2003). The reasons for observing different patterns likely include int eractions between plastic and adaptive responses to the environmental temperature gradient. Indeed, comparative and manipulative experiments (Angilletta et al. 2004a; Ballinger 1977; Ballinger and Congdon 1980; Dunham 1978; Niewiarowski and Roosenburg 1993; Sears and Angille tta 2004; Sinervo and Adolph 1989; Sinervo and Adolph 1994; Smith 1998), ecological mo delling (Adolph and Porter 1993), and construction of energy budgets (Congdon et al. 1982 ; Niewiarowski 2001) have elucidated the influence of intrinsic and environmental variables, such as temperature and food availability, as well as the expected plastic responses to these var iables.

PAGE 59

48 Several hypotheses about the processes through whic h Bergmann’s cline is exhibited in Sceloporus result from this extensive body of work. These hy potheses often focus on growth rates because growth rates of lizards are inextrica bly linked to resulting life history suites. Growt h needs to be allocated for in energy budgets, affect s traits like age and size at maturity and adult body size, and can be constrained by survival (Adol ph and Porter 1993; Angilletta et al. 2004a; Niewiarowski 2001; Sears and Angilletta 2004; Stear ns and Koella 1986). Furthermore, growth rates in Sceloporus even developmental rates, are known to be intrins ically different between populations (Angilletta et al. 2004b; Ferguson et a l. 1980; Ferguson and Talent 1993; Niewiarowski and Angilletta 2008; Niewiarowski and Roosenburg 1993; Oufiero and Angilletta 2006; Storm and Angilletta 2007) and to respond pla stically to environmental differences (Niewiarowski and Roosenburg 1993; Sears and Angill etta 2004; Sinervo and Adolph 1989; Sinervo and Adolph 1994). Adaptive, intrinsic responses can result in counter gradient patterns, which may eliminate geographic variation because of adaptive compensati on and/or result in variation that is opposite what would be expected through a plastic response ( Conover and Schultz 1995). In Sceloporus undulatus adaptive changes in growth rate and adult body si ze often work in a countergradient fashion, with increased intrinsic growth found in p opulations that experience colder environments (Angilletta et al. 2004b; Oufiero and Angilletta 20 06). Even with this increased intrinsic growth, the extrinsic growth is reduced because of the cold er environmental temperatures, yet adult body sizes are also larger, resulting in a Bergmann’s cl ine (Sears and Angilletta 2004). The larger adult body sizes can be explained by extended growt h periods allowed by greater juvenile survival (Angilletta et al. 2004b; Sears and Angill etta 2004). Along the gradients of environmental temperature and/or latitude, the positive relations hip between juvenile survival and large adult body size over the geographic range of S. undulatus coincides with theory suggesting that survival should increase as activity periods decrea se because chances of predator contact decrease with decreased activity time (Adolph and P orter 1993). A cogradient pattern is what would be expected in a purely plastic response (Conover and Schultz 1995), although cogradient variation is not necessarily inconsistent with adaptation,

PAGE 60

49 hence, experiments such as reciprocal transplants t hat are designed to examine the relative adaptive and plastic responses of traits to environ mental gradients. Plastic responses also have been observed in Sceloporus undulatus and have been suggested as null models (Adolph and Porter 1993; Adolph and Porter 1996; Sears and Angi lletta 2004; Sinervo and Adolph 1994). Growth rates are influenced by temperature through constraints on thermoregulation and activity periods, and by food availability through constrain ts on energy assimilation. Growth rates generally increase plastically with increased tempe ratures and/or increased food availability (Adolph and Porter 1993; Adolph and Porter 1996; Si nervo and Adolph 1994). To begin examining how, and at what scale, life history tact ics change from exhibiting plastic (null model) responses to that of adaptive responses, small scal e reciprocal transplant experiments must be conducted. The relationships between environmental temperatures, food availability, intrinsic and extrinsic growth rates, and juvenile survival are t he focus of the reciprocal transplants conducted in our experiments. We studied populations of the Eastern Fence Lizard ( Sceloporus undulatus ) and the Florida Scrub Lizard ( S. woodi ) on a fine geographic scale along the latitudinal/ environmental temperature gradient of the Florida peninsula to ex amine the process through which the large scale countergradient variation in adult body sizes results. The large scale countergradient trend in adult body size that is observed among S. undulatus populations is also exhibited between these closely related species. In Florida, Sceloporus undulatus is from a relatively cooler habitat and exhibits larger adult body sizes than S. woodi (Hartmann 1993; McCoy et al. 2004; Mobley 1998; Robbins 2010). Within species, S. undulatus has larger adult body sizes in the southern population compared to the northern population (cog radient variation) and S. woodi has similar adult body sizes in the north and south populations (lack of variation consistent with countergradient variation; Robbins 2010). In other words, the observed geographic variation in adult body size between the north and south populat ions of each species is consistent with selective forces insufficient to result in Bergman’ s cline. Because these populations are separated by a mere ~2 in latitude and experience only a 1 C average difference between the sites (Robbins 2010), the observed geographic varia tion in adult body sizes should reflect that of

PAGE 61

50 a plastic response. The cogradient variation in S. undulatus is consistent with this hypothesis, but the lack of variation in S. woodi is not. We used reciprocal transplant experiments to examin e the relative plastic and adaptive responses from populations that experience a 1 C d ifference in their monthly average temperatures (Robbins 2010). We specifically measu red precipitation, ground cover heterogeneity, food availability, and potential act ivity periods of each population-environment treatment and examined their relationships with juv enile growth rates and survival. Two separate reciprocal transplant experiments were conducted al ong the latitudinal/environmental temperature gradient of the Florida peninsula. One experiment used populations Sceloporus undulatus and the other used populations of S. wood ). The fundamental assumption is that we should fi nd greater potential activity periods for populations experiencing the southern environments. Thus, in S. undulatus we expect both northern and southern populations t o exhibit faster extrinsic growth in the southern environment because of longe r activity periods, as long as food availability is equal. To achieve larger adult body sizes in th e southern environment, juvenile survival should be similar to that experienced in the northern envi ronment, or somewhat lower because of longer potential activity periods. If lower juvenile surv ival is observed in the southern environment, then growth rates would need to be sufficiently fast to compensate. In S. woodi we expect to see what would be consistent with a countergradient, adaptiv e response, with faster intrinsic and extrinsic growth rates in the northern population to achieve larger adult body sizes than expected (through a plastic response), unless juvenile survival is h igher for both populations in the northern environment. Materials and Methods Study species. Sceloporus woodi lives in open scrub habitats on remnant Pliocene a nd Pleistocene sand ridges in central Florida. Open sc rub habitats consist of sparse sand pines, oak shrubs, and extensive bare ground. S. woodi occurs in disjunct, genetically divergent populati ons along the Florida ridge (Clark et al. 1999; McCoy e t al. 2004) and is a rare species, although locally abundant (McCoy and Mushinsky 1992). S. undulatus is common in the southeastern

PAGE 62

51 United States (Conant 1975) and abundant in sandhil l habitats of central Florida. Sandhill habitats consist of long-leaf pines, turkey oaks, a nd ground cover of wiregrass and fallen pine needles (Myers and Ewel 1990b). These two species are closely related. Sceloporus woodi precinctive to Florida scrub, presumably diverged from S. undulatus when rising waters of the Pleistocene isolated sev eral populations on the sandy ridges of Florida (Clark e t al. 1999). Sceloporus woodi is in the eastern S. undulatus clade and is actually more closely related to S. undulatus than many previous designations of subspecies in the S. undulatus complex (Leache 2009; Leache and Reeder 2002). Furthermore, hybridization resulting in via ble offspring is known to occur between these species in isolated areas where their respective ha bitats are adjacent (Jackson 1973b; Robbins et al. 2010). Collection and housing of female lizards. Female lizards ( N =119) were collected from four populations in Florida, one northern and one s outhern population of each species. Collecting occurred from March to September in 2004. The nort hern populations were collected from Marion County. Each species was collected from the ir respective habitats, which included a S. undulatus population ( N =33) from N 2902’18”, W 8133’35”and a S. woodi population ( N =31) from N 2906’29”, W 8148’34”(Fig. 3.1). The southe rn populations of S. undulatus ( N =27) and S. woodi ( N =28) were collected in Hillsborough County, N 27 4 5’ 60”, W 82 15’ 07”, and Highlands County, N 27 37’ 07”, W 81 15’20”, respectively. Lizards were captured using a noosing technique, given a unique toe clip for identificati on (Waichman 1992), contained individually in a cotton bag or plastic-ware, and then collectively t ransported in a cooler kept at 20-30 C to the laboratory at the University of South Florida, Tamp a, Florida, USA. Each lizard was provided fresh water and crickets daily and housed in a container (30 x 17 x 12 cm) that included a sand substrate, water d ish, and plastic cover object for basking and refuge. Heat lamps maintained temperature gradient s within containers that averaged 31 C during the daytime portion of a 12/12 hr day/night cycle.

PAGE 63

52 Egg incubation and hatchling husbandry. Each lizard and housing was checked daily for oviposition. After oviposition, each clutch ( N =77) of eggs ( N =397) was placed in a glass jar (120 ml) and buried completely in vermiculite that was premixed to a water potential of –450 kPa. Water potential for vermiculite was determined by P ackard et al (1987). All vermiculite was oven-dried at 100 C for at least 4 hours prior to mixing with distill ed water. Each jar was covered with plastic kitchen wrap, sealed with a rubber ban d and placed in an incubator set at a constant 28 C. Vermiculite was replaced for each clutch on day 25 of incubation. Eggs in the incubator were checked daily for hatchl ings ( N =237). Each hatchling was marked with a unique combination of toe-clips (Waic hman 1992) and housed in a 38-liter (10 gallon) terrarium in the laboratory prior to their release in the field. Hatchling mortality is great est during the first few weeks after hatching for many reptiles (Crenshaw 1955; Iverson 1991; Tinkle 1967; Warner and Shine 2005), so hatchlings were ho used in the laboratory for eight weeks before being released to ensure successful mark-rec apture survival analyses. We provided water and crickets (dusted with vitamin/mineral mix) dail y for the hatchlings. For each individual, hatchling sex was recorded, and their SVL, TL, and mass (to the nearest 0.0001 g) measured before release. Male and female hatchlings did not differ in any phenotype (all P > 0.09), thus they were combined for all analyses. Reciprocal transplants. After housing the gravid females, incubating their eggs, and raising the hatchlings for eight weeks all under id entical conditions, hatchlings were released into the field under a reciprocal transplant design. At each site, the reciprocal transplant design included two 40 x 40 m enclosures constructed of a 61 cm aluminum flashing fence that was buried 13 cm into the ground and reinforced by meta l posts (electrical conduit) at 1.5 m intervals. A 1 m perimeter within each enclosure was cleared a nd mowed to inhibit climbing and jumping out of the enclosure. Hatchlings were released in the enclosures (from September to December 2004) in a split-clutch design with approximately h alf of each clutch being released at their site of capture as residents and half at the other respecti ve site (north or south). Reciprocal transplants were conducted within species between the north and south populations, resulting in four

PAGE 64

53 treatments per species – that of the northern popul ation released in the northern environment (NN; S. undulatus N=30; S. woodi N=23), the northern population released in the so uthern environment (NS; S. undulatus N=30; S. woodi N=20), the southern population released in the southern environment (SS; S. undulatus N=18; S. woodi N=13), and the southern population released in the northern environment (SN; S. undulatus N=20; S. woodi N=12). To increase sample sizes where necessary, hatchlings caught in the field (N=76) were used to supplement the lab raised hatchlings. Body sizes (SVL, and ma ss and TL relative to SVL) were not different between field caught and lab raised hatchlings from each site (p-values from ANOVA all > 0.05). Each of the four sites was methodically searched ap proximately every 10 days by walking around the inside perimeter of each enclosure and then zig -zagging through the enclosure in one direction, turning, and ziz-zagging through the enc losure in the perpendicular direction. Searches occurred between 900-1500 hours from September 2004 to February 2005, and lasted at least 8 weeks after the last hatchling was released. When hatchlings were sighted, they were captured by noosing, identified by their unique toe clip com bination, and their SVL, TL, and mass were measured with a ruler to the nearest 0.5 mm and wit h a Pesola spring scale to the nearest 0.05 g, respectively. With these data growth rates and su rvivorship associated with each treatment can be assessed (see Data Analyses section below). Environmental covariates. We measured precipitation for each environment, and ground cover heterogeneity, canopy cover, food avai lability, and potential activity periods for each population-environment treatment. Precipitation (m m) was measured with a rain gauge that was checked and emptied during each site visit. We meas ured ground cover heterogeneity and canopy cover at each point of a 16 point grid withi n each enclosure. Points were 10 meters apart. We surveyed each point in all four cardinal directi ons and used the average value as the sample unit. We measured ground cover heterogeneity using a vertical density board (Nudds 1977) from 5 meters away to eliminate spatial overlap of data collection. Heterogeneity was evaluated near the ground from 0 – 66 cm. Canopy cover was measur ed with a spherical densiometer from 1.3 m above ground and one meter from each point. Both vegetation measurements were estimated

PAGE 65

54 as percent cover. Food availability was measured t hroughout the mark-recapture experiment using an array of pitfall traps (15 traps at each o f the 4 sites; 5 traps per enclosure and 5 outside the enclosures) that were opened approximately once a month for five trapping periods between August and February 2004-2005. Each trapping perio d lasted 3-7 days. The index of total biomass per trap per day was estimated by summing t he lengths of the individual arthropods caught in each trap during each time period and div iding by the number of days open. Arthropods greater than 5 mm in length were conside red too large for consumption and not included in the analysis (Jackson 1973b). Potential activity periods were estimated between e ach lizard capture occasion. We followed the procedure in Grant and Dunham (1988) w ith slight modifications. Active lizard body temperatures were recorded for individuals of each population in their respective habitats ( S. undulatus N=116 for Ocala and N=68 for Balm; S. woodi N=91 for Ocala, N=63 for Avon) with a quick-read cloacal thermomether. If eggs were felt when a female lizard was palpated, that individual was considered gravid, and body temperat ures of gravid females were not used to calculate activity periods. Using 90% of the activ e lizard body temperatures we derived a minimum and maximum active lizard body temperature and used these limits to bracket the operative environmental temperatures measured by te mperature logger arrays placed at each site. These logger arrays consisted of five ibutto ns (Thermochron, model # DS1921; Maxim Integrated Products, Dallas Semiconductor, Sunnyval e, CA; www.ibutton.com ) in a cross pattern that was 1 m across. Some logger arrays were rando mly placed on the ground (using a random number table and a coordinated grid superimposed ov er a map of the site) and others on trees. Those placed on trees consisted of five loggers str ung together in a line and evenly spaced across a meter. The string of loggers was placed a round tree trunks in a spiral pattern, alternately at 0.5 and 1 m above the ground. A tre e logger array was used because both species use tree trunks to perch and bask, although Sceloporus undulatus is more arboreal than S. woodi At each site, all trees within a 60 x 80 meter a rea (including inside the enclosures) were tagged and randomly selected by number for placemen t of the tree logger array.

PAGE 66

55 Logger arrays were moved (randomly placed) during e ach site visit, and the data periodically downloaded. Temperatures were recorde d every 15 minutes (n=822,565). If 10% or more of the operative temperatures in at least 2 lo gger arrays during any 15 minute period were within the lizard body temperature minimum and maxi mum, potential for lizard activity was assumed. We summed these 15 minute periods to calc ulate hours per day of potential lizard activity and then summed the daily activity periods to calculate the total potential activity periods between each lizard capture occasion. If daily act ivity periods were missing because of logger malfunction we added the mean daily activity period of the particular capture interval for each missing day. Spatial autocorrelation of temperatur e loggers was also tested by comparing temperature variation within arrays to variation am ong arrays and no difference was found. We used 15 minute periods from 5 days between 1400-160 0 hours to account for the angle of the sun. We used periods with averages of 42 1 C, w hich is relatively high, allowing for temperature variation and only occurring when the s un is present. We tested if the variances were different with an ANOVA using standard deviate s from the mean as the response variable and array as the factor. Data analyses. Hatchling body sizes at release were compared betwe en populations, within species, using ANOVA. Source population was used as a factor and SVL was used as a covariate when analyzing TL and mass. Growth rates were assessed within species for infl uence of source population and growth environment and compared among treatments. Individ ual daily growth rates were calculated by subtracting SVL at release from SVL at last capture and divided by the number of days in between. Only individuals with at least 2 weeks be tween measurements were used to allow for measurable growth. Because differences between gro wth rates of lab raised and field caught hatchlings were assessed within treatments and none were found (all P > 0.2), they were pooled for all analyses. To test for relative influence o f source population and growth environment on growth rates an ANOVA was used. Source population, growth environment, and their interaction were factors with daily growth rates ( S. undulatus N=37; S. woodi N=39) as the dependent

PAGE 67

56 variable and SVL at release as covariate to control for size dependent growth. Body condition (residuals of mass relative to SVL) was also used a s a covariate for S. woodi because a difference between source populations was found. T o assess potential differences in reaction norms planned comparisons were also made between sp ecific treatments with ANOVA using treatment as factor and the covariates listed above for each species. Planned comparisons were conducted within species and included comparisons b etween resident populations and within source populations between growth environments. Survival was analyzed using Cormack-Jolly-Seber cap ture-recapture models in the information-theoretical framework (Burnham and Ande rson 2002; Lebreton et al. 1992) of the Program MARK (White and Burnham 1999). We first mo delled survival for source population and growth environment (using both as grouping variable s for factor-only models) to find populationenvironment specific survival, and then modelled su rvival with only covariates associated with source population and growth environment (no groupi ng variables for covariate-only models) to find which covariates might explain treatment speci fic survival. We chose a global model and assessed how well the model fit the data, then foun d the best candidate models of survival in the four population-environment treatments for each spe cies. Survival ( ) and recapture rates (p) were estimated using the step down approach (Lebret on et al. 1992) where p was modeled first, then and then p again. The global model was chosen a priori to include source population, growth environment, and their interaction as groupi ng variables, with time included additively because some environmental covariates could be anal yzed through time (activity periods, precipitation, and food availability). Covariates were not used in the global model because we were not interested in estimating survival after ac counting for covariates, but rather what true survival was in each treatment. Plus, goodness-offit tests do not yet allow their incorporation (Burnham and Anderson 2002). Model fit to the data was tested using a bootstrap method and where overdispersion was found, the overdispersion parameter ( ) was used for correction (Lebreton et al. 1992). Parsimony was assessed thr ough a maximum likelihood approach by the lowest Akaike Information Criteria (AICc) value wit h a bias-correction in case sample size was small with respect to the number of estimated param eters (Burnham and Anderson 2002). When

PAGE 68

57 was included ( S. undulatus =3.70; S. woodi =2.74), a Quasi-AICc value (QAICc) that accounts for was used to assess parsimony (Burnham and Anderson 2002). Starting with the global model, we found the most parsimonious model of p among candidate models. The models of p were assessed again with the most parsimonious model of to complete the step down method. The most parsimonious model of p, which wa s the intercept only model for both species, was then used as a constant among the candidate mod els of Factor-only models were assessed first. Among the candidate model set of f actor-only models, the relative influence of environment and population was assessed with the QA ICc values and further assessed with likelihood ratio (LR) tests. Likelihood ratio test s were performed between the most parsimonious model including both grouping variables (growth env ironment and source population) and the sub-models containing only one grouping variable. The probability of survival was estimated for each population-environment treatment through model averaging of (Burnham and Anderson 2002). Covariate-only models were then assessed fo r the influence of each habitat and morphological variable on survival. Habitat variab les were included in the models if they were different between treatments and they were not corr elated with each other. Body sizes were also used as covariates at the individual level to accou nt for differences among individuals and any differences that may exist between source populatio ns. Each covariate was assessed for its influence on survival with QAICc values, and furthe r assessed using model averaged beta values (B) with unconditional confidence intervals (CI). Model averaging occurred among models that had QAICc values (the difference between the QAICc val ue for the particular model and the QAICc value for the most parsimonious model) of 2.0 or less. If the unconditional confidence interval (CI) did not include zero, the effect was considered statistically significant (Burnham and Anderson 2002). Habitat variables were compared between populationenvironment treatments, where possible, and those that showed differences were te sted for correlations among treatments within each species. Heterogeneity and canopy cover were examined among population-environment treatments with mixed model analyses using average values at each point as dependent variables (All N =64). Activity periods were examined among popula tion-environment treatments

PAGE 69

58 with mixed model analyses using the estimated activ ity periods associated with each populationenvironment treatment at each lizard capture interv al as sample units ( Sceloporus undulatus N =48; S. woodi N =40). Treatment and time interval were used as fac tors. Precipitation was examined between environments (not between populati on-environment treatments) with a mixed model analysis using total precipitation measured f or each lizard capture interval as sample units ( S. undulatus N =24; S. woodi N =20). Environment (north/south) and time interval were used as factors. Food availability was examined separately at two levels, between environments ( N =30 per sampling period) and between enclosures within environments ( i.e. at the populationenvironment treatment level, N =10 per sampling period), with repeated-measures AN OVAs across the 5 sampling periods. The two levels were analyzed separately because 5 traps at each site were outside of the enclosures. Total biomass per trap per day was the dependent variable. Environment or population-environment treatment was used as the factor, respectively. The biomass index data was log-transformed to meet assu mptions of the ANOVA. Correlation analyses among population-environment treatments we re used to further assess any relationships between growth, survival, and environ mental variables. Results Sceloporus undulatus. Hatchling body sizes at release were not affected b y source population in Sceloporus undulatus (Table 3.1) The observed resident growth rates also were not different between populations of S. undulatus (Fig. 3.2; Table 3.2). In the reciprocal transpla nt experiments, however, hatchling growth rates were i nfluenced by source population and marginally by growth environment. Individuals from the northern population grew faster than those from the southern population in both environm ents, which suggests an intrinsic influence associated with source population (Fig. 3.2; Table 3.2). Both populations of S. undulatus also grew faster in the southern, warmer environment tha n in the northern cooler environment, however, the trend was only marginally significant (Table 3.2). No source population x growth environment interaction was found, but planned pair wise comparisons did show differences in the shape of population-specific reaction norms. The s outhern population of S. undulatus exhibited

PAGE 70

59 hatchling growth rates that were significantly slow er in the cooler, northern environment than in the warmer, native environment, but the northern po pulation did not exhibit different growth rates between growth environments (Table 3.3). These res ults suggest a larger reaction norm (greater plasticity) associated with growth rates in the sou thern population. Together, these results suggest a shift in the overall reaction norm as wel l as in the shapes. Correlation analyses also found a negative relationship between growth rates and canopy cover in S. undulatus (Table 3.4). In the survival analysis, the intercept-only model was the most parsimonious recapture probability model before and after finding the most parsimonious survival probability model. The intercept only recapture probability model was ther efore used for all the survival models. The model averaged recapture probability was (estimated probability 1SE) 0.52 0.07 for Sceloporus undulatus All survival models that included a time parame ter had QAICc values greater than 10 ( QAICc =14-98), suggesting that differences in surv ival among intervals were of little influence to the overall modelling of surviv al. Thus, models including a time parameter were excluded. Survival was influenced more by source population t han growth environment in S. undulatus Model averaged survival was greater for the sout hern population in both growth environments (Fig. 3.3), which suggests an intrinsi c influence. Source population had an effect size of 51% (an effect size of 100% being equivalen t to one population having a survival probability twice that of the other) in the norther n growth environment and 41% in the southern growth environment. So, on average the southern po pulation had a monthly survival rate that was 46% higher than the northern population. Further support for a greater source population eff ect on survival than growth environment was found in likelihood ratio tests amo ng the models. The survival model with only source population was more parsimonious than the po pulation + environment model, which was more parsimonious than the model with only growth e nvironment (Table 3.5). Likelihood ratio tests between the full population + environment mod el and the source population ( P = 0.88) or growth environment ( P = 0.05) only models suggest that the growth enviro nment only model is a significantly worse descriptor of survival than the source population only model.

PAGE 71

60 Environmental and morphological covariates explaine d survival among treatments better than the general population-environment models in Sceloporus undulatus (QAICc values in Tables 3.5 & 3.6). Mass was used as an individual covariate to reflect body condition because there was not a difference in SVL between populatio ns (see Table 3.1). Individual mass (B = 2.56, CI = 0.41 to 4.71), average growth rates (B = -33.76, CI = -71.38 to 3.86), food availability (B = 4.86, CI = -3.43 to 13.16), and activity perio ds (B = -0.003, CI = -0.016 to 0.010) all influenced survival according to the most parsimoni ous S. undulatus models with QAICc values of 2.0 or less (Table 3.6). However, mass was the only covariate with a CI that did not include zero. Heterogeneity was greater in the south for Sceloporus undulatus and canopy cover was greater in the north. Thus, canopy cover showed a trend that was opposite to that of ground cover heterogeneity (Table 3.7). Precipitation was not significantly different between the north and south environments, however the trend showed gr eater precipitation in the south (Table 3.7). Activity periods were greater in the southern envir onment for both populations of S. undulatus (Table 3.7). Correlations among population-environ ment treatments included a positive correlation between heterogeneity and food availabi lity (r = 0.98, P = 0.021), and a negative correlation between canopy cover and activity perio d (r = -0.96, P = 0.040). Sceloporus woodi. Hatchlings from the southern population of Sceloporus woodi were longer and heavier than those from the northern pop ulation (Table 3.1). The observed resident growth rates also were marginally different between populations of S. woodi with faster growth rates associated with the northern population (Fig. 3.4; Table 3.8). In the reciprocal transplant experiments hatchling growth rates were only influe nced by source population. Individuals from the northern population grew faster than those from the southern population, regardless of growth environment, though the difference was marginally s ignificant (Fig. 3.4; Table 3.8). The trend in growth rates across growth environments was not sig nificant (Table 3.8), but notably, was opposite the trend found in S. undulatus with individuals from both populations growing fa ster in the cooler, northern environment. Furthermore, bod y condition was positively associated with

PAGE 72

61 growth rates (Table 3.8). No source population x g rowth environment interaction was found (Table 3.8), and planned pairwise comparisons suppo rted a conclusion that the shapes of the population-specific reaction norms were similar (Ta ble 3.9). Correlation analyses found no environmental correlates with growth rates in S. woodi (Table 3.4). In the survival analysis, the intercept-only model was the most parsimonious recapture probability model before and after finding the most parsimonious survival probability model. The intercept only recapture probability model was ther efore used for all the survival models. The model averaged recapture probability was (estimated probability 1SE) 0.39 0.06 for S. woodi All survival models that included a time parameter had QAICc values greater than 10 ( QAICc=21-90), suggesting that differences in surviv al among intervals were of little influence to the overall modeling of survival. Thus, models inc luding a time parameter were excluded. Survival was influenced more by growth environment than by source population in Sceloporus woodi Model averaged survival was greater in the south ern environment for both the northern and southern populations (Fig. 3.5). Sour ce population had little to no effect on survival in any environment (Fig. 3.5), however, growth envi ronment had an effect size of 22% on the northern population and 17% on the southern populat ion. Likelihood ratio tests between the full source population + growth environment model and th e source population or growth environment only models suggested no differences between these models (Table 3.10; P -values > 0.1). The order of parsimony (growth environment only > sourc e population + growth environment > source population only), however, was in the opposite dire ction among the three models, compared to the survival models of S. undulatus Environmental and morphological covariates explaine d survival among treatments better than the general population-environment models in S. woodi also (QAICc values in Tables 3.10 & 3.11). Mass was used as an individual covariate be cause there were differences between populations in SVL, mass, and body condition (Table 3.1) and mass was correlated with SVL (P<0.001) and body condition (P=0.016). Individual mass (B = 5.75, CI = -2.36 to 13.86), average growth rates (B = -11.14, CI = -54.93 to 32 .66), heterogeneity (B = -0.02, CI = -0.05 to 0.02), and food availability (B = 0.28, CI = -0.88 to 1.45) were the covariates in the most

PAGE 73

62 parsimonious S. woodi models (Table 3.11), however, all confidence inter vals included zero. Heterogeneity and canopy cover were both greater in the north for S. woodi Thus, the parallel trend between canopy cover and heterogeneity is dif ferent than the opposing trend between canopy cover and heterogeneity found between S. undulatus habitats (Table 3.7). Precipitation also was not significantly different between the no rth and south environments for S. woodi however the trend showed greater precipitation in t he south (Table 3.7). Activity periods did not differ between environments for both populations of S. woodi (Table 3.7). Correlations among source population-growth environment treatments inc luded a positive correlation between heterogeneity and canopy cover (r = 0.99, P = 0.011). Summary of overall trends. In Sceloporus undulatus extrinsic growth rates of residents in their native environments were similar but faster intrinsic growth rates were associated with the northern population. Greater i ntrinsic survival rates were associated with the southern population. In S. woodi hatchlings from the northern population had great er body sizes at hatching, faster extrinsic growth rates (residen ts in their native environments), and faster intrinsic growth rates. Greater intrinsic survival rates were associated with the southern growth environment, but not significantly. Across all tre atments and both species, survival was negatively related to activity periods, which was c onsistent with the negative, although nonsignificant, relationship within each species (Tabl e 3.4). Survival was not directly related with any environmental variables within either species (Tabl e 3.4). Discussion Minimal geographic variation in life history tactic s can have significant and complex intrinsic underpinnings that are masked by the inte raction between plastic and adaptive responses to the environment. These lizard populat ions are separated by a mere 2 latitudinal distance, which corresponds to experiencing a 1 C average monthly difference in temperature (Robbins 2010). In Sceloporus undulatus the larger adult body sizes in the southern popul ation were not a result of faster extrinsic juvenile grow th rates, although potential activity periods were

PAGE 74

63 greater in the southern environment. In fact, the extrinsic growth rates were similar between the north and south resident populations. Furthermore, we found population-specific, intrinsic, differences in juvenile growth rates and survival. The null hypotheses of cogradient variation in extrinsic juvenile growth rates and similar surviva l, therefore, were not supported. The larger adult body sizes in the southern population can be explained by greater intrinsic survival that translated into greater extrinsic survival. In S. woodi the similar adult body sizes between the north and south populations could be explained by f aster intrinsic and extrinsic juvenile growth rates observed in the northern population when in t heir native environment. We did not observe greater potential activity periods in the southern environment. Juvenile survival, although slightly higher for both populations in the southern environ ment, was not different between populations. The hypotheses of countergradient variation in extr insic growth rates, through intrinsic differences, and similar survival, therefore, were supported. In S. woodi the similarity in adult body sizes between populations is likely a result o f adaptive responses. Reaction norms and effects of environmental variabl es in Sceloporus undulatus In Sceloporus undulatus the effects of both source population and growth environment on growth rates reveal adaptive differences between populatio ns that are masked by plastic responses to the environment. The effect of source population o n growth rates suggests a shift in reaction norms between populations with individuals from the cooler, northern environment growing faster than those from the warmer, southern environment re gardless of the growth environment experienced (Fig. 3.2; Table 3.2 ). Thus, a popula tion level adaptive response has occurred between these populations. Although we observed fa ster growth from both source populations when in the southern environment, the observed geog raphic variation of resident growth rates was statistically indistinguishable and consistent with countergradient variation (Table 3.2). The effect of growth environment on growth rates influe nced the southern population more than the northern population. The southern population had s ignificantly faster growth in the warmer, southern environment compared with that in the cool er, northern environment (Table 3.3). The greater plastic response of the southern population compared to the northern population implies

PAGE 75

64 population-specific reaction norms of different sha pe. Overall, the plasticity of growth rates may have been influenced by canopy cover, activity peri od, and/or food availability. Although activity period was not directly associated with growth rate s among treatments, growth rates were directly associated, negatively, with canopy cover, which wa s negatively associated with activity period. The positive relationship between growth and activi ty is at least consistent with this hypothesis (Table 3.4). Food availability was not associated with growth rates at the treatment level, but the trends between growth environments were similar. F ood availability was greater in the southern environment where growth rates were also faster (Ta ble 3.7, Fig. 3.2). To understand the full extent to which these reaction norms are different and associated with each environmental variable, studies examining growth rates in multipl e environments are warranted. Survivorship was influenced by source population su ggesting possible adaptive significance, although maternal effects cannot be r uled out. Growth environment did not influence survival in the general models, and corre lation analyses found no relationships between survival and any environmental variable within Sceloporus undulatus (Table 3.4). Indeed, individual mass is the only covariate that signific antly influenced survival in the mark-recapture covariate models. If there is some adaptive signif icance, it may be linked simply to intrinsic survival or behavioral modifications that decrease conspicuousness to predators. The higher survival of the southern population may also be ass ociated with the greater food availability in the southern environment. We can rule out that greater food availability directly caused greater survival because the greater survival of the southe rn population was observed in both environments. The gravid females that were initial ly captured in the southern environment, however, may be higher quality parents and have hig her quality offspring (Roff 1997). We did not have the data to test specifically for a delayed re sponse to food availability. We do note, however, that there was no difference in offspring body sizes at release between the north and south populations. Illuminating population-specifi c reasons for this survival differential necessitates further study. The hypothesis that longer activity periods are ass ociated with lower juvenile survival was not supported by our data. In Sceloporus undualtus potential activity periods were greater in the

PAGE 76

65 southern environment, but extrinsic juvenile surviv al was not lower within either population because of the southern environment. Instead, popu lation-specific, intrinsic, juvenile survival was higher in the southern population regardless of env ironment. Perhaps intrinsic survival increased through a compensatory reaction to selection, but i t does not appear to be directly related to activity periods. Reaction norms and effects of environmental variabl es in Sceloporus woodi In Sceloporus woodi the observed geographic variation in resident pop ulation growth rates also followed a countergradient, with faster growth rate s observed in individuals from the northern population in their native, cooler environment (Fig 3.4; Table 3.8). This observation is consistent with the expected adaptive response. Faster growth rates occurred also for individuals from the northern population, regardless of the growth envir onment experienced (Fig. 3.4; Table 3.8). Reaction norms therefore marginally shifted in the same direction as they did in S. undulatus (Fig. 3.4). The shapes, however, of the population-speci fic reaction norms were similar in S. woodi (Fig. 3.4; Table 3.9). Notably, the influence of g rowth environment, although non-significant, was opposite in direction to that found in S. undulatus Both populations tended to grow faster when experiencing the cooler environment (Fig. 3.4). Bo dy condition, which was different between source populations at time of release, appears to b e the most influential variable on growth rates. Individuals with greater body condition had faster growth rates. The few differences that did exist between the environments, heterogeneity and canopy cover, had no discernable relationship with growth rates among treatments. Also, we should hav e observed longer activity periods in the southern environment, however, estimated potential activity periods were not different between treatments or environments. As a result, the hypot hesis that longer activity periods are associated with lower juvenile survival could not b e tested within S. woodi because potential activity periods were not different between the pop ulation-environment treatments. Interestingly, however, survivorship also did not differ among tre atments nor was it considerably influenced by any morphological or environmental covariates.

PAGE 77

66 Potential activity periods are estimated by compari ng the range of active body temperatures to the range of microhabitat temperatu res available. Potential activity periods, therefore, change according to the range of active body temperatures. In Sceloporus undulatus active body temperatures are similar among populati ons that span its large geographic range (Andrews 1998), which is why potential activity per iods generally increase as environmental temperatures increase. It is possible that active body temperatures are different among populations of S. woodi Also, the Florida scrub habitat associated with S. woodi has relatively sparse canopy cover, but is spatially heterogeneous with regard to the low lying vegetation. With much direct sunlight, microhabitat availability may shift spatially, but result in similar overall availability even when average environmental temper atures change. Examining these hypotheses would be an important next step in eluci dating the reasons for the unexpected similar potential activity periods observed between the pop ulation-environment treatments of S. woodi Large scale versus small scale trends in adult body sizes. Because life history tactics of adults in these particular populations w ere previously studied, we can examine how relationships among traits such as adult body size, juvenile growth rates, and juvenile survival fit within the larger body of theory. On a large geogr aphic scale, Sceloporus undulatus exhibits countergradient variation in adult body size and in trinsic growth rates. As environmental temperatures decrease adult body sizes increase and extrinsic growth rates decrease while intrinsic growth rates increase in a compensatory f ashion. Therefore, the larger adult body sizes are achieved through higher juvenile survival rates (Angilletta et al. 2004b; Sears and Angilletta 2004); Sears and Angilletta 2004), not faster extri nsic growth. The populations studied here were examined on a relatively small geographic scale. A dult body sizes in these two S. undulatus populations follow a cogradient pattern, larger adu lt body sizes in the warmer environment, and in these two S. woodi populations follow a countergradient pattern (lack of variation). Among these Florida populations, as environmental temperature d ecreases S. undulatus shows an increase in intrinsic growth rates and no change in extrinsic g rowth rates, while adult body size and population-specific survival decreases. In S. woodi as environmental temperature decreases we

PAGE 78

67 see an increase in intrinsic and extrinsic growth r ates, and no change in adult body size or juvenile survival. The relationships among environ mental temperatures, adult body sizes, growth rates, and survival on a fine geographic scale are not consistent with the large scale trends. Ignoring the environmental temperature gradient, th e relationships among adult body size, extrinsic growth rates, and juvenile survival are, however, consistent with the large scale explanations, but only in Sceloporus undulatus Because of greater survival, growth could occur during a longer time period to achieve the larger a dult body sizes (e.g. Angilletta et al. 2004b; Sears and Angilletta 2004). Intrinsic growth rates however, are already changing in a manner consistent with the large scale trends associated w ith the environmental temperature gradient. Populations experiencing cooler environments exhibi ted faster intrinsic growth rates, which actually occurred in both S. undulatus and S. woodi In S. woodi however, it was not higher survival that allowed more time to grow to larger a dult body sizes. Instead, faster intrinsic growth rates in the north resulted in faster extrinsic gro wth rates and subsequently larger adult body sizes than expected. As a result, adult body sizes were similar between north and south populations. On fine geographic scales there appea rs to be complex relationships among environmental temperatures and trade-offs among lif e history traits of Sceloporus lizards. Conclusions. In conclusion, even small environmental gradients c ause adaptive population level responses that can result in count ergradient variation where observed geographic variation appears non-existent, as in th e extrinsic growth rates of Sceloporus undulatus or the adult body sizes of S. woodi This study reveals that population-specific surv ival, which is likely intrinsic, may play an underappreci ated role in life history variation. It is especia lly the case when differences in intrinsic survival can be found between populations in proximity to each other. Between populations within both specie s, although stronger in S. undulatus higher intrinsic growth rates were associated with lower i ntrinsic and extrinsic juvenile survival in the cooler environment. How do we end up with the larg e scale trend of higher intrinsic growth rates associated with higher extrinsic juvenile survival in cooler environments? Is it strictly environmental pressures, such as predation or compe tition that ease up in much northern

PAGE 79

68 populations to increase extrinsic juvenile survival ? Does intrinsic juvenile survival increase after some geographic threshold is reached? Or is it both ? This study only begins to examine the process through which life histories switch from th e plastic responses to the adaptive responses that ultimately result in large scale geographic va riation. It appears that intrinsic growth rates ar e under strong selection and may be one of the first traits to adapt. Acknowledgements Many thanks go to the undergraduate army that helpe d us collect these data and care for the lizards in the laboratory, especially Lorelei S traub. We thank Tom Raffel for statistical help and Ray Martinez for invaluable engineering of spec ialized tools. We also thank Gary Huxel and Gordon Fox for reviewing this manuscript, for it ha s been much improved. This research was partially funded by Sigma Xi Grants-in-Aid of resea rch to TRR. Lizards were collected under collection permit WX05107 issued by the State of Fl orida Fish and Wildlife Conservation Commission. All protocols were reviewed and accept ed by the USF Institutional Animal Care and Use Committee, IACUC file #2778.

PAGE 80

69 References Adolph, S. C., and W. P. Porter. 1993. Temperature, activity, and lizard life histories. American Naturalist 142:273-295. —. 1996. Growth, seasonality, and lizard life histo ries: Age and size at maturity. Oikos 77:267278. Andrews, R. M. 1998. Geographic variation in field body temperature of Sceloporus lizards. Journal of Thermal Biology 23:329-334. Angilletta, M. J. 2001. Variation in metabolic rate between populations of a geographically widespread lizard. Physiological and Biochemical Zo ology 74:11-21. Angilletta, M. J., and A. E. Dunham. 2003. The temp erature-size rule in ectotherms: Simple evolutionary explanations may not be general. Ameri can Naturalist 162:332-342. Angilletta, M. J., P. H. Niewiarowski, A. E. Dunham A. D. Leache, and W. P. Porter. 2004a. Bergmann's clines in ectotherms: Illustrating a lif e-history perspective with sceloporine lizards. American Naturalist 164:E168-E183. Angilletta, M. J., T. D. Steury, and M. W. Sears. 2 004b. Temperature, growth rate, and body size in ectotherms: Fitting pieces of a life-history puz zle. Integrative and comparative biology 44:498-509. Ashton, K. G. 2002. Patterns of within-species body size variation of birds: strong evidence for Bergmann's rule. Global Ecology and Biogeography 11 :505-523. Ashton, K. G., and C. R. Feldman. 2003. Bergmann's rule in nonavian reptiles: Turtles follow it, lizards and snakes reverse it. Evolution 57:1151-11 63. Ashton, K. G., M. C. Tracy, and A. de Queiroz. 2000 Is Bergmann's rule valid for mammals? American Naturalist 156:390-415. Ballinger, R. E. 1977. Reproductive strategies fo od availability as a source of proximal variation in a lizard. Ecology 58:628-635. Ballinger, R. E., and J. D. Congdon. 1980. Food res ource limitation of body growth rates in Sceloporus scalaris (Sauria, Iguanidae). Copeia 198 0:921-923.

PAGE 81

70 Bergmann, C. 1847. Uber die Verhaltnisse der Warmeo konomie der Thiere zu ihrer Grosse. Gottinger Studien 3:595-708. Berven, K. A. 1982. The genetic basis of altitudina l variation in the wood frog Rana sylvatica. 1. An experimental analysis of life history traits. Ev olution 36:962-983. Burnham, K. P., and D. R. Anderson. 2002, Model sel ection and multimodal inference: a practical information-theoretic approach. New York, NY, USA, Springer-Verlag. Clark, A. M., B. W. Bowen, and L. C. Branch. 1999. Effects of natural habitat fragmentation on an endemic scrub lizard ( Sceloporus woodi ): an historical perspective based on a mitochondrial DNA gene genealogy. Molecular Ecology 8:1093-1104. Conant, R. 1975, A field guide to the reptiles and and amphibians of eastern and central North America Boston, Houghton Mifflin. Congdon, J. D., A. E. Dunham, and D. W. Tinkle. 198 2. Energy budgets and life histories of reptiles. Biology of Reptilia 13:233-271. Conover, D. O., and E. T. Schultz. 1995. Phenotypic similarity and the evolutionary significance of countergradient variation. Trends in Ecology & Evol ution 10:248-252. Crenshaw, J., W, Jr. 1955. The life history of the southern spiny lizard, Sceloporus undulatus undulatus Latreille. American Midland Naturalist 54 :257-298. Dunham, A. E. 1978. Food availability as a proximat e factor influencing individual growth rates in Iguanid lizard Sceloporus merriami Ecology 59:770-778. Dunham, A. E., B. W. Grant, and K. L. Overall. 1989 Interfaces between biophysical and physiological ecology and the population ecology of terrestrial vertebrate ectotherms. Physiological Zoology 62:335-355. Ferguson, G. W., C. H. Bohlen, and H. P. Woolley. 1 980. Sceloporus undulatus comparative life-history and regulation of a kansas populations Ecology 61:313-322. Ferguson, G. W., and L. G. Talent. 1993. Life histo ry traits of the lizard Sceloporus undulatus from 2 populations raised in a common laboratory en vironment. Oecologia 93:88-94. Grant, B. W., and A. E. Dunham. 1988. Thermally imp osed time constraints on the activity of the desert lizard Sceloporus merriami Ecology 69:167-176.

PAGE 82

71 Hartmann, P. P. 1993. Demography of a population of the Florida scrub lizard ( Sceloporus woodi ) in a sand pine scrub on the Lake Wales Ridge of cen tral Florida, Masters Thesis, University of South Florida, Tampa. Iverson, J. B. 1991. Patterns of survivorship in tu rtles (Order Testudines). Canadian Journal of Zoology-Revue Canadienne De Zoologie 69:385-391. Jackson, J. F. 1973. Phenetics and ecology of a nar row hybrid zone. Evolution 27:58-68. James, F. C. 1970. Geographic size variation in bir ds and its relationship to climate. Ecology 51:365-&. Jones, S. M., S. R. Waldschmidt, and M. A. Potvin. 1987. An experimental manipulation of food and water growth and time-space utilization of ha tchling lizards ( Sceloporus undulatus ). Oecologia 73:53-59. Leache, A. D. 2009. Species Tree Discordance Traces to Phylogeographic Clade Boundaries in North American Fence Lizards ( Sceloporus ). Systematic Biology 58:547-559. Leache, A. D., and T. W. Reeder. 2002. Molecular sy stematics of the Eastern Fence Lizard ( Sceloporus undulatus ): A comparison of parsimony, likelihood, and Bayes ian approaches. Systematic Biology 51:44-68. Lebreton, J. D., K. P. Burnham, J. Clobert, and D. R. Anderson. 1992. Modeling survival and testing biological hypotheses using marked animals a unified approach with case studies Ecological Monographs 62:67-118. McCoy, E. D., P. P. Hartmann, and H. R. Mushinsky. 2004. Population biology of the rare Florida scrub lizard in fragmented habitat. Herpetologica 6 0:54-61. McCoy, E. D., and H. R. Mushinsky. 1992. Rarity of organisms in the sand pine scrub habitat of Florida. Conservation Biology 6:537-548. Mobley, E. R. 1998. A base line population study of the southern fence lizard, Sceloporus undulatus undulatus in central Florida., University of Central Florid a, Orlando, Florida. Myers, R. L., and J. L. Ewel. 1990. Scrub and high pine. Ecosystems of Florida.:150-193. Niewiarowski, P. H. 1994. Understanding geographic life-history variation in lizards. Lizard ecology: Historical and experimental perspectives:3 1-49.

PAGE 83

72 —. 1995. Effects of supplemental feeding and therma l environment on growth rates of eastern fence lizards, Sceloporus undulatus Herpetologica 51:487-496. —. 2001. Energy budgets, growth rates, and thermal constraints: Toward an integrative approach to the study of life-history variation. American Na turalist 157:421-433. Niewiarowski, P. H., and M. J. Angilletta. 2008. Co untergradient variation in embryonic growth and development: do embryonic and juvenile performa nces trade off? Functional Ecology 22:895-901. Niewiarowski, P. H., and W. Roosenburg. 1993. Recip rocal transplant reveals sources of variation in growth rates of the lizard Sceloporus undulatus Ecology 74:1992-2002. Nudds, T. D. 1977. Quantifying the vegetative struc ture of wildlife cover. Wildlife Society Bulletin 5:113-117. Oufiero, C. E., and M. J. Angilletta. 2006. Converg ent evolution of embryonic growth and development in the eastern fence lizard ( Sceloporus undulatus ). Evolution 60:1066-1075. Robbins, T. R. 2010. Geographic variation in life h istory tactics, adaptive growth rates, and habitat-specific adaptations in phylogenetically si milar species: the eastern fence lizard, Sceloporus undulatus and the Florida scrub lizard, Sceloporus woodi Dissertation, University of South Florida, Tampa. Robbins, T. R., J. N. Pruitt, L. E. Straub, E. D. M cCoy, and H. R. Mushinsky. 2010. Transgressive aggression in Sceloporus hybrids confers fitness through advantages in male agonistic encounters. Journal of Animal Ecology 79:137-147. Roff, D. A. 1997, Evolutionary Quantitative Genetic s. New York, NY, Chapman & Hall. Sears, M. W., and M. J. Angilletta. 2004. Body size clines in Sceloporus lizards: Proximate mechanisms and demographic constraints. Integrative and Comparative Biology 44:433442. Sinervo, B., and S. C. Adolph. 1989. Thermal sensit ivity of growth rate in hatchling Sceloporus lizards environmental, behavioral, and genetic as pects. Oecologia 78:411-419. —. 1994. Growth plasticity and thermal opportunity in Sceloporus lizards. Ecology 75:776-790.

PAGE 84

73 Smith, G. R. 1998. Habitat-associated life history variation within a population of the striped plateau lizard, Sceloporus virgatus. Acta Oecologic a-International Journal of Ecology 19:167-173. Stearns, S. C., and J. C. Koella. 1986. The evoluti on of phenotypic plasticity in life-history traits predictions of reaction norms for age and size at m aturity. Evolution 40:893-913. Storm, M. A., and M. J. Angilletta. 2007. Rapid ass imilation of yolk enhances growth and development of lizard embryos from a cold environme nt. Journal of Experimental Biology 210:3415-3421. Tinkle, D. W. 1967, The life and demography of the side-blotched lizard, Uta stansburiana : Miscellaneous Publications, v. No. 132. Ann Arbor, MI, Museum of Zoology, University of Michigan. Tinkle, D. W., and A. E. Dunham. 1986. Comparative life histories of two syntopic Sceloporine lizards. Copeia 1986:1-18. Waichman, A. V. 1992. An alphanumeric code for toe clipping amphibians and reptiles. Herpetological Review 23:19-21. Warner, D. A., and R. Shine. 2005. The adaptive sig nificance of temperature-dependent sex determination: Experimental tests with a short-live d lizard. Evolution 59:2209-2221. White, G. C., and K. P. Burnham. 1999. Program MARK : survival estimation from populations of marked animals. Bird Study 46:120-138.

PAGE 85

74 n n r nrn Figure 3.1. Map of reciprocal transplant locations. Reciproc al transplants were conducted between the north and south populations within each species.

PAGE 86

75 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 N SGrowth rate (mm/day)Latitude of Environment 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 N SGrowth rate (mm/day)Latitude of Environment Figure 3.2. Growth rates of resident and reciprocally transpl anted hatchlings of S. undulatus in warmer, southern and cooler, northern habitats. Po ints are estimated marginal means. Error bars represent 1 standard error. (N) refers to Nort h and (S) refers to South with respect to source population.

PAGE 87

76 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 N SMonthly SurvivorshipLatitude of Environment 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 N SMonthly SurvivorshipLatitude of Environmentr r Figure 3.3. Survivorship of resident and reciprocally transpl anted hatchlings of Sceloporus undulatus in warmer, southern and cooler, northern habitats. Points are survival estimates from the program MARK. Error bars represent 1 standard error. (N) refers to North and (S) refers to South with respect to source population.

PAGE 88

77 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 N SGrowth rate (mm/day)Latitude of Environment 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 N SGrowth rate (mm/day)Latitude of Environment Figure 3.4. Growth rates of resident and reciprocally transpl anted hatchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats. Points are estimated marginal means. Error bars represent 1 standard error. (N) refers t o North and (S) refers to South with respect to source population.

PAGE 89

78 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 N SMontlhy SurvivorshipLatitude of Environment 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 N SMonthly SurvivorshipLatitude of Environment Figure 3.5. Survivorship of resident and reciprocally transpl anted hatchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats. Points are survival estimates from the program MARK. Error bars represent 1 standard erro r. (N) refers to North and (S) refers to South with respect to source population.

PAGE 90

79 Table 3.1. Effects of source population on morphological tra its at time of release into the field. Sceloporus undulatus Sceloporus woodi Mean 1SE ANOVA Mean 1SE ANOVA Trait at release Ocala (N) Balm (S) F P Ocala (N) Avon (S) F P SVL (mm) 31.2 0.4 31.0 0.5 0.14 0.71 28.3 0.5 30.7 0.7 7.02 0.01 Mass (g) 1.29 0.05 1.22 0.06 0.92 0.34 0.86 0.06 1.18 0.07 11.92 < 0.01 Body condition (mm/g) 1.28 0.02 1.24 0.02 2.5 1 0.12 0.95 0.02 1.04 0.03 6.98 0.01

PAGE 91

80 Table 3.2 ANOVA results for growth rates among resident po pulations and reciprocally transplanted hatchlings of Sceloporus undulatus Residents Reciprocal transplant Source F P F P SVL (mm) 1.350 0.266 0.823 0.371 Source population 0.039 0.847 5.408 0.027 Growth environment 3.553 0.069 Pop x env 0.811 0.374 SVL refers to snout-vent-length of hatchlings at time of release.

PAGE 92

81 Table 3.3. Results of post-hoc ANOVAs on population-specific (North and South) growth rates among growth environments for Sceloporus undulatus Source population North South Source F P F P SVL (mm) 1.369 0.260 0.105 0.750 Growth environment 0.424 0.525 14.566 0.002 This analysis tests plasticity of population-spec ific reaction norms for growth rates.

PAGE 93

82 Table 3.4. Relationships between environmental variables, gr owth, and survival across and within species. Across species Within species ( N =4) ( N =8) S. undulatus S. woodi Trait r P r P r P Growth Survival -0.45 0.259 -0.77 0.228 -0.84 0.158 Heterogeneity 0.62 0.099 0.75 0.247 0.45 0.551 Canopy -0.23 0.576 -0.98 0.025 0.55 0.447 Activity 0.13 0.761 0.90 0.101 0.82 0.180 Food availability 0.24 0.571 0.65 0.348 -0.23 0.770 Survival Heterogeneity 0.01 0.990 -0.24 0.761 -0.84 0.160 Canopy -0.67 0.068 0.63 0.373 -0.88 0.118 Activity -0.78 0.024 -0.55 0.451 -0.50 0.498 Food availability 0.08 0.842 -0.06 0.940 0.04 0.957 r = Pearson correlation coefficient. Significant pr obabilities are denoted in bold.

PAGE 94

83 Table 3.5. Survival models including source population (pop) and growth environment (env) for resident and reciprocally transplanted hatchlings o f Sceloporus undulatus in warmer, southern and cooler, northern habitats. Model QAICc QAICc i K QDev (pop)p( ) 185.2 0.0 0.46 1.00 3 179.1 ( )p( ) 187.0 1.8 0.19 0.41 2 183.0 (pop + env)p( ) 187.3 2.0 0.17 0.36 4 179.1 (pop env)p( ) 188.2 2.9 0.11 0.23 5 177.9 (env)p( ) 189.0 3.8 0.07 0.15 3 182.9 Candidate models were evaluated using the Akaike’ s Information Criterion (AIC). QAICc = quasi-likelihood adjusted and sample-size corrected AIC. QAICc = QAICc relative to most parsimonious model. i = Akaike weight of model. = model likelihood. K = number of parameters in model. QDev = quasi-likelihood adjusted deviance. ( ) = inter cept only model. = survival probability. p = recapture probability.

PAGE 95

84 Table 3.6. Survival models including only covariates of sour ce population and growth environment for resident and reciprocally transplan ted hatchlings of Sceloporus undulatus in warmer, southern and cooler, northern habitats. Model QAICc QAICc i K QDev (mass) (growth)p( ) 181.1 0.0 0.34 1.00 4 172.9 (mass) (food) (growth)p( ) 181.5 0.4 0.27 0.80 5 171.3 (mass) (activity) (food)p( ) 182.1 1.0 0.21 0.61 5 171.8 (mass) (activity) (growth)p( ) 182.4 1.3 0.17 0.51 5 172.2 Candidate models were evaluated using the Akaike’ s Information Criterion (AIC). QAICc = quasi-likelihood adjusted and sample-size corrected AIC. QAICc = QAICc relative to most parsimonious model. i = Akaike weight of model. = model likelihood. K = number of parameters in model. QDev = quasi-likelihood adjusted deviance. ( ) = inter cept only model. = survival probability. p = recapture probability. mass = individual hatchling masses at release. activity = potential activity periods; see text for estimation. food= food availability per trap per d ay. growth = average growth rate specific to each popul ation-environment unit.

PAGE 96

85 Table 3.7. Relationships among population-environment treatm ents with regard to environmental variables. Sceloporus undulatus Sceloporus woodi Treatments Treatments Environmental variable NN NS SS SN NN NS SS SN Heterogeneity (%) 23A 68B 56B 13A 81A 37B 44B 88A Canopy cover (%) 60A 53A 58A 71B 18A 2B 1B 21A Precipitation (mm) 282(23)A 367(31)A 367(31)A 282(23)A 110(11)A 151(15)A 151(15)A 110(11)A Activity period (hours) 890(74)A 1032(86)B 935(78)B 806(67)A 604(60)A 588(59)A 529(53)A 528(53)A Food availability (mm/trap/day) 0.93A 1.10B 1.10B 0.91A 0.90A 1.01AB 1.00AB 1.25B Matching superscripts denote statistically simila r values. Heterogeneity and canopy cover are shown as percent cover. Precipitation and activity period are shown as totals with the es timated marginal means across time intervals in par entheses. Treatments are shown as abbreviated latitudes denoting population of origin first and environment second. For example, NS mea ns the northern population in the southern environment.

PAGE 97

86 Table 3.8. Results of ANOVAs for growth rates among resident populations and reciprocally transplanted hatchlings of Sceloporus woodi Residents Reciprocal transplant Source F P F P SVL (mm) 0.009 0.926 0.150 0.701 Body condition 2.410 0.140 7.825 0.009 Source population 3.940 0.065 3.562 0.068 Growth environment 2.322 0.137 Pop x env 0.229 0.636 SVL (referring to snout-vent-length) and body con dition are of hatchlings at time of release. Significant probabilities are denoted in bold.

PAGE 98

87 Table 3.9. Results of post-hoc ANOVAs on population-specific (North and South) growth rates among growth environments for Sceloporus woodi Source population North South Source F P F P SVL (mm) 0.332 0.572 0.876 0.368 Body condition 10.888 0.004 0.060 0.811 Growth environment 0.138 0.714 3.213 0.098 This analysis tests plasticity of population-spec ific reaction norms for growth rates.

PAGE 99

88 Table 3.10. Survival models including source population (pop) and growth environment (env) for resident and reciprocally transplanted hatchlings o f Sceloporus woodi in warmer, southern and cooler, northern habitats. Model QAICc QAICc i K QDev (env)p( ) 183.3 0.0 0.31 1.00 3 135.2 ( )p( ) 183.4 0.1 0.30 0.96 2 137.4 (pop + env)p( ) 184.5 1.1 0.18 0.57 4 134.2 (pop)p( ) 184.9 1.5 0.15 0.46 3 136.7 (pop env)p( ) 186.6 3.2 0.06 0.20 5 134.2 Candidate models were evaluated using the Akaike’ s Information Criterion (AIC). QAICc = quasi-likelihood adjusted and sample-size corrected AIC. QAICc = QAICc relative to most parsimonious model. i = Akaike weight of model. = model likelihood. K = number of parameters in model. QDev = quasi-likelihood adjusted deviance. ( ) = inter cept only model. = survival probability. p = recapture probability.

PAGE 100

89 Table 3.11. Survival models including only covariates of sour ce population and growth environment for resident and reciprocally transplan ted hatchlings of Sceloporus woodi in warmer, southern and cooler, northern habitats. Model QAICc QAICc i K QDev (mass) (hetero)p( ) 179.3 0.0 0.31 1.00 4 171.0 (mass)p( ) 179.3 0.0 0.31 1.00 3 173.2 (mass) (growth)p( ) 180.7 1.4 0.15 0.49 4 172.5 (mass) (hetero) (food)p( ) 181.2 1.9 0.12 0.40 5 170.8 (mass) (hetero) (growth)p( ) 181.3 2.0 0.11 0.37 5 170.9 Candidate models were evaluated using the Akaike’ s Information Criterion (AIC). QAICc = quasi-likelihood adjusted and sample-size corrected AIC. QAICc = QAICc relative to most parsimonious model. i = Akaike weight of model. = model likelihood. K = number of parameters in model. QDev = quasi-likelihood adjusted deviance. ( ) = inter cept only model. = survival probability. p = recapture probability. svl = individual hatchling snout-vent length at release. hetero = vegetative heterogeneity between 0 and 66 cm above ground for each population-environment unit. food= food availabili ty per trap per day. growth = average growth rate specific to each population-environment unit.

PAGE 101

90 Chapter 4: Habitat-specific Adaptations in Growth Rates Play a Role in Species Distribution of Sceloporus Lizards in Florida Abstract Habitat-specific adaptations are important in deter mining subsequent species’ distributions because they often result in fitness trade-offs across environments. Indeed, many studies examining the relative fitness of native an d foreign populations in particular habitats have found greater fitness associated with the native po pulations in their native environment. Native populations do not always, however, have greater fi tness than foreign transplants. Furthermore, the ecological factors responsible for species dist ributions vary considerably from abiotic factors to biotic factors and their interactions, and are r arely well understood. To understand how habitat differentiation contributes to divergent selection, subsequent reproductive isolation, and species’ distributions, reciprocal transplant experiments be tween species and habitats are necessary. In this study, we examined habitat-specific adaptation s in juvenile growth rates and survival using a reciprocal transplant experiment between species an d habitats of the Eastern Fence Lizard, Sceloporus undulatus, and the Florida Scrub Lizard, S. woodi Two populations of each species were reciprocally transplanted. Because of previou s work on these populations, we also had minimum size at maturity that could be used to dete rmine whether delayed maturity occurs for these populations in the foreign habitats. Juvenil e survival rates also allowed us to estimate the probability that individuals from each population w ill reach the minimum size at maturity in each habitat. Habitat-specific adaptations were present in juvenile growth rates, but not juvenile survival. A home advantage in juvenile growth rate s was seen, in that each native species grew faster in their native environment than did the for eign species. Juvenile survival, however, was similar for both species in all environments, which suggests that reproductive isolation does not

PAGE 102

91 occur simply through decreased juvenile survival in the foreign environments. When juvenile growth rates and survival were examined together wi th population-specific minimum size at maturity, however, there were overall habitat-speci fic adaptations with regard to the probability of reaching size at maturity. Home-site advantages we re seen, in that each species had a higher average probability of reaching size at maturity in their native habitats, relative to the foreign species. Furthermore, when each species was in th e foreign habitat they had an extremely low probability, on average, of reaching size at maturi ty. Although multiple mechanisms may be synergistically involved in the reproductive isolat ion observed between S. undulatus and S. woodi it appears that the probability of reaching size at maturity in each others’ habitats is at least part of the story. Key words. Species concepts, lizards, squamates, juvenile surv ival, local adaptation Introduction Habitat-specific adaptations are important in deter mining subsequent species’ distributions because they often result in fitness trade-offs across environments. Indeed, many studies examining the relative fitness of native an d foreign populations in particular habitats have found greater fitness associated with the native po pulations in their native environments (Hereford 2009; Kawecki and Ebert 2004; Linhart and Grant 1996). These local adaptations are strong examples of the process of natural selection that can even lead to prezygotic or postzygotic reproductive isolating barriers that de lineate species bounderies (Coyne and Orr 2004; Hendry et al. 2007; Lowry et al. 2008; McKinn on et al. 2004; Nosil 2007; Rundle 2002; Rundle and Nosil 2005; Schluter 2001; Via et al. 20 00). Native populations do not always have greater fitness than foreign transplants, however ( e.g. Galloway and Fenster 2000; Hereford and Winn 2008; Rice and Mack 1991). Furthermore, the e cological factors responsible for species distributions include abiotic variables (Cumming 20 02; Root 1988) biotic variables (Bullock et al. 2000; Terborgh and Weske 1975) and interactions amo ng variables (Randall 1982; Taniguchi and Nakano 2000), and are rarely understood well. To understand how habitat differentiation

PAGE 103

92 contributes to divergent selection, subsequent repr oductive isolation, and species’ distributions, reciprocal transplant experiments between species a nd habitats are necessary. Reciprocal transplant experiments allow us to differentiate be tween environmental and population-specific influence on phenotypic variation and fitness, and many reciprocal transplant studies have found that habitat-specific adaptations play a role in re stricting gene flow between species (Angert and Schemske 2005; Hall and Willis 2006; Linhart and Gr ant 1996; Nagy and Rice 1997; Rieseberg and Willis 2007; Wang et al. 1997). Sceloporus lizards in Florida provide a system to test the ro le of habitat-specific adaptations in limiting species’ distributions. Tw o sister species, the Eastern Fence Lizard ( S. undulatus ) and the Florida Scrub Lizard ( S. woodi ), exhibit habitat specificity in Florida. In cent ral Florida, S. undulatus is found in sandhill habitat and S. woodi in open scrub habitat on remnant Pliocene and Pleistocene sand ridges. Sandhill con sists of long-leaf pines, turkey oaks, and ground cover of wiregrass and fallen pine needles. Open scrub habitat consists of sparse sand pines, oak shrubs, and extensive bare ground (Myers and Ewel 1990b). Fossil evidence of S. undulatus and genetic variation among S. woodi populations suggest that these species have been distinct in Florida for more than one million years (Clark et al. 1999; Myers and Ewel 1990a). It is unclear how long both species have b een living parapatrically, although perhaps longer than 100,000 years (Brooks 1972). Although data suggest that S. undulatus and S. woodi have been distinct for at least one million years, where the open scrub and sandhill habitats are adjacent these two species hybridize, producing via ble hybrids and no apparent hybrid breakdown (Jackson 1972; Robbins et al. 2010). Ana lyses of gut contents indicate that the diets of the two species are similar in composition as we ll (Jackson 1973b). The most conspicuous factor separating the two spec ies of Sceloporus is their habitat specificity, which may be induced through habitat-s pecific adaptations and lower fitness in the respective foreign habitats leading to reproductive isolation. We do know that S. undulatus males are more aggressive than S. woodi males and win agonistic encounters 100% of the tim e (Robbins et al. 2010), which implies a competitive advantage to S. undulatus at least to male lizards. Because hybridization results in viable h ybrids, genetic swamping in either direction is

PAGE 104

93 also possible. However, S. undulatus males have not overrun the scrub habitat and genet ic swamping has not been observed. Sceloporus undulatus and S. woodi remain distinct species associated with distinct habitats. In this study, we examine habitat-specific adaptati ons in juvenile growth rates and survival using a reciprocal transplant experiment b etween species and habitats. Because of previous work on these populations, we know minimum sizes at maturity, which can be used to determine whether delayed maturity occurs in the fo reign habitats. We use juvenile survival rates to estimate the probability that individuals from e ach population will reach the minimum size at maturity in each habitat. Because the two species remain distinct and exhibit habitat specificity, we predicted that individuals of each population wo uld have faster growth rates and greater survival when in their native habitat. We also pre dicted that each species in the foreign habitat would exhibit delayed maturity and/or a lower proba bility of reaching maturity, when compared to the native species in its native habitat. We measu red habitat-specific variables to elucidate which ecological factor(s) may be responsible for any hab itat-specific adaptations. Methods Collection and housing of female lizards. Female Sceloporus lizards ( N =109) were collected from four populations in Florida, one nor thern and one southern population of each species. Collecting occurred from March to Septemb er in 2005. The northern populations were collected from the Ocala National Forest, Marion Co unty. Each species was collected from their respective habitats, which included a S. undulatus population ( N =37) from N 2902’18”, W 8133’35”and a S. woodi population ( N =20) from N 2906’29”, W 8148’34”(Fig. 4.1). The southern populations of S. undulatus ( N =31) and S. woodi ( N =21) were collected from Balm Boyette Preserve, Hillsborough County, N 27 45’ 60 ”, W 82 15’ 07”, and Avon Park Air Force Range, Highlands County, N 27 37’ 07”, W 81 15’20 ”, respectively. Lizards were captured using a noosing technique, given a unique toe clip for identification (Waichman 1992), contained individually in a cotton bag or plastic-ware, and t hen collectively transported in a cooler kept at 20-30 C back to the campus of the University of So uth Florida, Tampa, Florida, USA.

PAGE 105

94 Each lizard was housed individually in the laborat ory, labelled by their toe clip, species, capture date, and site of origin, and provided fres h water and crickets daily. Containers (30 x 17 x 12 cm) included a sand substrate, water dish, and plastic cover object for basking and refuge. Heat lamps maintained temperature gradients within containers that averaged 31 C during the daytime portion of a 12/12 hr day/night cycle. Egg incubation and hatchling husbandry. Each lizard and housing was checked daily for oviposition. After oviposition, each clutch ( N =109) of eggs ( N =620) was placed in a glass jar (120 ml) and buried completely in vermiculite that was premixed to a water potential of –450 kPa. Water potential for vermiculite was determined by P ackard et al (1987). All vermiculite was oven-dried at 100 C for at least 4 hours prior to mixing with distill ed water. Each jar was covered with plastic kitchen wrap, sealed with a rubber ban d and placed in an incubator set at a constant 28 C. Vermiculite was replaced for each clutch on day 25 of incubation. Eggs in the incubator were checked daily for hatchl ings ( N =509). Each hatchling was marked with a unique combination of toe-clips (Waic hman 1992) and housed in a 38-liter (10+ gallon) terrarium in the laboratory prior to their release in the field. Hatchling mortality is great est during the first few weeks after hatching for many reptiles (Crenshaw 1955; Iverson 1991; Tinkle 1967; Warner and Shine 2005), so hatchlings were ho used in the laboratory for eight weeks before being released to ensure successful mark-rec apture survival analyses. We provided water and crickets (dusted with vitamin/mineral mix) dail y for the hatchlings. For each individual, hatchling sex was recorded, and their SVL, TL, and mass (to the nearest 0.0001 g) measured before release. Male and female hatchlings did not differ in any phenotype (all P > 0.25), thus they were combined for all analyses. Reciprocal transplants. After housing the gravid females, incubating their eggs, and raising the hatchlings for eight weeks all under id entical conditions, hatchlings were released into the field under a reciprocal transplant design. At each site, the reciprocal transplant design included two enclosures constructed of a 61 cm alum inum flashing fence that was buried 13 cm

PAGE 106

95 into the ground and reinforced by metal posts (elec trical conduit) at 1.5 m intervals. The enclosure for resident hatchlings was 40 x 40 m and the enclosure for transplanted individuals was 40 x 60 m. The enclosure for transplanted indi viduals was larger because it was holding individuals from two transplanted populations of th e same species. Only one of the transplanted populations at each site was included in this study but the densities within the enclosures were within the normal range found in the field (20 to 1 24 hatchling lizards per hectare; Crenshaw 1955; McCoy et al. 2004; Niewiarowski 1994). A 1 m perimeter within each enclosure was cleared and mowed to inhibit climbing and jumping o ut of the enclosure. Hatchlings (N=192 total) were released in the enclosures (from Septem ber to December 2005) in a split-clutch design with approximately half of each clutch being released at their site of capture as residents and half at the other respective site (within latit udes in sandhill or scrub habitats). Reciprocal transplants were conducted within latitudes, betwee n species, and therefore between the sandhill and scrub habitats. The design resulted in four tr eatments per species – that of the northern populations of each species released in the s(A)ndh ill habitat (treatment acronyms refer, for example, to (N)orth (U)ndulatus s(A)ndhill; NUA; S. undulatus N=31; NWA, S. woodi N=22), the northern populations of each species released in th e s(C)rub habitat (NUC; S. undulatus N=25; NWC, S. woodi N=25), the southern populations of each species r eleased in the s(A)ndhill habitat (SUA; S. undulatus N=29; SWA, S. woodi N=14), and the southern populations of each species released in the s(C)rub habitat (SUC; S. undulatus N=31; SWC, S. woodi N=15). To increase sample sizes where necessary, hatchlings c aught in the field (N=92) were used to supplement the lab raised hatchlings. For S. undulatus 25-40% of hatchlings released at each site were field caught, and for S. woodi 67-77% of those released were field caught. Each of the four sites was methodically searched approximately every 10 days by walking around the inside perimeter of each enclosure and then zig-zagging th rough the enclosure in one direction, turning, and ziz-zagging through the enclosure in the perpen dicular direction. Searches occurred between 900-1500 hours from September 2005 to March 2006, and lasted at least 8 weeks after the last hatchling was released. When hatchlings w ere sighted, they were captured by noosing, identified by their unique toe clip combination, an d their SVL, TL, and mass were measured with

PAGE 107

96 a ruler to the nearest 0.5 mm and with a Pesola spr ing scale to the nearest 0.05 g, respectively. With these data growth rates and survivorship can b e assessed (see Data Analyses section). Environmental covariates. We measured precipitation, ground cover heterogenei ty, canopy cover, and food availability associated with both habitats. Potential activity periods were also estimated for both species in each habitat. Pr ecipitation (mm) was measured with a rain gauge that was checked and emptied during each site visit. We measured ground cover heterogeneity and canopy cover at each point of a 1 6 point grid within each enclosure. Points were 10 meters apart. We surveyed each point in al l four cardinal directions and used the average value as the sample unit. We measured grou nd cover heterogeneity using a vertical density board (Nudds 1977) from 5 meters away to el iminate spatial overlap of data collection. Heterogeneity was evaluated near the ground from 0 – 66 cm. Canopy cover was measured with a spherical densiometer from 1.3 m above ground and one meter from each point. Both vegetation measurements were estimated as percent c over. Food availability was measured throughout the mark-recapture experiment using an a rray of pitfall traps (15 traps at each of the 4 sites) that were opened approximately once a month for five trapping periods between August and January 2005-2006. Each trapping period lasted 3-10 days. The index of total biomass per trap per day was estimated by summing the lengths o f the individual arthropods caught in each trap during each time period and dividing by the nu mber of days open. Arthropods greater than 5 mm in length were considered too large for consumpt ion and not included in the analysis (Jackson 1973b). Potential activity periods were estimated between e ach lizard capture occasion. We followed the procedure in Grant and Dunham (1988) w ith slight modifications. Active lizard body temperatures were recorded for individuals of each population in their respective habitats ( S. undulatus N=116 for Ocala and N=68 for Balm; S. woodi N=91 for Ocala, N=63 for Avon) with a quick-read cloacal thermomether. If eggs were felt when a female lizard was palpated, that individual was considered gravid, and body temperat ures of gravid females were not used to calculate activity periods. Using 90% of the activ e lizard body temperatures we derived a

PAGE 108

97 minimum and maximum active lizard body temperature and used these limits to bracket the operative environmental temperatures measured by te mperature logger arrays placed at each site. These logger arrays consisted of five ibutto ns (Thermochron, model # DS1921; Maxim Integrated Products, Dallas Semiconductor, Sunnyval e, CA; www.ibutton.com ) in a cross pattern that was 1 m across. Some logger arrays were rando mly placed on the ground (using a random number table and a coordinated grid superimposed ov er a map of the site) and others on trees. Those placed on trees consisted of five loggers str ung together in a line and evenly spaced across a meter. The string of loggers was placed a round tree trunks in a spiral pattern, alternately at 0.5 and 1 m above the ground. A tre e logger array was used because both species use tree trunks to perch and bask, although Sceloporus undulatus is more arboreal than S. woodi At each site, all trees within a 60 x 80 meter a rea (including inside the enclosures) were tagged and randomly selected by number for placemen t of the tree logger array. Logger arrays were moved (randomly placed) during e ach site visit, and the data periodically downloaded. Temperatures were recorde d every 15 minutes (n=911,702). If 10% or more of the operative temperatures in at least 2 lo gger arrays during any 15 minute period were within the lizard body temperature minimum and maxi mum, potential for lizard activity was assumed. We summed these 15 minute periods to calc ulate hours per day of potential lizard activity and then summed the daily activity periods to calculate the total potential activity periods between each lizard capture occasion. If daily act ivity periods were missing because of logger malfunction we added the mean daily activity period of the particular capture interval for each missing day. Spatial autocorrelation of temperatur e loggers was also tested by comparing temperature variation within arrays to variation am ong arrays and no difference was found. We used 15 minute periods from 5 days between 1400-160 0 hours to account for the angle of the sun. We used periods with averages of 42 1 C, w hich is relatively high, allowing for temperature variation and only occurring when the s un is present. We tested if the variances were different with an ANOVA using standard deviate s from the mean as the response variable and array as the factor.

PAGE 109

98 Data analyses. Growth rates ( Sceloporus undulatus N=48; S. woodi N=33) were assessed between species and habitats with latitude included as a blocking variable because differences between north and south populations of each species are known to exist (Robbins 2010). Interactions between species x habitat and latitude x habitat were also included in the ANCOVA model with SVL at release as covariate to co ntrol for size dependent growth. The species x habitat interaction was included because a significant interaction would suggest different reactions of both species to each habitat The latitude x habitat interaction was included to assess site-specific responses in growth rates b ecause some environmental variables were different among all four sites. Individual daily g rowth rates were calculated by subtracting SVL at release from SVL at last capture and divided by the number of days in between. Only individuals with at least 13 days between measurements were use d to allow for measurable growth. Growth rates of lab raised and field caught hatchlings fro m these populations have already been shown to be similar, so they were pooled for all analyses (Robbins 2010). Survival was analyzed using Cormack-Jolly-Seber cap ture-recapture models in the information-theoretical framework (Burnham and Ande rson 2002; Lebreton et al. 1992) of the Program MARK (White and Burnham 1999). We first mo delled survival using the same factors used in the ANCOVA for growth rates (habitat, speci es, latitude, species x habitat, and latitude x habitat) to find treatment specific survival, and t hen modelled survival with only covariates associated with each site and/or treatment to find which covariates might explain treatment specific survival. We chose the full factor model as a global model and assessed how well the model fit the data, then found the best candidate m odels of survival. Time was not used as a factor because it has already been shown not to inf luence monthly survival estimates over a similar time span (Robbins 2010). Survival ( ) and recapture rates (p) were estimated using the step down approach (Lebreton et al. 1992) where p w as modelled first, then and then p again. Covariates were not used in the global model becaus e we were not interested in estimating survival after accounting for covariates, but rathe r what true survival was in each treatment. Plus, goodness-of-fit tests cannot incorporate covariates (Burnham and Anderson 2002). Model fit to the data was tested using a bootstrap method and wh ere overdispersion was found, the

PAGE 110

99 overdispersion parameter ( ) was used for correction ( =2.97; Lebreton et al. 1992). Parsimony was assessed through a maximum likelihood approach by the lowest Akaike Information Criteria (AICc) value with a bias-correction in case sample size was small with respect to the number of estimated parameters (Burnham and Anderson 2002). When was included, a Quasi-AICc value (QAICc) that accounts for was used to assess parsimony (Burnham and Anderson 2002). Starting with the global model, we found the most p arsimonious model of p among candidate models. The models of p were assessed again with t he most parsimonious model of to complete the step down method. The most parsimonio us model of p, which was the intercept only model, was then used as a constant among the c andidate models of Factor-only models were assessed first. Among the candidate model set of factor-only models, parsimony was assessed with the QAICc values and further assessed with likelihood ratio (LR) tests. The probability of survival was estimated for each trea tment through model averaging of (Burnham and Anderson 2002) among the final candidate model set. The relative influence of each factor on survival was then assessed using model averaged beta values (B) with unconditional confidence intervals (CI). Covariate-only models w ere also assessed for the influence of each environmental and morphological variable on surviva l. Habitat variables were included in the models if they were different between treatments an d they were not correlated with each other. The individual covariate of SVL was also used to as sess the influence of body size at release. Each covariate was assessed for its influence on su rvival through QAICc values among the models, and further assessed using model averaged b eta values (B) with unconditional confidence intervals (CI). Model averaging occurre d among covariate models that had QAICc values less than 2.1. If the unconditional confide nce interval (CI) did not include zero, the effect was considered statistically significant (Burnham a nd Anderson 2002). Once we had habitatspecific growth rates we could estimate, using mini mum size at maturity from this study and previously collected data, the time it would take t o reach size at maturity in each habitat. Minimum size at maturity was estimated from two yea rs (2004 and 2005) of adult female reproductive data (Robbins 2010). For each populat ion, the smallest individuals (SVL) from each year that oviposited in the laboratory were average d. Using our habitat-specific juvenile survival

PAGE 111

100 rates for each population, we could also estimate t he probability associated with individuals surviving to size at maturity. Habitat variables were compared between sites and t reatments, where possible, and those that showed differences were tested for corre lations among sites. Heterogeneity (He) and canopy cover (C) were examined among sites with Man n-Whitney U tests because they did not meet the homogeneity of variance assumption (each s ite n=32). Activity periods (A) were examined among sites within species using mixed mod el analysis with the estimated average daily activity periods associated with each treatme nt at each lizard capture interval as sample units (each site n=14). Site and time interval wer e used as factors. Precipitation was examined between sites with a mixed model analysis using ave rage daily precipitation measured for each lizard capture interval as sample units (each site n=14). Site and time interval were used as factors. Food availability (F) was examined betwee n sites ( n =14 per site per sampling period) with repeated-measures ANOVAs across the 5 sampling periods. Total biomass per trap per day was the dependent variable. Site was used as the f actor. The biomass index data was logtransformed to meet assumptions of the ANOVA. If p rimary statistical analyses found significant differences, then pairwise comparisons were made. Least significant difference tests were used for parametric analyses and Mann-Whitney U tests fo r non-parametric analyses. Correlation analyses among sites and/or treatments were used to further assess any relationships between growth, survival, and environmental variables. Results Habitat significantly affected growth rates, but sp ecies x habitat and latitude x habitat interactions were also significant (Table 4.1). In the south both species grew faster in the scrub habitat but in the north individuals of Sceloporus undulatus grew slower in the scrub habitat compared to those in the sandhill (Table 4.1, Fig. 4.2). Notably, SVL was not a significant predictor of growth rates within this system (Table 4.1). The species x habitat interaction in hatchling growth rates shows that when a species wa s in its native environment, the native species grew faster than the foreign species. This home-site advantage was stronger in the

PAGE 112

101 northern latitude. The latitude x habitat interact ion suggests that site-specific factors are likely important. However, only precipitation was signifi cantly correlated with growth rates among sites and only within S. undulatus (Table 4.2). Within S. undulatus possible relationships between growth and heterogeneity, activity period, and surv ival exist (Table 4.2). Within S. woodi a possible relationship between growth and canopy cov er exists as well (Table 4.2). Monthly juvenile survivorship associated with the s andhill and scrub habitats were similar between species at both latitudes, but there was a significant latitude x habitat interaction. In the north both species had higher survival in the sandh ill and in the south both species had higher survival in the scrub (Fig. 4.3). The intercept-on ly model was the most parsimonious recapture probability model before and after finding the most parsimonious survival probability model. The intercept only recapture probability model was ther efore used for all the survival models. The model averaged recapture probability (p) was (estim ated probability 1SE) 0.40 0.04. The candidate model set was determined with QAICc and likelihood ratio tests. The full model w as included in the candidate model set because it did not explain survival significantly worse than the two best models (both 2>0.2, P>0.65) even though the QAICc value was greater than 2.0. Furthermore, the full model did explain survival si gnificantly better than the fourth ranked model ( 2=19.7, P=0.001). Overall, the candidate set of the most parsimonious survival models included all the factors and interactions. Each factor and interaction was included in at least one model (Table 4.3). Significant factors among the best ca ndidate, factor-only, survival models were determined by model averaged beta values (B) and un conditional confidence intervals (CI). Habitat, latitude, and the latitude x habitat inter action all had a CI that did not include zero. Species and the species x habitat interaction were not significant factors (Table 4.4). Again, the latitude x habitat interaction suggests that site-s pecific factors are important. There were 8 covariate-only models with QAICc values less than 2.1 (Table 4.5). Food avail ability was included as a covariate in the models because we th ought the difference among sites ( P =0.098; see results) warranted inclusion and the informatio n-theoretical framework would filter it out (as it did; see below) if it did not influence survival. Significant habitat covariates among the best candidate, covariate-only, survival models were als o determined by model averaged beta values

PAGE 113

102 (B) and unconditional confidence intervals (CI). H eterogeneity was negatively associated with survival, and SVL was positively associated with su rvival in this system. Activity periods, canopy cover, food availability, and growth rates were not associated with survival in this system (Table 4.6). Results of correlation analyses were consist ent with the negative relationship between heterogeneity and juvenile survival in both species but were not significant (Table 4.2). Correlation analyses also suggest possible relation ships between precipitation and survival and activity periods and survival within both species, however, they were not significant at the alpha=0.05 level (Table 4.2). There was also a mar ginally significant positive correlation among all populations between growth rates and survival ( Table 4.2). Our estimates of the average time it would take ind ividuals to reach size at maturity in the north suggest that each species would reach size at maturity faster than the foreign species in their native habitat (Table 4.7; Fig. 4.4). In the south, Sceloporus woodi would reach size at maturity faster than S. undulatus in both habitats (Table 4.7; Fig. 4.4). When we i nclude monthly survival probabilities, S. undulatus had the advantage in its native habitat in the nor th, and S. woodi had the advantage in its native habitat in the sou th (Table 4.7; Fig. 4.5). Both species have an extremely low probability of reaching size at ma turity in the northern scrub habitat, with S. woodi having a higher probability, but this advantage is extremely small (Table 4.7). In the south sandhill habitat, both species have an extremely sm all probability of reaching size at maturity as well, with S. woodi again having an extremely small advantage (Table 4 .7; Fig. 4.5). Overall, however, we found habitat-specific advantages in th e average probabilities of reaching size at maturity for the native species in their native env ironments (Fig. 4.6). Environments were different between the sites with regard to some environmental variables. Specifically, heterogeneity and canopy cover were different among all four sites and activity periods were different among some sites (T able 4.8). Precipitation ( P =0.91) and food availability ( P =0.098) were similar among all four sites (Table 4. 8). Correlation analyses did not detect relationships between environmental variable s (precipitation, heterogeneity, canopy cover, and food availability) among the four sites (all P >0.14). Activity periods, which were species specific, were also not correlated with any environ mental variables (all P >0.11).

PAGE 114

103 Discussion Habitat-specific adaptations were present in juveni le growth rates, but not juvenile survival. A home-site advantage in juvenile growth rates was seen, in that each native species grew faster in its native environment than did the foreign species. This home advantage was more pronounced between the northern populations. Juvenile survival, however, was similar for both species in all environments, which suggests th at reproductive isolation does not occur simply through decreased juvenile survival in the f oreign environments. When juvenile growth rates and survival were examined together with popu lation-specific minimum size at maturity there were overall habitat-specific adaptations wit h regard to the probability of reaching size at maturity. Home-site advantages were seen, in that each species had a higher average probability of reaching size at maturity in their n ative habitats, relative to the foreign species. Furthermore, when each species was in the foreign h abitat it had an extremely low probability, on average, of reaching size at maturity. Younger age and smaller size at maturity is though t to be selected when juvenile survival is low because reaching maturity faster can increas e the probability of successful reproduction (Stearns and Koella 1986). So, early age at maturi ty may compensate for lower juvenile survival. It can be argued that juvenile survival in Sceloporus undulatus should have been lower than S. woodi in the scrub habitat because S. woodi has a younger age and smaller size at maturity tha n S. undulatus Sceloporus undulatus and S. woodi are closely related, and it is presumed that S. woodi diverged from S. undulatus during the Pliocene/Pleistocene when ocean water r ose and isolated S. undulatus populations on the sand dune ridges of central Flo rida. The life history suite of S. woodi – relatively small adult body sizes, short lifespa ns, small clutches, and early maturity – is analogous to populations of S. undulatus at its southern range limits (Tinkle and Ballinger 1972). We did not observe lower juvenile survival in S. undulatus when experiencing the scrub habitats of S. woodi suggesting that low juvenile survival is not the cause of the small adult body sizes and early age at maturity observed in S. woodi and possibly by extension, southern

PAGE 115

104 populations of S. undulatus What then, is the reason? Our results suggest t hat further study of this ecological conundrum would provide much needed insight. The home-site advantage in juvenile growth rates w as more pronounced between the northern populations, likely because of site-specif ic heterogeneity and its effect on potential activity periods. Although not statistically signi ficant, growth rates were negatively related to heterogeneity and positively related to activity pe riods, especially in S. undulatus Furthermore, the differences in heterogeneity and potential acti vity periods between habitats were significant and greater between habitats in the north. Activit y period length is known to constrain growth rates (Sinervo and Adolph 1994). The similarities in juvenile survival of both spec ies between habitats suggest that juvenile survival does not act alone as a reproductive isola ting mechanism. The fact that these species do come into contact where their habitats are adjac ent, yet neither species has expanded its range into the others’ habitat does, however, sugge st some isolating mechanism is occurring. Our estimates of the low probabilities of reaching size at maturity of both species in each others’ habitat suggests that the more complex relationship s between juvenile growth rates, survival, and size at maturity may be acting as a reproductive is olating mechanism, but other unmeasured factors may be involved. Previous work has shown t hat male agonistic encounters are won 100% of the time by S. undulatus (Robbins et al. 2010), which may push S. woodi males out of S. undulatus habitat, but it does not explain why S. undulatus males do not move into the scrub habitats of S. woodi It is possible that lower fitness is associated with each species in the foreign habitat at other life stages as well. For example, hatching success may be lower for each species in each others’ habitat. It is also possible that the isolating mechanism between these two species simply comes down to a habitat preference. Sceloporus undulatus is more arboreal than S. woodi and may prefer habitats with greater tree density, which is observed in the sandhill habitats as evidenced by observation and in our can opy cover data. We do know that microhabitat preference is different between these species during anti-predator behavioral trials. When given a refuge choice between a tree stump and a shrub, S. undulatus chose the tree and S. woodi chose the shrub when a predator approached (unpubl ished data). Whatever

PAGE 116

105 mechanisms are synergistically involved in the repr oductive isolation observed between S. undulatus and S. woodi it appears that the probability of reaching size at maturity in each others’ habitats is at least part of the story. Acknowledgements Many thanks go to the undergraduate army that helpe d us collect these data and care for the lizards in the laboratory, especially Lorelei S traub. We thank Tom Raffel for statistical help and Ray Martinez for invaluable engineering of spec ialized tools. We also thank Gary Huxel and Gordon Fox for reviewing this manuscript, for it ha s been much improved. This research was partially funded by Sigma Xi Grants-in-Aid of resea rch to TRR. Lizards were collected under collection permit WX05107 issued by the State of Fl orida Fish and Wildlife Conservation Commission. All protocols were reviewed and accept ed by the USF Institutional Animal Care and Use Committee, IACUC file #2778.

PAGE 117

106 References Adolph, S. C., and W. P. Porter. 1993. Temperature, activity, and lizard life histories. American Naturalist 142:273-295. —. 1996. Growth, seasonality, and lizard life histo ries: Age and size at maturity. Oikos 77:267278. Andrews, R. M. 1998. Geographic variation in field body temperature of Sceloporus lizards. Journal of Thermal Biology 23:329-334. Andrews, R. M., T. Mathies, and D. A. Warner. 2000. Effect of incubation temperature on morphology, growth, and survival of juvenile Sceloporus undulatus Herpetological Monographs:420-431. Angert, A. L., and D. W. Schemske. 2005. The evolut ion of species' distributions: Reciprocal transplants across the elevation ranges of Mimulus cardinalis and M. lewisii. Evolution 59:1671-1684. Angilletta, M. J. 2001. Variation in metabolic rate between populations of a geographically widespread lizard. Physiological and Biochemical Zo ology 74:11-21. Angilletta, M. J., and A. E. Dunham. 2003. The temp erature-size rule in ectotherms: Simple evolutionary explanations may not be general. Ameri can Naturalist 162:332-342. Angilletta, M. J., P. H. Niewiarowski, A. E. Dunham A. D. Leache, and W. P. Porter. 2004a. Bergmann's clines in ectotherms: Illustrating a lif e-history perspective with sceloporine lizards. American Naturalist 164:E168-E183. Angilletta, M. J., T. D. Steury, and M. W. Sears. 2 004b. Temperature, growth rate, and body size in ectotherms: Fitting pieces of a life-history puz zle. Integrative and comparative biology 44:498-509. Ashton, K. G. 2002. Patterns of within-species body size variation of birds: strong evidence for Bergmann's rule. Global Ecology and Biogeography 11 :505-523. Ashton, K. G., and C. R. Feldman. 2003. Bergmann's rule in nonavian reptiles: Turtles follow it, lizards and snakes reverse it. Evolution 57:1151-11 63.

PAGE 118

107 Ashton, K. G., M. C. Tracy, and A. de Queiroz. 2000 Is Bergmann's rule valid for mammals? American Naturalist 156:390-415. Ballinger, R. E. 1977. Reproductive strategies fo od availability as a source of proximal variation in a lizard. Ecology 58:628-635. Ballinger, R. E., and J. D. Congdon. 1980. Food res ource limitation of body growth rates in Sceloporus scalaris (Sauria, Iguanidae). Copeia 198 0:921-923. Ballinger, R. E., D. L. Droge, and S. M. Jones. 198 1. Reproduction in a Nebraska sandhills population of the northern prarie lizard Scelopours undulatus garmani American Midland Naturalist 106:157-164. Bell, G. 1977. The life of the smooth newt ( Triturus vulgaris ) after metamorphosis. Ecological Monographs 47:279-299. Bergmann, C. 1847. Uber die Verhaltnisse der Warmeo konomie der Thiere zu ihrer Grosse. Gottinger Studien 3:595-708. Berven, K. A. 1982. The genetic basis of altitudina l variation in the wood frog Rana sylvatica. 1. An experimental analysis of life history traits. Ev olution 36:962-983. Brooks, H. K. 1972. The Geology of the Ocala Nation al Forest in Snedaker, ed., The Ecology of the Ocala National Forest. Tallahassee, U.S. Forest Service, U.S.D.A. Bullock, J. M., R. J. Edwards, P. D. Carey, and R. J. Rose. 2000. Geographical separation of two Ulex species at three spatial scales: does competition limit species' ranges? Ecography 23:257-271. Burnham, K. P., and D. R. Anderson. 2002, Model sel ection and multimodal inference: a practical information-theoretic approach. New York, NY, USA, Springer-Verlag. Clark, A. M., B. W. Bowen, and L. C. Branch. 1999. Effects of natural habitat fragmentation on an endemic scrub lizard ( Sceloporus woodi ): an historical perspective based on a mitochondrial DNA gene genealogy. Molecular Ecology 8:1093-1104. Conant, R. 1975, A field guide to the reptiles and and amphibians of eastern and central North America Boston, Houghton Mifflin.

PAGE 119

108 Congdon, J. D. 1989. Proximate and evolutionary con straints on energy relations of reptiles. Physiological Zoology 62:356-373. Congdon, J. D., A. E. Dunham, and D. W. Tinkle. 198 2. Energy budgets and life histories of reptiles. Biology of Reptilia 13:233-271. Conover, D. O., and E. T. Schultz. 1995. Phenotypic similarity and the evolutionary significance of countergradient variation. Trends in Ecology & Evol ution 10:248-252. Coyne, J. A., and H. A. Orr. 2004, Speciation. Sund erland, MA, Sinauer Associates, Inc. Crenshaw, J., W, Jr. 1955. The life history of the southern spiny lizard, Sceloporus undulatus undulatus Latreille. American Midland Naturalist 54 :257-298. Cumming, G. S. 2002. Comparing climate and vegetati on as limiting factors for species ranges of African ticks. Ecology 83:255-268. Demarco, V. 1992. Embryonic development times and e gg retention in 4 species of Sceloporine lizards. Functional Ecology 6:436-444. Demarco, V. G. 1989. Annual variation in the season al shift in egg size and clutch size in Sceloporus woodi. Oecologia 80:525-532. Derickson, W. K. 1976. Ecological and physiological aspects of reproductive strategies in 2 lizards. Ecology 57:445-458. Dunham, A. E. 1978. Food availability as a proximat e factor influencing individual growth rates in Iguanid lizard Sceloporus merriami Ecology 59:770-778. Dunham, A. E., B. W. Grant, and K. L. Overall. 1989 Interfaces between biophysical and physiological ecology and the population ecology of terrestrial vertebrate ectotherms. Physiological Zoology 62:335-355. Dunham, A. E., and D. B. Miles. 1985. Patterns of c ovariation in life-history traits of squamate reptiles the effects of size and phylogeny recons idered. American Naturalist 126:231257. Ferguson, G. W., C. H. Bohlen, and H. P. Woolley. 1 980. Sceloporus undulatus comparative life-history and regulation of a kansas populations Ecology 61:313-322.

PAGE 120

109 Ferguson, G. W., and L. G. Talent. 1993. Life histo ry traits of the lizard Sceloporus undulatus from 2 populations raised in a common laboratory en vironment. Oecologia 93:88-94. Gadgil, M., and W. Bossert. 1970. Life historical c onsequences of natural selection. The American Naturalist 104:1-24. Galloway, L. F., and C. B. Fenster. 2000. Populatio n differentiation in an annual legume: Local adaptation. Evolution 54:1173-1181. Grant, B. W., and A. E. Dunham. 1988. Thermally imp osed time constraints on the activity of the desert lizard Sceloporus merriami Ecology 69:167-176. —. 1990. Elevational covariation in environmental c onstraints and life histories of the desert lizard Sceloporus merriami Ecology 71:1765-1776. Hall, M. C., and J. H. Willis. 2006. Divergent sele ction on flowering time contributes to local adaptation in Mimulus guttatus populations. Evoluti on 60:2466-2477. Hartmann, P. P. 1993. Demography of a population of the Florida scrub lizard ( Sceloporus woodi ) in a sand pine scrub on the Lake Wales Ridge of cen tral Florida, Masters Thesis, University of South Florida, Tampa. Hendry, A. P., P. Nosil, and L. H. Rieseberg. 2007. The speed of ecological speciation. Functional Ecology 21:455-464. Hereford, J. 2009. A Quantitative Survey of Local A daptation and Fitness Trade-Offs. American Naturalist 173:579-588. Hereford, J., and A. A. Winn. 2008. Limits to local adaptation in six populations of the annual plant Diodia teres. New Phytologist 178:888-896. Huey, R. B. 1991. Physiological consequences of hab itat selection. American Naturalist 137:91115. Huey, R. B., and E. R. Pianka. 1981. Ecological con sequences of foraging mode. Ecology 62:991-999. Iverson, J. B. 1991. Patterns of survivorship in tu rtles (Order Testudines). Canadian Journal of Zoology-Revue Canadienne De Zoologie 69:385-391.

PAGE 121

110 Jackson, J. F. 1972. The population phenetics and b ehavioral ecology of the Florida scrub lizard, Sceloporus woodi Dissertation thesis, University of Florida, Gains eville. —. 1973a. Distribution and population phenetics of Florida scrub lizard, Sceloporus woodi Copeia:746-761. —. 1973b. Phenetics and ecology of a narrow hybrid zone. Evolution 27:58-68. Jackson, J. F., and S. R. Telford. 1974. Reproducti ve ecology of the Florida scrub lizard, Sceloporus woodi. Copeia:689-694. James, F. C. 1970. Geographic size variation in bir ds and its relationship to climate. Ecology 51:365-&. Jones, S. M., and R. E. Ballinger. 1987. Comparativ e life histories of Holbrookia maculata and Sceloporus undulatus in western Nebraska. Ecology 68:1828-1838. Jones, S. M., R. E. Ballinger, and W. P. Porter. 19 87a. Physiological and environmental sources of variation in reproduction prarie lizards in a food rich environment. Oikos 48:325-335. Jones, S. M., S. R. Waldschmidt, and M. A. Potvin. 1987b. An experimental manipulation of food and water growth and time-space utilization of ha tchling lizards ( Sceloporus undulatus ). Oecologia 73:53-59. Kawecki, T. J., and D. Ebert. 2004. Conceptual issu es in local adaptation. Ecology Letters 7:1225-1241. Leache, A. D. 2009. Species Tree Discordance Traces to Phylogeographic Clade Boundaries in North American Fence Lizards ( Sceloporus ). Systematic Biology 58:547-559. Leache, A. D., and T. W. Reeder. 2002. Molecular sy stematics of the Eastern Fence Lizard (Sceloporus undulatus): A comparison of parsimony, likelihood, and Bayesian approaches. Systematic Biology 51:44-68. Lebreton, J. D., K. P. Burnham, J. Clobert, and D. R. Anderson. 1992. Modeling survival and testing biological hypotheses using marked animals a unified approach with case studies Ecological Monographs 62:67-118. Linhart, Y. B., and M. C. Grant. 1996. Evolutionary significance of local genetic differentiation in plants. Annual Review of Ecology and Systematics 27 :237-277.

PAGE 122

111 Lowry, D. B., R. C. Rockwood, and J. H. Willis. 200 8. Ecological reproductive isolation of coast and inland races of Mimulus guttatus. Evolution 62: 2196-2214. McCoy, E. D., P. P. Hartmann, and H. R. Mushinsky. 2004. Population biology of the rare Florida scrub lizard in fragmented habitat. Herpetologica 6 0:54-61. McCoy, E. D., and H. R. Mushinsky. 1992. Rarity of organisms in the sand pine scrub habitat of Florida. Conservation Biology 6:537-548. McKinnon, J. S., S. Mori, B. K. Blackman, L. David, D. M. Kingsley, L. Jamieson, J. Chou et al. 2004. Evidence for ecology's role in speciation. Na ture 429:294-298. Miles, D. B., R. Noecker, W. M. Roosenburg, and M. M. White. 2002. Genetic relationships among populations of Sceloporus undulatus fail to s upport present subspecific designations. Herpetologica 58:277-292. Mobley, E. R. 1998. A base line population study of the southern fence lizard, Sceloporus undulatus undulatus in central Florida., University of Central Florid a, Orlando, Florida. Myers, R. L., and J. L. Ewel. 1990a. Ecosystems of Florida. Ecosystems of Florida.:i-xviii, 1-765. —. 1990b. Scrub and high pine. Ecosystems of Florid a.:150-193. Nagy, E. S., and K. J. Rice. 1997. Local adaptation in two subspecies of an annual plant: Implications for migration and gene flow. Evolution 51:1079-1089. Niewiarowski, P. H. 1994. Understanding geographic life-history variation in lizards. Lizard ecology: Historical and experimental perspectives:3 1-49. —. 1995. Effects of supplemental feeding and therma l environment on growth rates of eastern fence lizards, Sceloporus undulatus Herpetologica 51:487-496. —. 2001. Energy budgets, growth rates, and thermal constraints: Toward an integrative approach to the study of life-history variation. American Na turalist 157:421-433. Niewiarowski, P. H., and M. J. Angilletta. 2008. Co untergradient variation in embryonic growth and development: do embryonic and juvenile performa nces trade off? Functional Ecology 22:895-901. Niewiarowski, P. H., and W. Roosenburg. 1993. Recip rocal transplant reveals sources of variation in growth rates of the lizard Sceloporus undulatus Ecology 74:1992-2002.

PAGE 123

112 Nosil, P. 2007. Divergent host plant adaptation and reproductive isolation between ecotypes of Timema cristinae walking sticks. American Naturalis t 169:151-162. Nudds, T. D. 1977. Quantifying the vegetative struc ture of wildlife cover. Wildlife Society Bulletin 5:113-117. Oufiero, C. E., and M. J. Angilletta. 2006. Converg ent evolution of embryonic growth and development in the eastern fence lizard ( Sceloporus undulatus ). Evolution 60:1066-1075. Randall, G. M. 1982. The dynamics of an insect popu lation throughout its altitudinal distribution: Coleophora alticolella (Lepidoptera) in northern En gland. Journal of Animal Ecology 51:993-1016. Rice, K. J., and R. N. Mack. 1991. Ecological genet ics of Bromus tectorum. III. The demography of reciprocally sown populations. Oecologia 88:91-1 01. Rieseberg, L. H., and J. H. Willis. 2007. Plant spe ciation. Science 317:910-914. Robbins, T. R. 2010. Geographic variation in life h istory tactics, adaptive growth rates, and habitat-specific adaptations in phylogenetically si milar species: the eastern fence lizard, Sceloporus undulatus and the Florida scrub lizard, Sceloporus woodi Dissertation, University of South Florida, Tampa. Robbins, T. R., J. N. Pruitt, L. E. Straub, E. D. M cCoy, and H. R. Mushinsky. 2010. Transgressive aggression in Sceloporus hybrids confers fitness through advantages in male agonistic encounters. Journal of Animal Ecology 79:137-147. Roff, D. A. 1997, Evolutionary Quantitative Genetic s. New York, NY, Chapman & Hall. —. 2002, Life History Evolution. Sunderland, MA, Si nauer Associates, Inc. Root, T. 1988. Energy constraints on avian distribu tions and abundances. Ecology 69:330-339. Rose, B. 1981. Factors affecting activity in Sceloporus virgatus Ecology 62:706-716. Rundle, H. D. 2002. A test of ecologically dependen t postmating isolation between sympatric sticklebacks. Evolution 56:322-329. Rundle, H. D., and P. Nosil. 2005. Ecological speci ation. Ecology Letters 8:336-352. Schaffer, W. M., and P. F. Elson. 1975. Adaptive si gnificance of variations in life-history among local populations of alantic salmon in North Americ a. Ecology 56:577-590.

PAGE 124

113 Schluter, D. 2001. Ecology and the origin of specie s. Trends in Ecology & Evolution 16:372-380. Sears, M. W., and M. J. Angilletta. 2004. Body size clines in Sceloporus lizards: Proximate mechanisms and demographic constraints. Integrative and Comparative Biology 44:433442. Sinervo, B., and S. C. Adolph. 1989. Thermal sensit ivity of growth rate in hatchling Sceloporus lizards environmental, behavioral, and genetic as pects. Oecologia 78:411-419. —. 1994. Growth plasticity and thermal opportunity in Sceloporus lizards. Ecology 75:776-790. Smith, G. R. 1998. Habitat-associated life history variation within a population of the striped plateau lizard, Sceloporus virgatus. Acta Oecologic a-International Journal of Ecology 19:167-173. Stearns, S. C. 1976. Life history tactics a revie w of the ideas. Quarterly Review of Biology 51:347. Stearns, S. C., and R. E. Crandall. 1981. Quantitat ive predictions of delayed maturity. Evolution 35:455-463. Stearns, S. C., and J. C. Koella. 1986. The evoluti on of phenotypic plasticity in life-history traits predictions of reaction norms for age and size at m aturity. Evolution 40:893-913. Storm, M. A., and M. J. Angilletta. 2007. Rapid ass imilation of yolk enhances growth and development of lizard embryos from a cold environme nt. Journal of Experimental Biology 210:3415-3421. Taniguchi, Y., and S. Nakano. 2000. Condition-speci fic competition: implications for the altitudinal distribution of stream fishes. Ecology 81:2027-2039 Terborgh, J., and J. S. Weske. 1975. The role of co mpetition in distribution of Andean birds. Ecology 56:562-576. Tinkle, D. W. 1967, The life and demography of the side-blotched lizard, Uta stansburiana : Miscellaneous Publications, v. No. 132. Ann Arbor, MI, Museum of Zoology, University of Michigan. —. 1969. Concept of reproductive effort and its rel ation to evolution of life histories of lizards. American Naturalist 103:501-&.

PAGE 125

114 —. 1972. The dynamics of a Utah population of Sceloporus undulatus Herpetologica 28:351-359. Tinkle, D. W., and R. E. Ballinger. 1972. Sceloporus undulatus study of intraspecific comparative demography of a lizard. Ecology 53:570&. Tinkle, D. W., and A. E. Dunham. 1986. Comparative life histories of two syntopic Sceloporine lizards. Copeia 1986:1-18. Tinkle, D. W., H. M. Wilbur, and S. G. Tilley. 1970 Evolutionary strategies in lizard reproduction. Evolution 24:55-&. Via, S., A. C. Bouck, and S. Skillman. 2000. Reprod uctive isolation between divergent races of pea aphids on two hosts. II. Selection against migr ants and hybrids in the parental environments. Evolution 54:1626-1637. Vinegar, M. B. 1975. Life-history phenomena in 2 po pulations of lizard Sceloporus undulatus in southwestern New Mexico. American Midland Naturalis t 93:388-402. Waichman, A. V. 1992. An alphanumeric code for toe clipping amphibians and reptiles. Herpetological Review 23:19-21. Wang, H., E. D. McArthur, S. C. Sanderson, J. H. Gr aham, and D. C. Freeman. 1997. Narrow hybrid zone between two subspecies of big sagebrush (Artemisia tridentata: Asteraceae) .4. Reciprocal transplant experiments. Evolution 51 :95-102. Warner, D. A., and R. Shine. 2005. The adaptive sig nificance of temperature-dependent sex determination: Experimental tests with a short-live d lizard. Evolution 59:2209-2221. White, G. C., and K. P. Burnham. 1999. Program MARK : survival estimation from populations of marked animals. Bird Study 46:120-138. Wiens, J. J., and T. W. Reeder. 1997. Phylogeny of the spiny lizards ( Sceloporus ) based on molecular and morphological evidence. Herpetologica l Monographs 11:1-101. Wiley, R. H. 1974. Evolution of social-organization and life-history patterns among grouse. Quarterly Review of Biology 49:201-227.

PAGE 126

115 n n r nrn Figure 4.1. Map of reciprocal transplant locations for habita t-specific adaptations study. Reciprocal transplants were conducted within latitu de (North and South) and between speciesspecific habitats.

PAGE 127

116 0.00 0.02 0.04 0.06 0.08 0.10 0.12 S(A)ndhillS(C)rubGrowth rate (mm/day)Habitat 0.00 0.02 0.04 0.06 0.08 0.10 0.12 S(A)ndhillS(C)rubGrowth rate (mm/day)Habitat 0.00 0.02 0.04 0.06 0.08 0.10 0.12 S(A)ndhillS(C)rubGrowth rate (mm/day)Habitat 0.00 0.02 0.04 0.06 0.08 0.10 0.12 S(A)ndhillS(C)rubGrowth rate (mm/day)Habitat Figure 4.2. Daily growth rates of resident and reciprocally t ransplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. Po ints are estimated marginal means. Error bars represent 1 standard er ror. The solid black circles ( ) represent S. undulatus and the open circles ( ) represent S. woodi

PAGE 128

117 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 S(A)ndhillS(C)rubMontlhy SurvivorshipHabitat 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 S(A)ndhillS(C)rubMonthly SurvivorshipHabitat 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 S(A)ndhillS(C)rubMonthly SurvivorshipHabitat 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 S(A)ndhillS(C)rubMonthly SurvivorshipHabitat Figure 4.3. Monthly juvenile survivorship of resident and rec iprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. Po ints are survival estimates from the program MARK. Error ba rs represent 1 standard error. The solid black circles ( ) represent S. undulatus and the open circles ( ) represent S. woodi

PAGE 129

118 4 6 8 10 12 14 16 18 20 22 24 S(A)ndhillS(C)rubMonths until maturityHabitat 4 6 8 10 12 14 16 18 20 22 24 S(A)ndhillS(C)rubMonths until maturityHabitat 4 6 8 10 12 14 16 18 20 22 24 S(A)ndhillS(C)rubMonths until maturityHabitat 4 6 8 10 12 14 16 18 20 22 24 S(A)ndhillS(C)rubMonths until maturityHabitat Figure 4.4. Time in months until maturity is reached for resi dent and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. Points are estimates based on population specific d ata. The solid black circles ( ) represent S. undulatus and the open circles ( ) represent S.woodi

PAGE 130

119 -0.05 0.00 0.05 0.10 0.15 0.20 S(A)ndhillS(C)rubProbability of reaching size at maturityHabitat -0.05 0.00 0.05 0.10 0.15 0.20 S(A)ndhillS(C)rubProbability of reaching size at maturityHabitat -0.05 0.00 0.05 0.10 0.15 0.20 S(A)ndhillS(C)rubProbability of reaching size at maturityHabitat -0.05 0.00 0.05 0.10 0.15 0.20 S(A)ndhillS(C)rubProbability of reaching size at maturityHabitat Figure 4.5. Probability of reaching size at maturity of resid ent and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. Points are estimates based on population specific d ata. The solid black circles ( ) represent S. undulatus and the open circles ( ) represent S. woodi

PAGE 131

120 -0.05 0.00 0.05 0.10 0.15 0.20 S(A)ndhillS(C)rubProbability of reaching size at maturityHabitat !"rrr Figure 4.6. Average probability of reaching size at maturity of resident and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between sandhill and scrub habitats. Points are estimates based on population specific data. The solid black circles ( ) represent S. undulatus and the open circles ( ) represent S. woodi

PAGE 132

121 Table 4.1. Results of ANOVA for juvenile growth rates of res ident and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. Factor F P SVL (cov) 2.19 0.143 Species 0.00 0.978 Latitude 3.33 0.072 Habitat 4.28 0.042 Species x habitat 5.53 0.021 Latitude x habitat 11.30 0.001 Significant probabilities are denoted in bold.

PAGE 133

122 Table 4.2. Relationships between environmental variables, gr owth, and survival across all treatments and within species. All treatments Within species ( N =4) ( N =8) S. undulatus S. woodi Trait r P r P r P Growth Survival 0.70 0.056 0.90 0.097 0.50 0.510 Heterogeneity -0.81 0.188 -0.08 0.923 Canopy -0.11 0.890 -0.90 0.110 Precipitation 0.99 0.004 0.47 0.528 Activity 0.21 0.617 0.88 0.116 0.74 0.265 Food availability 0.01 0.995 0.20 0.803 Survival Heterogeneity -0.95 0.055 -0.90 0.102 Canopy 0.04 0.959 -0.08 0.922 Precipitation 0.93 0.069 0.92 0.085 Activity 0.35 0.397 0.94 0.065 0.91 0.095 Food availability -0.43 0.547 -0.43 0.547 r = Pearson correlation coefficient.

PAGE 134

123 Table 4.3. Survival models including only factors of residen t and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. rrrrrrrrrrrrrrModel QAICc QAICc K QDev (H) (L) (LxH) p( ) 377.8 0.0 0.59 1.00 5 277.4 (H ) (S) (L) (LxH) p( ) 379.2 1.4 0.29 0.49 6 276.7 (H) (S) (L) (SxH) (LxH) p( ) 381.1 3.3 0.12 0.19 7 276.5 Factors are habitat (H), latitude (L), species (S ), and interactions. Candidate models were evaluated using the Akaike’s Information Criterion (AIC) and likelihood ratio tests among models (see methods). QAICc = quasi-likelihood adjusted and sample-size corrected AIC. QAICc = QAICc relative to most parsimonious model. rrAkaike weight of model.rr rmodel likelihood.rrKrrnumber of parameters in model.rrQDevrrquasi-likelihood adjusted deviance. ( ) = intercept only model. = survival probability. p = recapture probability.

PAGE 135

124 Table 4.4. Results of model averaging across the candidate m odel set of the factor only survival models. CI Factor B Lower Upper Habitat (H) 2.03 0.72 3.33 Latitude (L) 1.99 0.67 3.32 Latitude x Habitat (LxH) 3.77 -5.59 -1.95 Species (S) 0.13 -0.33 0.59 Species x Habitat (SxH) 0.05 -0.18 0.28 B refers to the beta coefficient and CI refers to the 95% unconditional confidence interval with the lower and upper limits below. Values in bold a re significant because the CI does not include zero.

PAGE 136

125 Table 4.5. Survival models including only covariates of resi dent and reciprocally transplanted hatchlings of Sceloporus undulatus and S. woodi between habitats at north and south latitudes. rrrrrrrrrrrrrrModel QAICc QAICc K QDev (He) (G) (svl) p( ) 370.7 0.0 0.20 1.00 5 360.6 (He) (svl) p( ) 370.9 0.2 0.18 0.90 4 362.8 (He) (C) (svl) p( ) 371.3 0.6 0.15 0.75 5 361.2 (He) (F) (G) (svl) p( ) 371.8 1.1 0.12 0.57 6 359.6 (F) (G) (svl) p( ) 372.1 1.4 0.10 0.50 5 361.9 (A) (He) (svl) p( ) 372.4 1.7 0.09 0.43 5 362.3 (A) (He) (G) (svl) p( ) 372.7 1.9 0.08 0.38 6 360.4 (He) (C) (G) (svl) p( ) 372.8 2.0 0.07 0.36 6 360.6 Covariates include heterogeneity (He), growth rat es (G), canopy cover (C), food availability (F), activity periods (A), and individual snout-vent len gths at time of release (svl). Candidate models were evaluated using the Akaike’s Information Crite rion (AIC). QAICc = quasi-likelihood adjusted and sample-size corrected AIC. QAICc = QAICc relative to most parsimonious model. rrAkaike weight of model.rr rmodel likelihood.rrK rnumber of parameters in model.rrQDevrrquasi-likelihood adjusted deviance. ( ) = interc ept only model. = survival probability. p = recapture probability.

PAGE 137

126 Table 4.6. Results of model averaging across the candidate m odel set of the covariate only survival models. CI Factor B Lower Upper SVL 0.19 0.02 0.35 Heterogeneity (He) -0.03 -0.05 -0.01 Activity periods (A) 8.8x10-5 -2.5x10-4 4.3x10-4 Canopy cover (C) -1.9x10-3 -8.1x10-3 4.3x10-3 Food availability (F) -1.80 -5.40 1.80 Growth rates (G) 15.03 -8.29 38.34 B refers to the beta coefficient and CI refers to the 95% unconditional confidence interval with the lower and upper limits below. Values in bold a re significant because the CI does not include zero.

PAGE 138

127 Table 4.7. Population specific life history data showing the numbers used to calculate the habitat-specific pro babilities of reaching size at maturity for Sceloporus undulatus and S. woodi Treatment Hatchling SVL at release Daily growth rate Minimum size at maturity Days until maturity Months until maturity Monthly survival rate Prob. of reaching size at maturity SUA 30.2 0.4 0.059 0.012 59.3 493.2 16.4 0.368 0.123 0.000000072 SWA 31.1 1.7 0.051 0.022 48.8 347.1 11.6 0.329 0.129 0.0000025 SUC 30.2 0.4 0.090 0.008 59.3 323.3 10.8 0.735 0.086 0.036 SWC 31.1 1.7 0.101 0.009 48.8 175.2 5.8 0.759 0.091 0.2 NUA 29.2 0.4 0.074 0.007 51.3 298.6 10 0.774 0.070 0.077 NWA 28.2 0.4 0.055 0.010 48.3 365.1 12.2 0.741 0.086 0.026 NUC 29.2 0.4 0.036 0.015 51.3 613.9 20.5 0.275 0.112 3.2E-12 NWC 28.2 0.4 0.067 0.010 48.3 300 10 0.301 0.118 0.0000061 Treatment acronyms refer, for example, to (N)orth (U)ndulatus in the s(A)ndhill habitat (see text). Error rates are displayed as plus or minus 1 standard error where appropriate.

PAGE 139

128 Table 4.8. Environmental variables among latitude-habitat tr eatments. Latitude-habitat treatments Environmental variable NA NC SA SC Heterogeneity (%) 18A 84.7B 62.7C 40.5D Canopy cover (%) 67.8A 19.9B 57.5C 1.6D Precipitation (mm) 366(1.6)A 300(1.8)A 337(1.6)A 383(2.1)A Activity period (hours) S. undulatus 1523(7.1)AC 1340(6.3)B 1319(6.3)BC 1608(7.9)A S. woodi 994(4.7)AB 923(4.3)A 857(4.1)A 1042(5.2)B Food availability (mm/trap/day) 0.709A 0.781A 0.872A 0.829A Matching superscripts denote statistically simila r values at the alpha=0.05 level. Heterogeneity and canopy cover are shown as percent cover. Preci pitation and activity period are shown as totals with estimated marginal means across interva ls in parentheses. Treatments are shown as abbreviated latitudes and habitats. For example, NA means the (N)orthern population in the s(A)ndhill habitat.

PAGE 140

About the Author Travis R. Robbins was born in La Jolla, California, and grew up in Encinitas, Vista, and Fallbrook, where he graduated from Fallbrook Union High School. He studied small business management at Mira Costa Community College and tran sferred to Arizona State University where he pursued a degree in Mechanical Engineering, only to realize that his interests really lied in the natural sciences. Thus, he earned a B.S. degree in Wildlife Conservation Biology from Arizona State University. During his stint at Arizona Stat e, he also attended Columbia University at their Biosphere 2 Center (affectionately called “The Moth er Bubble”) in Oracle, Arizona, where he received the Burgess Award for Excellence in Field Study and an Earth Scholar Certificate. He earned his Ph.D in Ecology and Evolutionary Biology at the University of South Florida. He has been a part of multiple research teams, assisting i n research at Cedar Creek Natural History Area in Bethel, Minnesota, Smithsonian Tropical Research Institute on Barro Colorado Island in Panama, and La Selva Biological Station in Costa Ri ca. He has authored multiple peer-reviewed publications in ecology and evolutionary biology an d received grants for his work from the Sigma Xi Research Society and the American Museum of Natu ral History. He now lives in State College, Pennsylvania, with his wife, Kristan.


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 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004651
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Robbins, Travis.
0 245
Geographic variation in life history tactics, adaptive growth rates, and habitat-specific adaptations in phylogenetically similar species :
b the eastern fence lizard,_sceloporus undulatus undulatus_, and the florida scrub lizard, _sceloporus woodi_
h [electronic resource] /
by Travis Robbins.
260
[Tampa, Fla] :
University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: To understand the evolutionary and ecological significance of geographic variation in life history traits, we must understand whether the patterns are induced through plastic or adaptive responses. The Eastern Fence Lizard, Sceloporus undulatus, exhibits countergradient variation (larger body sizes, et cetera, in northern, cooler environments; presumed adaptive) in life history traits across its large geographic range. However, cogradient variation (the expected result from a plastic response, although not necessarily inconsistent with adaptation) has been suggested as a null hypothesis, especially on fine geographic scales because of relatively small environmental changes. Here we focus on life history variation on a fine geographic scale to test whether cogradient variation is exhibited even though countergradient variation is exhibited at larger scales, and if so, what mechanisms are involved in the switch. We examined north and south populations (~2deg latitude between) of the S. undulatus, and the Florida Scrub Lizard, S. woodi, by measuring adult body sizes, reproduction, and hatchling body sizes over a two year period and conducting reciprocal transplants of juvenile lizards each year. Our results indicate cogradient variation (larger body size in the southern population experiencing a warmer environment) in life history traits of S. undulatus and countergradient variation, a lack of variation in adult body size, in S. woodi along the Florida peninsula. Thus, S. undulatus exhibits cogradient variation at fine geographic scales and countergradient variation at larger scales. Reciprocal transplants revealed that the larger adult body sizes in the southern population of S. undulatus could be explained by longer growth periods allowed by greater intrinsic survival. In S. woodi, the larger than expected adult body sizes in the north could be explained by faster intrinsic and extrinsic juvenile growth rates in the northern population. Because S. undulatus and S. woodi remain distinct species associated with distinct, though adjacent, habitats, we also looked for habitat-specific adaptations. The second reciprocal transplant (between species and habitats) revealed habitat-specific adaptations in juvenile growth rates, but not juvenile survival. Each native species grew faster and had a higher average probability of reaching size at maturity in their native environment than did the foreign species.
590
Advisor: Henry R. Mushinsky, Ph.D.
653
Intrinsic survival
Bergmann's rule
Intrinsic growth
Squamates
Local adaptation
690
Dissertations, Academic
z USF
x Biology-Integrative
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4651