Citation
Burmese Python Use of Gopher Tortoise Burrows in Southwestern Florida

Material Information

Title:
Burmese Python Use of Gopher Tortoise Burrows in Southwestern Florida
Creator:
Hengstebeck, Kodiak Christopher
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Wildlife Ecology and Conservation
Committee Chair:
ROMAGOSA,CHRISTINA M
Committee Co-Chair:
REED,ROBERT N
Committee Members:
SMITH,LORA L
O'DONNELL,KATHERINE MARY
Graduation Date:
12/14/2018

Subjects

Subjects / Keywords:
burmese
burrow
invasion
invasive
python
refuge
thermal
tortoise

Notes

General Note:
Invasive species threaten communities of wildlife around the world by establishing, thriving, and dominating in new ecosystems. Burmese pythons are an invasive species in southern Florida, and are substantially impacting native species within the Florida Everglades and surrounding areas. To effectively manage invasive species, such as Burmese pythons, understanding their life history and biology is crucial, as this information can be used in prevention and control efforts. We added to this information by assessing gopher tortoise burrow use by Burmese pythons, a behavior commonly seen in southwestern Florida. We conducted three complimentary studies focused on measuring the frequency and drivers of python burrow use, as well as implications gopher tortoise burrow use could have on the geographic range of pythons in Florida. The average proportion of burrows occupied by pythons in our study site was low, however we found that pythons select burrows nonrandomly, with the highest probability of python presence occurring in more compact burrows and burrows located in areas with the most canopy cover. Further, we found that internal environments within tortoise burrows typically maintain temperatures conducive for Burmese pythons to survive, even as far north as southwestern Georgia. Tortoise burrows are commonly used by many native species to avoid temperature extremes, and could potentially be used by pythons to survive in areas that would typically be too cold.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
12/31/2019

Downloads

This item has the following downloads:


Full Text

PAGE 1

BURMESE PYTHON USE OF GOPHER TORTOISE BURROWS IN SOUTHWESTERN FLORIDA By KODIAK C. HENGSTEBECK A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGR EE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2018

PAGE 2

2018 Kodiak C. Hengstebeck

PAGE 3

To my family

PAGE 4

4 ACKNOWLEDGMENTS I would first like to sincerely thank my advisor, Christina Romagosa, for her continued co unsel and support throughout this process. I thank my lab mates, parents and siblings for their encouragement. Finally, I thank the SWFL snake team for always being willing to help me find pythons.

PAGE 5

5 TABLE OF CONTENTS P age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ............................. 9 ABSTRAC T ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Background ................................ ................................ ................................ ............. 12 Burmese Python Ecology ................................ ................................ ................. 13 Gopher Tortoise Ecology ................................ ................................ .................. 15 Research Objectives ................................ ................................ ............................... 16 2 RATES OF GO PHER TORTOISE BURROW USE BY BURMESE PYTHONS IN SOUTHWESTERN FLORIDA ................................ ................................ ................. 19 Synopsis ................................ ................................ ................................ ................. 19 Background ................................ ................................ ................................ ............. 19 Methods ................................ ................................ ................................ .................. 23 Study Area and Site Selection ................................ ................................ .......... 23 Python Detection Probability ................................ ................................ ............ 24 Burrow Selection ................................ ................................ .............................. 24 Burrow Surveys ................................ ................................ ................................ 25 Habitat Variables ................................ ................................ .............................. 26 Data Analysis ................................ ................................ ................................ ... 27 Results ................................ ................................ ................................ .................... 28 Python Detection ................................ ................................ .............................. 28 Burrow Surveys ................................ ................................ ................................ 28 Habitat Variables ................................ ................................ .............................. 29 Discussion ................................ ................................ ................................ .............. 30 3 BURROW SELECTION BY BURMESE PYTHONS IN SOUTHWESTERN FLORIDA ................................ ................................ ................................ ................ 38 Synopsis ................................ ................................ ................................ ................. 38 Background ................................ ................................ ................................ ............. 38 Study Area ................................ ................................ ................................ .............. 41 Methods ................................ ................................ ................................ .................. 42

PAGE 6

6 Telemetry Tracking ................................ ................................ ........................... 42 Burrow Detection ................................ ................................ .............................. 42 Habitat Variables ................................ ................................ .............................. 43 Data Analysis ................................ ................................ ................................ ... 44 Results ................................ ................................ ................................ .................... 45 Discussion ................................ ................................ ................................ .............. 46 4 GOPHER TORTOISE BURROWS AS SUITABLE REFUGE FOR INVASIVE BURMESE PYTHONS NORTH OF THEIR CURRENT RANGE IN FLORID A ....... 54 Synopsis ................................ ................................ ................................ ................. 54 Background ................................ ................................ ................................ ............. 55 Study Area ................................ ................................ ................................ .............. 58 Python Occupied Site ................................ ................................ ....................... 58 Northern Sites ................................ ................................ ................................ .. 58 Methods ................................ ................................ ................................ .................. 59 Burrow Selection ................................ ................................ .............................. 59 Habitat Characteristics ................................ ................................ ..................... 60 Data Analysis ................................ ................................ ................................ ... 61 Results ................................ ................................ ................................ .................... 61 Discussion ................................ ................................ ................................ .............. 62 5 TRAPPING BURMESE PYTHONS FROM GOPHER TORTOISE BURROWS ...... 73 Synopsis ................................ ................................ ................................ ................. 73 Background ................................ ................................ ................................ ............. 73 Study Area ................................ ................................ ................................ .............. 75 Methods ................................ ................................ ................................ .................. 76 Trap Construction ................................ ................................ ............................. 76 Python Detection ................................ ................................ .............................. 77 T rapping ................................ ................................ ................................ ........... 78 Results ................................ ................................ ................................ .................... 78 Discussion ................................ ................................ ................................ .............. 80 6 CONCLUSION ................................ ................................ ................................ ........ 88 LIST OF REFERENCES ................................ ................................ ............................... 91 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 101

PAGE 7

7 LIST OF TABLES Table P age 2 1 Conditional detection probability of burmese pythons in b urrows ....................... 36 2 2 Details of covariates used in models of python burrow u se ................................ 36 2 3 Ranking of single season logistic regression models used to describe the prob ability of Burmese python use of gopher tortoise burro ws in southwestern Florida ................................ ................................ .......................... 37 3 1 Number of VHF telemetry locations per Burmese python used in analysis ........ 53 3 2 Details of covariates used In models of python burrow s election ....................... 53 3 3 Ranking of single season mixed effects logistic regression models (using Python ID as a random intercept) used to describe Burmese python burrow se lection in southwestern Florida ................................ ................................ ....... 53 4 1 Number of iButton temperature log gers used at each sampling site .................. 71 4 2 Results of analyses of variance (ANOVA) ................................ .......................... 71 4 3 Ranking of single season linear regression models used to describe Minimum Burro w Temperature ................................ ................................ ........... 71 4 4 Ranking of single season linear regression models used to desc ribe Maximum B urrow Temperature ................................ ................................ .......... 72 4 5 Ranking of single season linear regression models used to desc ribe Average Burrow Temperature ................................ ................................ ........................... 72 5 1 Deta ils of trapping attempts ................................ ................................ ................ 87

PAGE 8

8 LIST OF FIGURES Figure P age 1 1 The southwestern range of Burmese pythons in Florida and documentations of Burmese pythons thro ughout all of southern Florida ................................ ...... 18 2 1 A map of the study area broken up into four zones ................................ ............ 34 2 2 Rates of burrow use (total % of burrows occupied) by Burmese python s in the primary research site ................................ ................................ .................... 35 3 1 A map of the upland scrub habitat used for this study, located within Rookery Bay NERR in southwestern Florida. ................................ ................................ ... 51 3 2 RO C curve and respective area under the curve (AUC) of the best m odel of python burrow selection ................................ ................................ ...................... 52 4 1 Locations of sites used to assess physical and microclimatic characterist ics of gopher tortoise burrows ................................ ................................ .................. 65 4 2 Maximum, minimum and average burrow temperatures of each burrow at four dif ferent sites from January May 2018 ................................ ..................... 66 4 3 External temperatures and average internal temperatures of gopher tortoise b urrows (n=9) from the Joseph Jo nes Ecological Research Center ................... 67 4 4 External temperatures and average internal temperatures of gopher tortoise burrows (n=8) f rom Ashton Biological Preserve ................................ ................. 68 4 5 External temperatures and average internal temperatures of gopher tortoise burrows (n=7) f rom Archbold Biological Station ................................ .................. 69 4 6 External temperatures and average internal temperatures of gopher tortoise burro ws (n=7) from Rookery Bay NERR ................................ ............................. 70 5 1 The python trap designed to capture large snakes as the y exit gopher tortoise burrows ................................ ................................ ................................ .. 84 5 2 The python trap in situ after having cap tured a 4.33 m / 34.6 kg python ............ 85 5 3 The python trap in situ after having cap tured a 1.96 m / 2.95 kg python ............ 86

PAGE 9

9 LIST OF ABBREVIATIONS ABP Ashton Biological Preserve ABS Archbold Biological Station AIC AICc ANOVA Analysis of Varia nce AUC Area Under the Curve Avg Average GIS Geographic Information System JJERC Joseph Jones Ecological Research Center LM Linear model LTDS Line transect distance sampling Max Maximum Min Minimum RBNERR Rookery Bay National Estuarine Research Re serve ROC Receiver Operating Characteristic TL Total Length

PAGE 10

10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science BURMESE PYTHO N USE OF GOPHER TORTOISE BURROWS IN SOUTHWESTERN FLORIDA By Kodiak C. Hengstebeck December 2018 Chair: Christina M. Romagosa Major: Wildlife Ecology and Conservation Invasive species threaten communities of wildlife around the world by establishing, thr iving, and dominating in new ecosystems. Burmese pythons are an invasive species in southern Florida, and are substantially impacting native species within the Florida Everglades and surrounding areas To effectively manage invasive species, such as Burmes e pythons, understanding their life history and biology is crucial, as this information can be used in prevention and control efforts. We added to this information by assessing gopher tortoise burrow use by Burmese pythons, a behavior commonly seen in sout hwestern Florida. We conducted three complimentary studies focused on measuring the frequency and drivers of python burrow use, as well as implications gopher tortoise burrow use could have on the geographic range of pythons in Florida. The average proport ion of burrows occupied by pythons in our study site was low, however we found that pythons select burrows nonrandomly, with the highest probability of python presence occurring in more compact burrows and burrows located in areas with the most canopy cove r. Further, we found that internal environments within tortoise burrows typically maintain temperatures conducive for Burmese pythons to survive, even as far north as southwestern Georgia. Tortoise

PAGE 11

11 burrows are commonly used by many native species to avoid temperature extremes, and could potentially be used by pythons to survive in areas that would typically be too cold

PAGE 12

12 CHAPTER 1 INTRODUCTION Background In the past forty years, human population growth and frequent movement of people have greatly increased the rate of non native species introductions around the world (Pimentel et al., 2005 ). Introduced species are not always harmful, but become a problem when they are invasive, causing damage to the environment and nati ve species ( Vitousek et al., 1997 ). On ce established, invasive species commonly thrive and dominate in new ecosystems, causing dramatic changes and permanently altering native commu nities (Gurevitch & Padilla, 2004). Invasive species threaten the natural functionality of ecosystems across the globe (Clout & Williams, 2009). I n Florida, invasive Burmese pythons ( Python bivittatus ) have established throughout southern parts of the state and are negatively impacting native wildlife ( Snow et al., 2007 ; Dorcas et al., 2012; McCleery et al., 2015 ). Burmese pythons were f irst observed in 1979 in the southern Everglades and were likely established by the mid 1980 s ( Meshaka et al. 2000 ) Their range has since been expanding and pythons can now be found in southwestern Florida (Andreadis, 2011) Unlike the rest of their Florida range, Burmese pythons in s outhwest ern Florida have access to upland scrub habitats and have been using animal burrows within these habitats presumably for refuge and reproductive purposes ( pers. obs.). Many of the burrows that Burmese pythons use in southwestern Florida are those constructed by gopher tortoises ( Gopherus polyphemus ; Metzger, 2013; Bartoszek et al., 2018 ). The potential impacts that pythons may have on native upland species including gopher tortoises are currentl y unknown.

PAGE 13

13 The use of tortoise burrows by pythons could impact many native upland species in addition to gopher tortoises including several threatened or endangered species such as Eastern indigo snake s ( Drymarchon couperi ), Florida pine snake s ( Pituophi s melanoleucus mugitus ), and Florida mic e ( Podomys floridanus ) Gopher tortoises play a vital role within upland ecosystems and are considered a keystone species (Eisenberg 1983). They are ecosystem engineers, capable of digging burrows that often extend to over 4.5 m in length (Hansen 1963). Tortoise burrows serve as a refuge not only for resident tortoise s but also for many other native species of vertebrate s and invertebrate s (Landers & Speake 1980; Diemer & Speake 1981; Campbell & Christman 1982 ). The presence of Burmese pythons within these burrows could negatively affect native burrow using species in many ways including direct predation, competitiv e interactions, displacement, and the spread of disease. Burmese Python Ecology Burmese python s ar e large constrictor snake s growing up to 6 m total length (TL) and weighing as much as 90 kg ( Minton & Minton, 1973; Ernst & Zug 1996). Having such a large body size, their scope of potential prey is broad varying anywhere from rodents and meso mammals to birds and large mammals like deer ( Bhupathy & Vijayan, 1989; Ernst & Zug, 1996; Snow et al., 2007; Dove et al., 2011). Burmese pythons are native to Southeast Asia, where they are generally considered a tropical species ( Barker & Barker 2008 ) They are typically found in lowland habitats throughout their native range including mangrove forests, rainforests, wet grasslands, and coastal plains, and commonly associate with bodies of water (Barker & Barker 2008). They also associate with upland habitats wi t hin parts of their native range and occasionally use animal burrows for refuge and reproductive purposes ( Sharma & Kandel, 2016).

PAGE 14

14 Burmese pythons were first brought to Florida for the pet trade ( Meshaka et al., 2004; Snow, 2006; Kraus, 2009 ). They were su ecosystems via releases and escapes of captive individuals ( Snow 2006 ). F irst recorded in the wild in Florida in 1979 (Meshaka et al., 2000 ), pythons are n ow well established throughout s outh ern Florida ( Figure 1 1 ) w ith evidence of breeding populations ( i.e., hatchling pythons) documented since at least 2006 (Snow et al., 2007 in s outh ern Florida has greatly affected many co occurring species. They have been implicated in the sever e declines of meso mammal populations within their invaded range and are also likely affecting populations of large mammals and birds ( Dorcas et al., 2012; McCleery et al., 2015 ). They have also been documented eating threatened and endangered species, su ch as Key Largo woodrat s ( Neotoma floridana smalli ) and wood stork s ( Mycteria americana ; Snow et al., 2007; Reed & Rodda, 2009). Invasive species management is among the most challenging conservation issues confronting the state of Florida this century (M eshaka 2011). Burmese pythons are difficult to eradicate or manage once they are established as their cryptic coloration and tendency to remain under water or thick vegetation make the m difficult to detect (Piaggio et al., 2014). There is also uncertaint y in the geographic range Burmese pythons could establish in the United States. Some species distribution models predict that pythons are restricted to the range they presently inhabit in southern Florida (Pyron et al., 2008). However, Rodda et al (2011) corrected several inaccuracies reported in Pyron et al (2008), and found that Burmese pythons could potentially establish throughout most of Florida

PAGE 15

15 temperate climate zones in China and the Himalayas (G roombridge & Luxmoore 1991; Zhao & Adler 1993 ; Whitaker et al., 2004 ), suggesting that they may be more tolerant of temperate conditions than widely thought. In 2013 the Conservancy of Southwest Florida, in collaboration with a variety of state and fed eral agencies, initiated a Burmese python removal campaign for Collier County in s outhwest ern Florida. Telemetry tagged pythons were re released into Rookery Bay National Estuarine Research Reserve (RBNERR) to research spatial ec ology and capture untagged Smith et al., 2016 ). Many of the telemetry tagged pythons had home ranges that included both lowland and upland habitats, and pythons within upland habitats commonly used gopher tortoise burrows (pers. obs.). Several py thon aggregations occurring within gopher tortoise burrows have been observed, including an aggregation of seven pythons in a single burrow in 2015 (Bartoszek et al., 2018). Gopher Tortoise Ecology The gopher tortoise is a long lived species of tortoise n ative to the s outheast ern U.S. ranging from southern South Carolina to southern Florida, and west to southeastern Louisiana ( Ernst & Barbour, 1989; Ernst et al., 1994 ). Gopher tortoises typically inhabit xeric environments, preferring well drained sandy so ils of longleaf pine and oak uplands, sandhills, sand pine and oak uplands, upland hammocks, scrub habitats and ruderal communities ( e.g. roadsides, grove edges, clearings, and old fields; Auffenburg & Franz 1982; Diemer 1986). They are mostly he rbivor ous, feeding primarily on grasses and herbs, but have also been observed eating fruits, bones, charcoal, and insects (Carr, 1952; Auffenberg, 1969; Wright, 1982; Diemer, 1986; Birkhead et al ., 2005).

PAGE 16

16 Gopher tortoises play several important roles within the ir ecosystems. Their herbivory often helps maintain the openness and composition of the vegetation in their habitat, at least on localized scales (MacDonald & Mushinsky, 1988). Gopher tortoises are also well known for their burrow digging capabilities, occ asionally digging burrows that exceed 4.5 m in length ( Hallinan, 1923; Hansen 1963). These burrows can benefit the surrounding vegetation by returning leached nutrients back to the surface (Auffenberg, 1969; Landers 1980 ; Kalisz & Stone 1984,) B urrows also serve a variety of native species that coexist with gopher tortoises b y providing protection and refuge from extreme external temperatures a nd nearby predators. (Hubbard, 1893, 1894; Hallinan, 1923; Young & Goff, 1939; Brode, 1959; Hutt, 1967; Auffenb erg, 1969; Landers & Speake, 1980; Woodruff, 1982). Gopher tortoise populations have declined across their range, generally due to habitat destruction, habitat degradation, and human predation ( Garner, 1981; Auffenberg & Franz, 1982; Wright, 19 82; Lohoefe ner & Lohmeier, 1984 ). It is estimated that gopher tortoise populations have declined by 80% in the last century (Auffenberg & Franz, 1982), and they have been listed as a threatened species throughout Florida since 2007. The loss of gopher tortoises would negatively affect many native species of animals and plants that are dependent on gopher tortoises fo r the benefits that their burrows provide. An additional stressor, such as a new invasive predator like the Burmese python, could be detrimental to popula tions of tortoises and their native burrow cohabitants. Research Objectives The primary objective of this project was to assess and describe burrow use by Burme se pythons in their invaded s outhwest ern Florida range. We addressed this

PAGE 17

17 objective by conductin g three complimentary studies focused on measuring the frequency and drivers of python burrow use, as well as implications gopher tortoise burrow use could have on the geographic range of pythons in Florida. Our study also included a fourth component where we developed a method of capturing and removing pythons from tortoise burrows. First, we assessed how frequently gopher tortoise burrows were occupied by pythons within our study site in southwestern Florida (Chapter 2) Second, we conducted a comparative study with telemetry tagged pythons to assess burrow selection and third order habitat selection ( i.e., how the components of upland habitat are used by pythons; Johnson, 1980; Chapter 3). Third we compared characteristics of the gopher tortoise burrow e nvironment along a latitudinal gradient to assess the potential for pythons to use burrows as overwintering refugia north of their current range (Chapter 4) Finally, we developed and tested a burrow trap to remove and capture pythons from burrows without damaging the structural integrity of the burrow itself (Chapter 5) These studies add to our understanding of burrow use by Burmese pythons and fulfill the primary objective of assessing and describing this behavior.

PAGE 18

18 Figure 1 1. The southwestern ran ge of Burmese pythons in Florida and documentations of Burmese pythons throughout all of s outh ern Florida (obtained from EDDmaps 01/15/2018 ).

PAGE 19

19 CHAPTER 2 RATES OF GOPHER TORTOISE BURROW USE BY BURMESE PYTHONS IN SOUTHWESTERN FLORIDA Synopsis In the past fi fty years, the rate of invasive species introductions has greatly increased These species threaten communities of wildlife by establishing, thriving, and dominating in new ecosystems. They can cause dramatics changes, profoundly altering communities all o ver the world. Burmese pythons are an invasive species in southern Florida, and are substantially impacting mammal diversity within the Florida Everglades. To effectively manag e invasive species, such as Burmese pythons, understanding their life history an d biology is crucial, as this information can be used in prevention and control efforts. We added to this information by assessing gopher tortoise burrow use by Burmese pythons, a behavior commonly seen in southwestern Florida. We used repeated surveys to assess the extent to which Burmese pythons occupy tortoise burrows, and used generalized linear models to evaluate the influence that burrow, habitat, and climate characteristics have on this behavior. The average proportion of burrows occupied by pythons in our study site was low. However, we observed pythons using burrows for both refuge and reproductive purposes, suggesting that, when available, burrows may be relatively important to their life history. Additionally, the probability of python presence wi thin burrows was most affected by sampling period, rather than any burrow, habitat, or climate characteristics, providing evidence of temporal variation in burrow use by pythons. Background Invasive species are a serious threat to the natural functionalit y of ecosystems across the globe (Clout & Williams, 2009). Invasive animals threaten native wildlife

PAGE 20

20 communities and are a leading cause of species endangerment and extinction second only to habitat loss (Lowe et al., 2000). Roughly 400 of the 958 species listed as threatened or endangered under the U.S. Endangered Species Act are considered to be at risk primarily because of competition with or predation by non nati ve species (Pimentel et al., 2005) I nvasive species are among the leading causes of extinc tion in birds and the second leading cause of extinction in fish and mammals (Clavero & Garca Berthou 2005). At least sixty three non native species of amphibians and reptiles are currently esta blished within Florida making it the most invasive species rich region in the continental United States ( Krysko et al., 2016 ). Florida serves as a major port of entry into the United States for non native animals, many of which originate from tropical and semitropical regions of Central America, South America, an d Africa (Meshaka 2011). When these imported non native sub tropical climate is often conducive to their establishment (Engeman et al., 2011). Florida is currently among the most invasive species rich regions in th e world. Several species of non native constricting snakes belonging to the families Pythonidae (Cope 1864) or Boidae (Gray 1825) have estab lished populations in Florida, including Burmese python s ( Python bivittatus ), Northern African python s ( Python seba e ), and boa constrictor s ( Boa constrictor ; Snow et al., 2007; Reed et al., 2010 ). While all three python is currently the most widely established and ecologically damaging Burmese pytho ns have been implicated in severe declines of native meso mammal populations in

PAGE 21

21 s outh ern Florida and are also likely impacting native populations of large mammals and birds ( Dorcas et al., 2012; McCleery et al., 2015 ). As the range of Burmese pythons ha s expand ed throughout s outh ern Florida, pythons have gained access to different types of habitat. Southwest ern Florida is amo ng the most recently invaded regions, with pythons first being documented in the early 2000s (Andreadis, 2011 ) Pythons in this reg ion of Florida have been frequently documented using upland scrub habitats ( pers. obs.). Pythons commonly use burrows of other animals within these upland habitats, a behavior occasionally seen within their native range of Southeast Asia ( Sharma & Kandel, 2016 ), but not documented in Florida until 2010 ( Metzger, 2013 ). Many burrows that Burmese pythons have been documented using in southwestern Florida are constructed by gopher tortoises ( Gopherus polyphemus ). This behavior could negatively affect not onl y gopher tortoises, but other native animal species that rely on gopher tortoise burrows for refuge and protection. More than 360 native animal species use gopher tortoise burrows to some extent, including several state or federally listed species such as Eastern indigo snake s ( Drymarchon couperi ), Florida mice ( Podomys floridanus ), and Florida pine snakes ( Pituophis melanoleucus mugitus ), among others ( Brode, 1959 ; Hutt, 1967 ; Auffenberg, 1969; Landers & Speake, 198 0; Woodruff, 1982; Eisenberg, 1983; Jacks on & Milstrey, 1989). Gopher tortoises are also protected, listed as vulnerable under the IUCN Red List of Threatened Species since 1982, and listed as threatened in Florida since 2007. Gopher tortoises are considered to be a keystone species and ecosystem engineers, as the burrows they construct provide refuge for many native species ( Hubbard, 1893; Hallinan, 1923;

PAGE 22

22 Young & Goff, 1939; Brode, 1959; Hutt, 1967; Auffenberg 1969, Landers, 1980; Landers & Speake, 1980; Woodruff, 1982; Kalisz & Stone, 1984 ). Bu rmese pythons could negatively affect native burrow using species through predation, competition, displacement, or the spread of disease Monitoring and adaptive management plans are vital to establish and maintain healthy ecosystems (Ball et al., 2005). An important requirement for effectively managing populations of invasive species, such as Burmese python s is understanding the ir life history and biology (Engeman et al., 2011). This information can be used to better predict potential habitat use and ra nge expansion so preventative efforts can be more focused. Our objective was to assess rates of gopher tortoise burrow use by Burmese pythons withi n our primary research site in s outhwest ern Florida. We developed a hypothesis and predictions based on our knowledge of python behavior and life history. W e hypothesized that tortoise burrow characteristics and habitat variables would significantly affect the probability of python presence within burrows. First as Burmese pythons are a largely tropical specie s ( Barker & Barker, 2008 ), we expected pythons to use burrows most frequently when external temperature and humidity were lowest. Second we expected the probability of python presence to be highest in longer burrows, as long burrows are typically deeper a nd maintain more stable microclimates ( i.e., consistent temperature and humidit y). Finally, scrub habitat in Florida typically consists of patchy vegetation and intermittent open sandy areas ( Christman, 1988; Fernald, 1989; Myers, 1990 ). A s previous studie s have shown that pythons in Everglades National Park commonly use habitats with dense over/understories ( Walters

PAGE 23

23 et al., 2016 ) w e expected the probability of python presence to be highest in burrows located in patches of vegetation with the highest perce ntage s of canopy cover rather than intermittent open sandy areas. Methods Study Area and Site Selection We conducted the study within Rookery Bay National Estuarine Research Reserve, located in Collier County near Naples, Florida. Situated near the edge of the Greater Everglades Ecosystem, Rookery Bay is comprised predominantly of mangrove forests, which covers over 45,000 ha of the reserve. Patches of upland hammock and scrub habitats are also found intermittently within mangrove forests. These are some of the southernm ost patches of upland scrub habitat in Florida, and are among the first upland areas that Burmese pythons have had access to over the course of their establishment in Florida. Our study area was a scrub patch within the reserve that has bee n monitored for pythons since 2013 ; researchers have documented pythons occupying gopher tortoise burrows every year since then ( Bartoszek in prep ) The study area consisted of roughly 40 ha of upland scrub habitat suitable for gopher tortoises. I used Ge ographic Information System (GIS, ArcGIS Desktop) to identify four different zones w ithin the 40 ha of scrub habitat (Figure 2 1), which were separated by firebreaks or wetlands. I then used data provided by the Conservancy of Southwest Florida ( Chapter 1 ) to identify the zone with the most Burmese python activity in the past five years. This zone, comprised of roughly 15 ha of upland scrub habitat, was selected as the primary research site

PAGE 24

24 Python Detection Probability We used telemetry tagged pythons pro vided by the Conservancy of Southwest Florida to assess the probability of detecting pythons within tortoise burrow s. Six tagged Burmese pythons (4 male, 2 female) had home ranges that included our primary research site, and ranged in size from 174 cm TL / 2.05 kg to 422 cm TL / 43.09 kg E ach python was tracked via homing on a weekly or biweekly basis during the breeding and nesting seasons. In s outh ern Florida, python breeding typically occurs from late December to mid March, while python nesting typicall y occurs from late March to mid June ( pers. obs. ). Whenever a tagged python was tracked to a tortoise burrow, we used a burrow camera (described below) The six pytho ns were tracked to a total of sixteen different burrows f rom January 2017 to June 2017 and one python was tracked to a burrow in May 2016. We survey ed each python occupied burrow three separate times with one to three minutes between each scoping attempt to assess conditional detection probability We surveyed burrows using a camera designed by Environmental Management Systems (Canton, GA) specifically to scope animal burro ws. The camera consisted of an eight m flexible hose leading to a camera with LED lights mounted on the front. The camera unit then attached via cabl e to a battery and monitor housed within a P elican TM case. Burrow Selection We used line tr ansect distance sampling (LTDS; Burnham et al ., 1980 ) to systematically survey the primary research site for gopher tortoise burrows. W e first walked the pe rimeter of the upland habitat to delineate suitable tortoise habitat. We then used GIS (ArcGIS Desktop) to create digital transects throughout the delineated upland

PAGE 25

25 h a bitat spaced 10 m apart and running in an East West direction. We randomly selected 100 transects to survey. T ransects were surveyed once by a single observer, and all gopher tortoise burrows visible from transect s were recorded, flagged, and numbered. We detected 391 burrows during transect surveys, and selected 190 of the 391 detected burr ows for repeated sampling. We randomly selected 100 of 190 burrows and the remaining ninety burrows were all located within 50 m of burrows that were occupied by pythons at least once in the past three years. These burrows were selected so that all areas where pythons had been detected in the past three years would be included in the repeated sampling for this objective. Burrow Surveys We used repeated sampling to assess rates of burrow use by Burmese pythons during their breeding and nesting seasons. We surveyed all 190 selected burrows once per week from January 11, 2017 to April 21, 2017. For survey efficiency, t he 190 burrows were split into five groups consisting of roughly forty burrows each. Each group was sampled over the course of one day, typical ly between 0900 and 1500 hrs We randomly selected the group each day to reduce sampling bias. Each burr ow was surveyed an average of eleven times (max 12, min 7) over the course of the sampling season, resulting in a total of 1,879 burrow surveys. The fol lowing features were recorded when each burrow was surveyed: python presence or absence, tortoise presence or absence, and other commensal vertebrate species presence or absence. When a python was detected, a python trap (Chapter 5) was set to catch and remove it We used trail cameras to test the assumption that pythons are not active ( i.e., remain in burrows) during each sampling period. Thirty gopher tortoise burrows were

PAGE 26

26 randomly selected within the area of the primary research site with the mos t documented occurrences of python burrow use 2014 2017, and a single trail camera (4 Browning TM Dark Ops Extreme, 26 Bushnell TM Essential E3 ) was placed on each burrow. Cameras were programmed to use high sensitivity infrared motion sensors to take phot os whenever triggered. Each camera was placed within 1 m of a burrow, facing the entrance, and remained from January 13, 2018 to May 3, 2018, resulting in a total of 79,200 trapping hours. We reviewed i mages once manually and recorded any vertebrate animal s captured on camera as well as their time of capture. Habitat Variables We measured the following burrow characteristics and habitat variables to assess whether any affected the probability of python presence : burrow length (cm), burrow entrance width (c m), and burrow orientation (N, E, S W). Burrow length was defined as the approximate length from the entrance of the burrow to the back wall, and was measured by marking the burrow scope with white marker. Burrow entrance width was defined in accordance w ith Smith et al. (2009) as the width 50 cm inside the burrow and was measured with calipers. Finally, burrow orientation was defined as the direction that the burrow entrance faced and was measured using a compass. We convert ed the continuous degree measur ements into the four cardinal directions (316 o 45 o = N, 46 o 135 o = E, 136 o 225 o =S, 226 o 315 o =W). The following habitat variables were also measured for each burrow: canopy cover (%), distance to water (m), air temperature ( o C ), and relative humidi ty (%). Canopy cover was defined as the percentage of vegetative cover waist height and was measured at the apron of each burrow using a densiometer. Distance to water was defined as the distance from each burrow to the nearest permanent water source and was measured using GIS (ArcGIS Desktop).

PAGE 27

27 Distance to water was classified as proximate (0 100 m), near (100 200 m), or far (20 0 + m ). Air temperature and relative humidity were recorded at a semi shaded central location within the scrub habitat of the primary research site and were measured using an iButton Hygrochron temperature and humidity data logger. Data Analysis We performed repeated presence absence surveys to estimate the proportion of burrows occupied by pythons within our primary research s ite during each sampling period. We investigated which factors affect the probability of python presence within tortoise burrows using logistic regression in program R version 3.1.2. We evaluated a suite of 16 a priori models that represented our hypothesi zed predictors of python presence within tortoise burrows (Table 2 2). We also assessed conditional detection probability to minimize potential sampling bias (Zeller et al., 2011). Sampling bias, typically caused by the failure to detect the species of int erest when it is indeed present, can result in underestimates and ultimately lead s to misguided conservation decisions (Linkie et al., 2007). Due to the high detection probability of pythons within burrows (0.922 ) and low probability of type II error, we assumed perfect detection probability for all models Models were ranked using Ak aike Information Criterion (AIC; Akaike, 2011 ) and AIC weight ( The weight of any particular model depends on the set of candidate models and varies from 0 (no support) to 1 (complete support). I considered any model within 2.0 AIC units of the best model to be competing models, and evaluated the model goodness of fit by calculating the area under the receiver curve statistic (AUC ; Hanley & McNeil, 1982 ).

PAGE 28

28 Results Python Detection We used six telemetry tagged pythons to evaluate conditional detection probability ( i.e., the probability of detection, given we know th e python is present) o f Burmese pythons within burrows. Pythons were tracked to a total of seventeen burrows (max 5, min 1 avg 2.83 ) and each burrow was searched with a burrow scope two or three times several minutes apart, resulting in a total of forty s ix detection attempts. In fifteen of seventeen burrows, p ythons were detected in all attempts. Python detection was only reduced when burrow length exceeded that of the burrow camera, or when the burrow contained water in which the python could fully subme rge. In total, p ythons were detected in forty three of forty six attempts (Table 2 1 ) for a n average detection rate of 0.922 Burrow Surveys During burrow surveys, we detected Burmese pythons on seven occasions, all on different dates and in different bur rows. The proportion of burrows occupie d by pythons varied from 0 to 3.1 % per day, with an average of 0. 36 % of burrows occupied throughout the sampling season We detected f ive pythons in tortoise burrows in January 2017, one python in March 2017, and one python in April 2017 (Figure 2 2). We found reproductively active male pythons, reproductively active and egg bound female pythons, juvenile pythons, and an old python nest within burrows. Detected pytho ns (4 male, 3 female) ranged in size from 1.74 m TL / 2.05 kg to 4.33 m TL / 34.6 kg. Burmese pythons were the most common snake and the fifth most common vertebrate within tortoise burrows during scoping Gopher tortoises we re seen most

PAGE 29

29 frequently, with on average 24% of burrows being occupied throughout the sampling season. Other common animals were b rown anoles ( Anolis sagrei avg 1 1 % ) cricket frogs ( Acris gryllus avg 1 1 % ) and cotton mice ( Peromyscus gossypinus avg 0. 69 % ) Other notable animals included Eastern indigo snakes ( avg 0. 17 % ) and cotton rats ( Sigmodon hispidus avg 0. 31 % ) During camera trapping Burmese pythons were detected by cameras on three separate occasions at different burrows. Two detectio ns occurred on January 21, 2018 several minutes apart ( 15:56 16:00 ) at burrows within 10 m of each other, suggesting that the same python was detected twice. The third python was detected on May 2, 2018 at 19:25 hr All detections occurred at times after burrow surveys (0900 1500 hrs), and support the assumption that pythons did not move amon g burrows while we were actively surveying. Habitat Variables The best approximating model for python presence indicated that sampling period ( 0.0631 + 0.035 ; p = 0.07 ) rather than any burrow characteristics or habitat variables, was the best predic tor of python presence or absence within burrows ( Table 2 2). However, the AUC indicated the model was only moderately accurate at predicting python presence (AUC = 0.71 ) Additionally, five models were within 2.0 AIC units; the top three model weights wer e similar and overall had 53% of the model weights (Table 2 2) These models indicated that sampling period, minimum external temperature, and burrow length were the best predictors of python presence within burrows, although sampling period was the only m arginally significant variable ( p = 0.07).

PAGE 30

30 Discussion Our study provides further information on the behavior and life history of Burmese pythons in s outhwest ern Florida. Pythons used gopher tortoise bur rows within upland habitats in s outhwest ern Florida however the average proportion of burrow s occupied during the sampling season was extremely low (0. 359 % ) This low rate of burrow use accompanied by the extremely high detection probability of pythons within burrows (0.922 ) suggest s that pythons may no t be using tortoise burrows at this site as frequently as previously thought or that the population density of pythons in and around the site is low Low population density in the area could be attributed to invasion rate ( i.e., the research site is near t he presumed northern margin of the current range of Burmese pythons in Florida ) or python removal s P ython removal efforts in the area (see Chapter 1) have been high, resulting in the removal of m ore than twenty mature pythons and four python nests from th e study site since 2014. Our results also provide evidence that rates of burrow use by pythons vary temporally. Five of the seven pythons found within burrows were detected during mid to late January, suggesting that may be when burrow use by pythons occ urs most frequently in this region of Florida. Four of the five top models contained sampling period as a covariate, further reflecting temporal variability of python burrow use. However, likely due to the low number of pythons found within burrows during the sampling season, all models were fairly uninformative and all model covariates were non significant. Therefore we cannot precisely estimate how strongly burrow characteristics and habitat variables affect the probability of python presence within burro ws. Regardless, these results match fairly closely with those recorded from

PAGE 31

31 previous years which show that telemetry tagged pythons are found within burrows most frequently during the months of January and February (Bartoszek in prep). While the detection probability of pythons in most habitats is th ought to be extremely low (<1%) (Dorcas & Willson 2013), and the overall proportion of burrows used by pythons in this study was low, we found the detection probability of py thons within tortoise burrows to be extremely high (0.922 ). It is worth noting, however, that detection in this case represents the conditional detection probability. Regardless, we found that python detection within burrows was only reduced when burrow length exceeded that of the burrow cam era, or when the burrow contained water in which the python could fully submerge We found reproductively active male pythons, reproductively active and egg bound female pythons, juvenile pythons, and an old python nest within burrows, confirming that Burm ese pythons are using tortoise burrows for both repr oductive purposes and as refugia While Burmese pythons are typically associated with lowland habitats (Barker & Barker 2008), some features of upland habitats are very conducive for their establishment. We detected relatively high rates of rodents, including cotton mice and cotton rats, within tortoise burrows, which a re among the most common prey items in Burmese python gut contents ( Romagosa in prep ). Additionally, sources of fresh water, such as ponds and wetlands, are common within most upland habitats in the region suggesting that Burmese pythons could potentially thrive in these upland sites. High prey densities refuge availability, and water could facilitate higher survivorship for pythons of all age classes.

PAGE 32

32 While the overall rate of python burrow use was low, Burmese pythons were found in burrows more frequently than any other species of snake; these findings, however could be confounded by lower detection probabilities of smaller snake species which were not tested as part of this study. Pythons share similar diets with some native upland snake species, including Eastern indigo snakes, rat snakes ( Pantherophis sp. ) and Eastern diamondback rattlesnakes ( Crotalus adamanteus ) all of which were d etected within burrows in this survey Pythons could negatively impact populations of these native snakes through competitive interactions or the spread of harmful non native parasites, which Burmese pythons have been shown to carry (Miller et al., 2018 ). Indigo snakes, however, could potentially impede the establishment of pythons within upland habitats, as indigo snakes have been observed consuming juvenile Burmese pythons (Andreadis et al., 2018). Pythons were found co occupying burrows with gopher tort oises on two occasions throughout the study While predation by pythons on gopher tortoises has not been observed other large constrictors including green anacondas have been observed consuming various species of Testudines (Rivas 1999). H atchling and juv enile gopher tortoises are also depredated by many upland species including several species of snakes (Douglass & Winegarner 1977, Butler & Sowell 1996). Additionally, an aggregation of seven mature Burmese pythons were found co occupying a burrow with a g opher tortoise in 2015 indicating that multiple occupancy can occur (Bartoszek et al., 2018). In this event, the pythons blocked the tunnel of the burrow, hindering the ivities such as feeding and thermoregulating Pythons could also displace gopher tortoises

PAGE 33

33 from their burrows, resulting in prolonged exposure to harsh external temperatures and leaving tortoise s vulnerable to predators Finally, our study has provided e vidence that burrow surveys conducted with the purpose of detecting Burmese pythons can be incorporated into python management efforts. T he high detection probability of pythons within tortoise burrows can be used to potentially locate pythons within uplan d habitats. The results of our study provide evidence that burrow use by pythons peaks in mid to late January, suggesting that searches for pythons in tortoise burrows within upland habitats should be focused during this time if possible Maximizing the e fficiency of python management will help prevent ecosystem damage and protect countless native upland species.

PAGE 34

34 Figure 2 1. A map of the study area broken up into four zones Zone 1 consists of roughly 10.0 ha of upland scrub habitat, zone 2 consists of roughly 15.0 ha of upland scrub habita t, zone 3 consists of roughly 6.50 ha of upland scrub habitat, and zone 4 co nsists of roughly 9.00 ha of upland scrub habitat. Zone 2 was selected as the primary research site for this study.

PAGE 35

35 Figure 2 2. Rates of burrow use (total % of burrows occupied) by Burmese pythons in the primary research site. Rates are broken up by sampling period, each consisting of one day between the hours of 0900 and 1500.

PAGE 36

36 Table 2 1. Conditional d etection p ro babilit y of Burmese p ythons in b urrows Burrow ID a Date b Python Size (m) c Burrow Size (cm) d Survey 1 e Survey 2 e Survey 3 e B1 05 / 06 / 2016 3.60 30 N N B2 01 / 23 / 2017 3.21 22 Y Y Y B3 01 / 24 / 2017 3.21 25 Y Y Y B4 02 / 07 / 2017 3.21 27 Y Y Y B 5 02 / 07 / 2017 4.22 28 Y Y Y B6 02 / 14 / 2017 3.21 24 Y Y B7 02 / 21 / 2017 1.74 5.0 Y Y Y B8 03 / 07 / 2017 3.60 24 Y Y Y B9 03 / 13 / 2017 3.60 30 Y Y B10 03 / 14 / 2017 1.74 5.0 Y Y B11 03 / 18 / 2017 3.60 24 Y Y Y B12 04 / 01 / 2017 1.7 4 5.0 Y Y Y B13 05 / 03 / 2017 3.60 24 Y Y Y B14 05 / 08 / 2017 1.82 23 Y Y B15 06 / 08 / 2017 4.22 31 Y Y N B16 06 / 08 / 2017 4.17 27 Y Y Y B17 06 / 13 / 2017 4.22 31 Y Y Y a Burrow ID is the identification of each burrow surveyed. b Date indicat es the date that each respective survey of python detection probability took place. c Python size is the TL of each python, d Burrow Size is the width of the burrow measured 50 cm within the burrow, and e Survey indicates whether or not the python was succe ssfully detected during each attempt (Y = detected, N = not detected). Table 2 2. Details of c ovariates u sed i n m odels of p ython b urrow u se Variable Description Units Max Min Avg Period Sampling period (1 day each) 1 12 Temp Minimum external t emperature o C 22.7 6.72 15.3 Humidity Minimum external humidity % 74.0 21.0 47.8 Length Approximate length of burrow Cm 655 98.5 305 Width Width 50 cm inside burrow Cm 49.0 12.5 25.4 Orientation Orientation of burrow entrance N,E,S,W Tortoise To rtoise present in burrow or not Y or N Cover Canopy cover % 46.9 0.00 5.20 Water Distance to water m 394 137 275

PAGE 37

37 Table 2 3 Ranking of single season logistic regression models used to describe the probability of Burmese python use of gop her tortoise burrows in s outhwest ern Florida. Model Name a K b AIC b AIC b b (Period) 1 92.78444 0 0.1897 (Temp + Length + Period) 3 92.85794 0.0735 0.1829 (Temp + Period) 2 93.09687 0.31243 0.1623 (Humidity + Period) 2 94.59762 1.81318 0.0766 (Te mp + Length) 2 94.61259 1.82815 0.0761 (Temp) 1 94.88629 2.10185 0.0663 (Temp Length) 2 96.15356 3.36912 0.0352 (Humidity + Length) 2 96.19933 3.41489 0.0344 (Tortoise) 1 96.27768 3.49324 0.0331 (Temp + Length + Width) 3 96.41584 3.6314 0.0309 (Tem p + Orientation) 2 96.49803 3.71359 0.0296 (Temp + Length + Cover) 3 96.6032 3.81876 0.0281 (Temp + Tortoise) 2 96.6259 3.84146 0.0278 (Humidity + Length + Width) 3 97.99379 5.20935 0.014 (Temp + Orientation + Cover) 3 98.42297 5.63853 0.0113 (Global) 9 102.1512 9.36675 0.0018 a Model covariates include sampling period (Period), minimum external temperature (Temp), minimum external humidity (Humidity), burrow length (Length), burrow entrance width (Width), burrow entrance orientation (Orientation), go pher tortoise presence within the burrow (Tortoise), canopy cover (Cover) and distance to water (water) b K = number of variables in model, AIC = Akaike AIC value of each model and the lowest AIC model, = AIC weight.

PAGE 38

38 CHAPTER 3 BURROW SELECTION BY BURMESE PYTHONS IN SOUTHWESTERN FLORIDA Synopsis Individuals of the same species wit hin the same local region often occur in several distinct types of habitat occasionally resulting in variation of certain life history traits depending on the type of habitat in which each individual occurs To better understand these variations, it is im portant to understand how individuals of a population select the habitat they occupy, commonly analyzed by defining which habitats a species uses in relation to the habitats that are available Knowing this information is central to understanding the ecolo gy of that animal and can greatly benefit management efforts. Burmese pythons are an invasive species in southern Florida, substantially impacting mammal diversity within the Florida Everglades and surrounding areas In southwestern Florida, pythons occup y upland habitats and commonly use animal burrows within these habitats for vari ous purposes. We used telemetry tagged pythons to assess third order resource selection by Burmese pythons occurring in upland habitats in southwestern Florida, and used genera lized linear mixed models to evaluate which burrow and habitat characteristics pythons select in these areas We found that pythons select burrows nonrandomly, with the highest probability of python presence occurring in more compact burrows and burrows located in areas with the most canopy cover. This information may help better predict potential c olonization so that prevention and control efforts can be focused more specifically. Background use of an environment, including its diet and the h abitat it occupies, is central to understanding the ecology of that animal (Johnson 1980).

PAGE 39

39 Individuals of the same species within the same local region often occur in several distinct types of habitat (Pulliam & Danielson 1991). In these cases, certain l ife history traits such as development rates, life spans, birth rates, and death rates may differ depending on the type of habitat in which each individual occurs (Pulliam & Danielson 1991). This information can benefit management efforts for animals of c oncern including protected species or potentially destructive invasive species. I nvasive Burmese pythons ( Python bivittatus ) have become established in southern Florida and are significantly impacting native wildlife. Pythons have been implicated in the severe declines of meso mammal populations in South Florida (Dorcas et al., 2012; McCleery et al., 2015 ) and have also been documented feeding on American alligators ( Alligator mississippiensis ) large mammals, and a wide variety of birds (Reed & Rodda 20 09; Dove et al., 2011 ) Burmese pythons are native to Southeast Asia, ranging from Thailand to southern China, and west to India ( Groombridge & Luxmoore 1991; Whitaker et al., 2004; Zhao & Adler 1993 ) They predominantly inhabit lowland s including mangro ve forests, rainforests, wet grasslands, and coastal plains, and are commonly associate d with bodies of water (Barker & Barker 2008). Although typically considered to be a tropical species, the native range of Burmese pythons extends well into the temper ate climate regions of China and the Himalayas (Groombridge & Luxmoore, 1991; Whitaker et al., 2004; Zhao & Adler 1993), suggesting that they may be more tolerant of temperate conditions than widely thought. I n s outhwest ern Florida, invasive pythons have been documented using both lowland habitats, including mangrove forests and flooded grasslands, and dry upland

PAGE 40

40 habitats, including upland scrub and pine hammocks (pers. obs.) Pythons that use upland habitats in this region of Florida have been commonly ob served using gopher tortoise ( Gopherus polyphemus ) burrows (Metzger, 2013; Bartoszek et al., 2018) Preliminary data suggest that r eproductively mature pythons primarily u se tortoise breeding and nesting seasons for aggregating, mating, and laying eggs while juvenile pythons have been documented using burrows during every month of the year (Bartoszek unpublished data ). As this behavior by Burmese pythons was first documented in Florida in 2010 ( Metzger, 2013 ), little is known abo ut the extent of their burrow use (see Chapter 2) or the impacts it may have on native upland species. An essential requirement for effectively managing populations of invasive species, such as the Burmese python, is understanding the life history and bio logy of that particular species (Engeman et al., 2011). This information may help better predict potential c olonization so that prevention and control efforts can be focused more specifically. Habitat selection is commonly analyzed by d efining which habita ts a species uses in relation to the habitats that are available (Bra dshaw et al., 1995). Availability is defined as the accessibility of a habitat or component of a habitat to an individual, although conclusions about availability often depend on the inve stigator s notion of what components of a habitat are actually available to the animal (Johnson 1980). Habitat use is considered selective, as opposed to random, if habitat types are used disproportionately to their availability (Johnson 1980). Habitat s election is hierarchical, broken down into the following orders: first order selection describes the physical or geographical range of a species, second order selection describes the home

PAGE 41

41 range of an individual or social group, third order selection descri bes the use of various habitat components within the home range, and fourth order selection describes the use of micro habitats and the procurement of food items (Johnson, 1980) Our objective was to assess burrow selection a measure of third order reso urce selection (Johnson, 1980) by Burmese pythons at our research site in s outhwest ern Florida. We used a comparative design with telemetry tagged pythons to investigate potential differences between occupied and available burrows W e hypothesized that py thons select burrows nonrandomly based on the burrow characteristics and surrounding habitat First we expected the probability of python presence to be highest in longer burrows, as long burrows are typically deeper and maintain more stable microclimates ( i.e., consistent temperature and humidity ) Second, as Burmese pythons commonly associate with water and lowland areas (Barker & Barker, 2008), we expected the probability of python presence to be highest in burrows near water. Finally, scrub habitat typ ically consists of patchy vegetation and intermittent open sandy areas ( Christman, 1988; Fernald, 1989; Myers, 1990 ). As previous studies have shown that pythons in Everglades National Park commonly use habitats with dense over/understories ( Walters et al. 2016 ), we expected the probability of python presence to be highest in burrows located in patches of vegetation with highest percentage s of canopy cover rather than intermittent open sandy areas. Study Area I conducted the study within ~40 ha of upland scrub habitat in Rookery Bay National Estuarine Research R eserve (Figure 3 1) located near Naples in s outhwest ern Florida (see Chapter 2)

PAGE 42

42 Methods Telemetry Tracking We used telemetry tagged pythons provided to us by the Conservancy of Southwest Florida to conduct f ieldwork for this study (see Chapter 1) Our study area was located within the home range s of seven telemetry tagged Burmese pythons (5 male, 2 female). These pythons ranged in size, with the smalle st python measuring 174 cm TL / 2.05 kg, and the larges t p ython measuring 422 cm TL / 43.1 kg. Five of the tagged pythons (3 male, 2 female) were reproductively mature, while two (2 male, 0 female) were juvenile. We tracked tagged pythons on a weekly or biweekly basis to monito r their locations. Whenever a python was tracked to a tortois e burrow, we scoped the burrow with a burrow camer a to confirm python occupancy. From January 23, 2017 to June 5, 2017, we tracked telemetry tagged pythons to a total of seventeen different burrows. We also tracked telemetry tagged pythons to three different burrows from February 16, 2018 to March 5, 2018, result ing in a total of twenty python occupied burrows used in the analysis of burrow selection. Each python was tracked to an average of 2.86 burrows (max 5, min 1). We measured burrow and habita t characteristics of all python occupied burrows. Burrow Detection We compared burrows that were occupied by tagged pyth ons to all other detected burrows that were within a 15 0 m radius. We selected 15 0 m based on the distances traveled b y three different telemetry tagged pythons during four different overnight dispersal events. During dispersal events, each python moved from one burrow to another over the course of one night. Dispersal distances were 312 m, 88 m, 69 m, and 63 m, for an av erage nightly dispersal distance of 133 m. We used 150 m, as

PAGE 43

43 opposed to the entire home range of each telemetry tagged animal as described in Gillies et al. (2006 ), due to the relatively limited area of suitable gopher tortoise habitat and expansive home r ange sizes of Burmese pythons ( Hart et al ., 2015 ). We used line transect distance sampling ( LTDS; Burnham et al ., 1980 ) to systematically survey the 15 0 m area around the python occupied burrows with the purpose of detecting all other burrows. We used a Ge ographic Information System (GIS, ArcGIS Desktop) to create digital transects within the 15 0 m area. Transects were spaced 10 m apart from each other and ran in an East West direction. All transects were surveyed once by a single observer, and all burrows visible from transects were recorded, flagged, and numbered. We then randomly selected 50% of the burrows that were detected during transect surveys for analysis. We measured characteristics of the randomly selected burrows along with habitat variables of the immediate area. Habitat Variables We measured burrow characteristics and habitat variables to assess third order resource selection (Johnson, 1980) by Burmese pythons. The following burrow characteristics were measured for each burrow: burrow length ( cm), burrow entrance width (cm), and burrow orientation (N, S, E, W). Burrow length was recorded as the approximate length from the entrance of the burrow to the back wall, and was measured by marking the burrow scope with a white marker. Burrow entrance w idth was recorded in accordance with Smith et al. (2009) as the width 50 cm within the burrow and was measured using calipers Finally, burrow orientation was recorded as the direction that the burrow entrance faced and was measured using a compass. The co ntinuous degree measurements were then converted into the four cardinal directions (316 o 45 o = N, 46 o 135 o = E, 136 o 225 o =S, 226 o 315 o =W) We also measured t he following habitat

PAGE 44

44 variables for each burrow: canopy c over (%) and distance to water (m) Canopy cover was recorded as the percentage of vegetati ve cover waist height and was measured at the apron of each burrow using a densiometer Distance to water was recorded as the distance from each burr ow to the nearest permanent source of water and was measured using GIS ( ArcGIS Desktop) Distance to water was classified as proximate (0 100 m), near (100 200 m), or far (200 + m ). Data Analysis We used telemetry tagged pythons to compare occupied and available burrows, which represents third order resource selection (Johnson 1980). In total, we used twenty burrow locations from two juvenile pythons (30 cm 25 0 cm TL) and five adult pythons ( 25 0 + cm TL ). Samples were unbalanced, varying from one to five locations per individual python (Table 3 1). All detected burrows located within a 150 m radius of a python were considered available to that python at that time We investigated which factors affect the probability of python presence within burrows using mixed effects logistic regression in program R version 3.1.2. To capture variation between individual evaluated a suite of 13 a priori models that represented our hypothesized predictors of python presence within bu rrows (Table 3 2). We ranked m odels using Ak aike Information Criterion (AIC; Akaike, 2011 ) and AIC weight ( models and varies from 0 (no support) to 1 (complete support). I considere d any model within 2.0 AIC units of the best model to be competing models, and evaluated the model goodness of fit by calculating the area under the receiver curve statistic (AUC ; Hanley & McNeil, 1982 ).

PAGE 45

45 Results We tracked seven telemetry tagged Burmese p ythons to a total of twenty different burrows. T he size of burrows occupied by pythons ranged from 5.0 cm entrance width to 31 cm entrance width, with an average of 22 cm entrance width per burrow Burrow entrance width was highly correlated with python ag e class (R 2 =0.96), with juvenile pythons occupying much smaller burrows. We observed pythons occupying gopher tortoise burrows, nine banded armadillo ( Dasypus novemcinctus ) burrows, and very small burrows presumably created by rodents. Nine of the burrows were in active use by gopher tortoises, and seven of the nine were occupied by both the resident gopher tortoise and a Burmese python at the same time. Two telemetry tagged python s were observed going through ecdysis within burrow s. Eighteen of the burrows occupied by pythons contained a single python. One burrow contained two pythons: a telemetry tagged mature male python and an untagged slightly smaller python. One burrow contained three pythons: a telemetry tagged mature male python, a n untagged slightly smaller mature male python, and a n untagged large mature female python. Two of the single occupied burrows were used by mature female pythons for nesting. The best approximating model of burrow selection and third order habitat selection by pythons incl uded two variables, burrow entrance width and canopy cover (Table 3 2) This model indicated that these variables were the best predictors of python burrow use as it was weighted much higher than any other candidate models ( = 0.9857, Table 3 2 ) The AUC indicated the model accurately predicted python presence and absence (AUC = 0.88 ; Figure 3 2 ) Python presence within burrows was negatively 0.9801 + 0.28 p<0.001 ) and positively + 0.24 p<0.001 ).

PAGE 46

46 Discussion Our study provides f urther information about burrow use and upland hab itat use by Burmese pythons in s outhwest ern Florida. Burmese pythons commonly use gopher tortoise burrows within up land habitats in this region of Florida, a behavior only recently documented in Florida ( Metzger, 2013 ), although this may be due to minimal access to upland habitats throughout most of the southern regions of the Greater Everglades Ecosystem where pythons have been established for the longest periods of time ( Snow et al., 2007). Our results provide considerable support for our hypothesis that pythons are selecting burrows nonrandomly based on characteristics of the burrows and the surrounding habitat. Res ults from our study suggest that pythons occurring within upland habitats may sel ect for burrows that are smaller and burrows located within areas containing the highest percentage of canopy cover Preference for compact burrows by pythons in s outhwest ern Florida corresponds with the findings by Mukherjee et al. (2017), who found that Indian pythons in Keoladeo National Park India most commonly use small, compact chambers within burrows. Reasons for this are unknown, but could be related to more efficient thermoregulation, due to the smaller volume within the compact chambers, or the avoidance of burrows that allow space for potential predators. Preference for dense canopy cover corresponds with the findings of Walters et al. (2016) who found that Burmese p ythons within Everglades National Park select for broad leafed habitats, which are characterized by dense over/understory. Dense canopy cover likely provides suitable cover for pythons to better avoid potential predators and ambush possible prey.

PAGE 47

47 Our data show that all age classes of pythons, from hatchlings to mature adults, use burrows for various purposes. We found that all age classes appear to use burrows for refuge presumably for escaping environmental extremes and avoiding potential predators. We a lso found evidenc e of both juvenile and adult pythons undergoing ecdysis within burrows Two python s were observed in the process of shedding within burrow s suggesting that pythons may use the safety of burrows to complete ecdysis (a period where their vi sual acuity is reduced ) or potentially use the higher humidity within burrows to aid in the ecdysis process Several observations provid e evidence that pythons use burrows for various reproductive purposes, as well. One burrow contained an aggregation of t hree pythons including two mature males and one mature female suggesting that the snakes aggregated for reproductive purposes This aggregation behavior has also b een observed in previous years (Bartoszek et al., 2018). Additionally, two tagged female pyt hons laid and incubated their respective clutches of eggs inside gopher tortoise burrows during this study. The s e observations suggest that burrows may be very important to the life history of Burmese pythons in areas where pythons have access to them. We observed pythons co occupying burrows with gopher tortoises on seven occasions throughout the study although potential impacts that pythons may have on tortoises are currently unknown Two occasions of co occupancy involved the two mature female pythons that each laid and incubated eggs within their respective burrow. These pythons each occupied their respective tortoise burrow for over two months, resulting in the displacement of the resident tortoises ( observed via camera trap data) Displacement from i ts burrow could potentially result in prolonged exposure to harsh

PAGE 48

48 e xternal temperatures and leave the tortoise vulnerable to predators. Additi onally, a large aggregation of seven reproductively mature Burmese pythons was found co occupying a burrow with a gopher tortoise in 2015 (Bartoszek et al., 2018). In this event, likely affecting its thermoregulation. Fortunately, predation by pythons on gopher tortoises has not been observed although other large constrictors including green anacondas have been observed depredating various species of Testudines throughout their native range (Rivas 1999). H atchling and juvenile gopher tortoises are also depredated by many upland species including several species of snakes (Douglass & Winegarner 1977; Butler & Sowell 1996). Additionally, we found that Burmese pythons select for smaller burrows, which could impact juvenile and subadult gopher tortoises, age classes with low survi val rates in many populations. This could affect tortoise recruitment rates and have implications on gopher tortosies at a population level. Gopher tortoises are considered to be a keystone species, and greatly benefit the ecosystem s in which they are foun d ( Landers & Speake 1980; Diemer & Speake 19 81; Campbell & Christman 1982; Eisenberg 1983 ) Burmese pythons are significantly impacting native animals within their Florida range ( Dorcas et al., 2012; McCleery et al., 2015 ), and could potentially have n egative effects on species that are endemic to upland habitats, including gopher tortoises. Burmese pythons are generally associated with lowland habitats and bodies of water ( Barker & Barker 2008 ), however hab itat use by pythons throughout s outhwest ern F lorida has been highly variable (per s. obs.) Pythons use a range of habitats in this region, including lowla nd habitats, agricultural matric es, and upland

PAGE 49

49 habitats ( pers. obs. ). Some features of upland habitats are very conducive for the establishment of pythons Upland habitats in s outhwest ern Florida have fairly high densities of various mammal species including rodents, raccoons, opossums, marsh rabbits, bobcats, feral hogs, and white tailed deer (per s. obs.), all of which have been found in gut content s of Burmese pythons in Florida (Romagosa unpublished data). Additionally, sources of fresh water, such as ponds and wetlands, are common within most upland habitats in the region. High densities of prey items combined with high availability of refuge and water could facilitate higher rates of survivorship for pythons of all age classes. I n cases where animal species within the same local region occur in several distinct types of habitat, such as th e case with Burmese pythons in s outhwest ern Florida, certa in life history traits including development rates, life spans, birth rates, and death rates may differ depending on the type of habitat in which each individual occurs (Pulliam & Danielson 1991). To maximize the success of management efforts, it is impor tant to understand the ecology of Burmese pythons within all habitats they utilize. T he data we have gathered for this study can be used to identify and predict areas within upland habitats that are most suitable for invasive pythons, thus improving python management efforts within Florida. Our finding that pythons select burrows located in areas with dense canopy cover suggests that searches should be focused in these areas. R educing cover through management techniques such as prescribed fire may also make habitats less suitable and discourage pythons from establishing Finally due to prolonged durations of burrow occupancy by female pythons during their nesting season i mplementing targeted searches during that time could help manage adult

PAGE 50

50 female pythons and python nests, potentially impacting Burmese pythons at a population level.

PAGE 51

51 Figure 3 1. A map of the upland scrub habitat used for this study, located wit hin Rookery Bay NERR in s outhwest ern Florida.

PAGE 52

52 Figure 3 2. ROC curve and respective a rea under the c urve (AUC) of the best mod el of python burrow selection (c anopy c over + b urrow e ntrance w idth).

PAGE 53

53 Table 3 1. Number of VHF telemetry locations per Burmese python used in analysis Python ID No. of locations Sex Size P1 3 Male 1.74 m / 2 .05 kg P2 2 Male P3 3 Male 3.21 m / 20.87 kg P4 4 Female 4.22 m / 43.09 kg P5 5 Male 3.60 m / 22.45 kg P6 1 Female P7 2 Male 3.20 m / 15.06 kg Table 3 2. Details of c ovariates u sed i n m odels of p ython b urrow s election Variable Description Unit s Max Min Avg Length Approximate l ength of burrow Cm 655 62.0 296 Width Width 50 cm inside burrow Cm 49.0 5.00 25.1 Orientation Orientation of burrow entrance N,E,S,W Tortoise Tortoise present in burrow or not Y or N Cover Canopy cover % 8 8.5 0.00 7.30 Water Distance to water m 394 44.7 269 Table 3 3 Ranking of single season mixed effects logistic regression models (using Python ID as a random intercept) used to describe Burmese python burrow selection in southwestern Florida. Model N ame a K b AIC b AIC b b (Width + Cover) 2 117.6976 0 0.9857 (Water Cover) 3 127.3871 9.6895 0.0078 (Cover) 1 128.9855 11.2879 0.0035 (Tortoise + Cover) 2 130.3543 12.6567 0.0018 (Orientation + Cover) 2 130.9063 13.2087 0.0013 (Width + Tortoise) 2 149.4748 31.7 772 <0.001 (Width Water) 3 150.678 32.9804 <0.001 (Width) 1 154.7898 37.0922 <0.001 (Length + Tortoise) 2 165.3325 47.6349 <0.001 (Length + Orientation) 2 168.2153 50.5177 <0.001 (Length) 1 168.5317 50.8341 <0.001 (Orientation + Water) 2 171.7172 5 4.0196 <0.001 (Water) 1 171.7484 54.0508 <0.001 a Model covariates include burrow length (Length), burrow entrance width (Width), burrow entrance orientation (Orientation), gopher tortoise presence within the burrow (Tortoise), canopy cover (Cover), and distance to water (Water). b K = number of variables in model, AIC = Akaike Information Criterion, AIC = difference between the

PAGE 54

54 CHAPTER 4 GOPHER TORTOISE BURROWS AS SUITABLE REFUGE FOR INVASIVE BURMESE PYTHONS NORTH OF THEIR CURRENT RANGE IN FLORIDA Synopsis Species distributions are commonly influenced by a variety of factors, and often shift and expand over time. The range margins of distributions can serve as areas of diversification where species adapt to novel environments. Determining how species respond to novel environments and which factors influence species adaptations is very applicable to invasive species ecology. I nvasi ons of non native species often spur rapid evolutionary events, and several studies have provided evidence of invasive species undergoing rapid evolutionary change during their invasion, especially at the range margins of the invasion. In southwestern Florida, near the presumed northern margin of their invaded range, Burmese pythons have been commonly observed occupying upland habitats and using gopher tortoise burrows within these habitats for various purposes. As gopher tortoise burrows typically experience relatively little temperature fluctuation, they are c ommonly used by many native species, including gopher tortoises, for refuge and to avoid temperature extremes. We assessed the internal environements of gopher tortoise burrows along a latitudinal gradient throughout Florida and southwestern Georgia to eva luate the possibility of Burmese pythons using tortoise burrows to survive environmental extremes north of their current range. We found that internal environments within tortoise burrows vary based on latitude and burrow length, but overall maintain tempe ratures conducive for Burmese pythons to survive, even as far north as southwestern Georgia. Burmese pythons have not been observed using subterranean refugia to overwinter for extended periods in their invaded or native

PAGE 55

55 ranges, but could potentially adopt such a behavior and use burrows to survive in areas that would typically be too cold. Background distribution and range limits is one of the fundamental challenges of ecology (Krebs, 1978). Species ranges ar e often variable, shifting and expanding over time (Brown et al., 1996; Davis & Shaw 2001; Gaston 2003 ). This mobility makes it challenging for ecologists to recognize which conditions promote niche conservatism versus those that facilitate adaptive dive Gomulkiewicz, 1997; Sexton et al., 2009). Range margins can serve as population sinks where adaptation is prevented by small population size and maladaptive gene flow or as areas of diversification where niche evolution occurs during adaptation to novel environments (Gaston, 2003; Sexton et al., 2009). Unfortunately, few empirical studies have been conducted examining the factors contributing to population differentiation at range margins. Determining which factors influence species adaptations at range margins is very applicable to invasive species ecology. I nvasi ons of non native species often spur rapid evolutionary events, resulting in populations that are genetically d ynamic over both time and space ( Rez nick & Ghalambor, 2001 ; Lee, 2002). Several studies have provided evidence of invasive species undergoing rapid evolutionary change during their invasion, especially at the range margins of the invasion ( Thomas et al., 2001; Lee 2002; Travis & Dytham, 2002; Cox 2004 ; Phillips et al., 2006; Darling et al., 2008 ). Some of the most common examples of adaptation by invasive species include physiological tolerance to environmental stress, often involving responses to novel

PAGE 56

56 climate conditions ( Potvin & Simon, 1989; Williams & Moore, 1989; Lee, 1999; Huey et al., 2000 ). I nvasive Burmese pythons ( Python bivittatus ) have established throughout southern Florida and are negatively impacting native wildlife ( Snow et al., 2007 ). B urmese python s have been implicated in the severe declines of meso mammal populations within their invaded range ( Dorcas et al., 2012; McCleery et al., 2015 ) and have also been documented eating threatened and endangered species, such as the Key Largo wood rat ( Neotoma floridana smalli ) and the wood stork ( Mycteria americana ) (Snow et al., 2007; Reed & Rodda, 2009). There is uncertainty in the geographic range Burmese pythons could establish in the United States. Some species distribution models predict that pythons are restricted to the range they presently inhabit in southern Florida (Pyron et al., 2008). However, Rodda et al (2011) corrected several inaccuracies reported in Pyron et al (2008), and found that Burmese pythons could potentially establish th roughout most of Florida. The extends well into temperate climate zones in China and the Himalayas (Groombridge & Luxmoore 1991; Zhao & Adler 1993 ; Whitaker et al., 2004 ), suggesting that they may be more tolerant of tempera te conditions than widely thought. Geographic distribution, particularly among ectothermic animals, is commonly limited by temperature, especially minimum temperatures during winter months (Ultsch, 2006). However, many animals survive extreme conditions b y limiting their exposure. Ectothermic species commonly survive in cold climates by using behavioral thermoregulation, often involving the selection of subterranean or aquatic hibernation sites, to avoid exposure to freezing temperatures (Storey, 1990). In many cases, the

PAGE 57

57 likelihood of winter survival is dependent on the timely discovery and selection of proper overwintering refugia (DeGregorio et al., 2012). In southwestern Florida, near the presumed northern margin of the Burmese ution, pythons use both lowland and upland habitats (per s. obs.). These habitats include mangrove forests, flooded grasslands, agricultural matrices, pine and oak hammocks, and Florida scrub, among others. Pythons occurring within xeric upland habitats in southwestern Florida commonly use animal burrows a behavior that is also occasionally seen within their native range of Southeast Asia ( Sharma & Kandel, 2016) Many of the burrows that pythons use in this region of Florida are those created by nine banded armadillos ( Dasypus novemcinctus ) or gopher tortoises ( Gopherus polyphemus ). As gopher tortoise burrows typically experience relatively little temperature fluctuation (DeGregorio et al., 2012), they are commonly used by many native species, including the gopher tortoise itself, for refuge and to avoid temperature extremes (Lips, 1991). Burrow use by pythons is significant because it may have implications for the geographic range that pythons could potentially inhabit in the southeastern United States. Our objective was to assess gopher tortoise burrows as suitable overwintering refugia sites for Burmese pythons north of their current range in Florida. We used a comparative design to investigate and describe potential physical and microclimatic differences between gopher tortoise burrows located within the southwestern range of Burmese pythons in Florida and gopher tortoise burrows located north of that range. We expected microclimatic characteristics of tortoise burrows to show varia tion among regions, but maintain temperatures suitable for pythons to survive.

PAGE 58

58 Study Area Python Occupied Site We conducted part of the study within Rookery Bay National Estuarine Research Reserve (RBNERR) located near Naples in s outhwest ern Florida. Situated near the edge of t he Greater Everglades Ecosystem, RBNERR is comprised predominantly of mangrove forests a habitat that covers over 45,000 ha of the reserve. Patches of upland hammock and scrub habitats can also be found intermittently within the mangrove forests. These ar e some of the southern most patches of upland scrub habitat in Florida, and are among the first upland areas that Burmese pythons have had access to over the course of their establishment in Florida. One such scrub patch located within the reserve has been monitored for pythons since 2014 with documented occurrences of pythons occupying gopher tortoise burrows every year since then. This patch served as the python occupied study area for this objective. Northern Sites We also conducted the study at severa l sites located north of the current range of Burmese pythons. Northern sites were selected based on their geographic location and amount of respective gopher tortoise habitat. These sites included the Archbold Biological Station (ABS ; 27.1829 N, 81.3521 W; 130 km from RBNERR ), Ashton Biological Preserve (ABP ; 29.5431 N, 82.5800 W; 400 km from RBNERR ), and Joseph Jones Ecological Research Center (JJERC ; 31.2205 N, 84.4792 W; 630 km from RBNERR ). ABS is locat ed in south c entral Florida at the southern tip of Lake Wales Ridge near the headwaters of the Greater Everglades Ecosystem. Of the three northern sites, ABS is the closest to the current range of Burmese pythons in Florida. ABS is comprised of relict sandy dunes, including scrub habitats and flatwo ods, along

PAGE 59

59 with wetlands, ranchlands, and lakes. AB P is located in north c entral Florida on the edge of Brooksville Ridge. It is comprised of rolling hills with xeric upland and sandhill habitats, along with wooded hammocks and wetlands. Finally, JJERC is lo cated in s outhwest ern Georgia on the Dougherty Plain. It is comprised of a diverse range of ec ological communities including longleaf pine forests mixed pine hardwoods, riparian pine hardwood forests, agricultural fields, depressional wetlands, rivers, and creeks. All three of these sites contain populations of gopher tortoises and relatively high densities of tortoise burrows Methods Burrow Selection We selected ten gopher tortoise burrows from each of the three northern sites (ABS, ABP, & JJERC) and th e python occupied site (RBNERR) for use in this study. Bur row locations within the python occupied site (RBNERR) were provided from a previous study (Chapter 2), and we randomly selected ten of the previously located unoccupied tortoise burrows to be use d for this study. Only unoccupied, but structurally sound, tortoise burrows were used for this study to increase the likelihood that temperature loggers remained in burrows for the duration of the sampling season. Previous experience showed that gopher tor toises commonly buried or removed temperature loggers located within their burrows. We used organized searching to locate the ten burrows to be used at each northern site (ABS, ABP, & JJERC). Areas of suitable gopher tortoise habitat were first identified prior t o surveying, and a group of one to four people then systematically surveyed the habitat until a burrow was located. Once located, we scoped the burrow using a burrow camera to determine its occupancy status. If unoccupied by a tortoise, the burrow w as selected to be used for the study.

PAGE 60

60 Several areas of suitable tortoise habitat were surveyed for burrows at each site to minimize potential pseudo sampling and to ensure that multiple areas were represented. Habitat Characteristics We measured habitat v ariables, physical characteristics, and microclimate characteristics of each selected burrow to compare tortoise burrows along a latitudinal gradient At each burrow we recorded canopy co ver (%), as the percentage of vegetati ve cover waist height and above measured at the burrow apron using a densitometer. We collected air temperature ( o C) from weather stations located within or in close proximity to each site. The following burrow characteristics were also measured for each burrow: burrow length (cm), burr ow entrance width (cm), burrow orientation (N, S, E, W) and burrow temperature ( o C) We recorded burrow length as the approximate length from the entrance of the burrow to the back wall, and measured it by marking the burrow scope with a white marker We recorded burrow entranc e width in accordance with Smith et al. (2009) as the width 50 cm within the burrow and measured it using calipers We recorded b urrow orientation as the direction that the burrow entrance faced and measured it using a compass. We th en converted t he continuous degree measurements into the four cardinal directions (316 o 45 o = N, 46 o 135 o = E, 136 o 225 o =S, 226 o 315 o =W) Finally, we recorded burrow temperature using an iButton Thermochron temperature data logger ( 40 o C to +85 o C capability) set to record once every four hours and placed near the back of the burrow. We connected each temperature logger to a thin rope and pushed them to the back of each burrow using a small hook on a burrow scope ( Environmental Management Systems; Canton, GA) Temperature loggers were left in burrows from January 2018 to May 2018.

PAGE 61

61 Data Analysis We measured physical and microclimatic characteristics of gopher tortoise burrows within the four sites. We used analysis of variance (ANOVA) in program R v ersion 3.1.2 to determine how physical and microclimatic characteristics of gopher tortoise burrows varied within and between each site. We also used linear models (LM) in program R version 3.1.2 to determine the influence of latitude and various physical burrow characteristics on the minimum temperature, maximum temperature, and average temperature within gopher tortoise burrows. We ranked m odels using Akaike Information Criterion corrected for small sample size (AIC c ; Sugiura, 1978 ) and AIC c weight ( and varies from 0 (no support) to 1 (complete support). I considered any model within 2.0 AIC c units of the best model to be competing models. Results Four of forty temperature loggers failed to collect data Three more temperature loggers were pulled out of their respective burrows by animals during the sampling season, and were not recovered. Two more temperature loggers were pulled out of their respective burrows in February 2 018, so their data were no t used in analysis. In total, thirty one temperature loggers recorded temperature data from January 15, 2018 until May 23, 2018 (Table 4 1). Minimum burrow temperature (F = 22.46, p < 0.01), maximum burrow temperature (F = 37.99, p < 0.01), and average burrow temperature (F = 345.6, p < 0.01) varied among sites (Table 4 2). ANOVA results from both the minimum burrow temperatures and the maximum burrow temperatures showed slight variation within

PAGE 62

62 each site, while results from the av erage burrow temperatures showed very little evidence of variance within each site. We used linear regression to determine the influence of latitude and various physical burrow characteristics on the minimum temperature, maximum temperature, and average temperature within gopher tortoise burrows. The best approximating model of minimum burrow temperature included sampling site and burrow length (Table 4 3) The best approximating model of maximum burrow temperature included canopy cover and burrow length (Table 4 4) Finally, t he best approximating model of average burrow temperature included sampling site and canopy cover (Table 4 5) Discussion Our study provides further information about the characteristics of gopher tortoise burrows and how the physic al and microclimatic conditions vary along a latitudinal gradient. Our results suggest that gopher tortoise burrows can serve as suitable overwintering refugia for Burmese pythons north of their current range in Florida. We measured physical and microclima tic characteristics of burrows within Rookery Bay National Estuarine Research Reserve (RBNERR), located on the northern margin of the Burmese pythons are commonly observed occupying gopher tortoise burrows throughout RBNERR, located in southwestern Florida. Microclimatic characteristics of burrows, including minimum burrow temperature, maximum burrow temperature, and average burrow temperature, all showed significant variat ion among sites (Table 4 2). Despite the aforementioned inter site variation in thermal ch aracteristics, thirtry of thirty one burrows maintained temperatures above those described as lethal for Burmese pythons, even when external temperatures dropped wel l below the lethal limit

PAGE 63

63 (~5 o C, Dorcas et al., 2011). The one burrow that dropped (3 o C) below the lethal limit described for Burmese pythons was the shortest burrow included in the study, measuring approximately 90 cm total length. In their native range Burmese pythons are generally associated with lowland habitats and bodies of water (Barker & Barker 2008 ), however habitat use by pythons throughout their s outhwest ern Florida range has been highly variable (per s. obs.) Pythons use a range of habitats i n this region, including mesic lowland habitats, agricul tural matric es, and xeric upland habitats (Bartoszek unpublished data) Pythons commonly use upland habitats, contradicting the presumption made in Pyron et al., (2008) that the proposed expansion of Burmese pythons into the continental United States would require an expansion of the tropical marshland comprising most of the Everglades. In addition to access to tropical ma rshland habitats, the species distribution models presented in Pyron et al (2008 ) predict ed that pythons are limited by their critical thermal minimum, restricting them to the range they presently inhabit in the Un ited States However, Rodda et al (2011) corrected several inaccuracies reported in Pyron et al (2008), and found that B urmese pythons could potentially establish throughout most of Florida. M any speci es distribution models only include climate data, and neglect to include variables related to availability of prey and suitable refuge. The importance of burrows and other sub terranean refugia is often overlooked, as they provide protection against temperature extremes, fire, and predation ( Campbell & Clark, 1981; Alkon & S altz, 1988; Reichman & Smith, 1990; Mukherjee et al., 2017). Furthermore, little is known about Burmese py thon ecology or the factors that delimit

PAGE 64

64 any part of the ir native range (Rodda et al., 2009), suggesting that these distribution models may be speculative, at best. Two studies conducted in north central Florida and South Carolina have provided evidence th at Burmese pythons may not be capable of surviving severe winters, although pythons may withstand conditions substantially cooler than those typically experienced in South Florida (Avery et al., 2010; Dorcas et al., 2011). In both cases, pythons exhibited inappropriate thermoregulatory behavior or did not have access to proper subterranean refugia, resulting in their deaths (Avery et al., 2010; Dorcas et al., 2011). While overwintering behavior by Burmese pythons has not yet been described in its native or nonnative range the Indian python ( Python molurus Linnaeus 1758) to which the Burmese python is closely related has been reported to overwinter in animal burrows for up to f our months in parts of northern Pakistan ( Minton, 1966; Rodda et al., 2009) Ad climate zones in China and the Himalayas (Groombridge & Luxmoore 1991; Zhao & Adler 1993 ; Whitaker et al., 2004 ), suggesting that they may be more tolerant of temperate conditions than widely thought. The purpose of this study was to determine whether microclimatic conditions within gopher tortoise burrows are conducive for Burmese pythons to survive extreme temperatures, should they adopt any overwintering behavior. We found that g opher tortoise burrows, given the appropriate length, maintain temperatures above the lethal limits of Burmese pythons (Dorcas et al., 2011), even when external temperatures drop well below those limits. We recommend further field trials be conducted, usin g natural refugia and pythons that are accustomed to using gopher tortoise burrows for refuge.

PAGE 65

65 Figure 4 1. Locations of sites used to assess physical and microclimatic characteristics of gopher tortoise burrows. JC = Joseph Jones Ecological Research Center, AS = Ashton Biological Preserve, AR = Archbold Biological Research Station, RB = Rookery Bay National Estuarine Research Reserve.

PAGE 66

66 Figure 4 2. Maximum, minimum and average burrow temperatures of each burrow at four different sites from January May 201 8

PAGE 67

67 Figure 4 3. External temperatures and average internal temperatures of gopher tortoise burrows (n=9) from the Joseph Jones Ecologic al Research Center, located in s outhwest ern Georgia, well north of the current range of Burmese pythons. External t emperatures obtained from University of Georgia climate station located within JJER C. Python lethal limit set at 5 o C, the conservative limit described in Dorcas et al., (2011).

PAGE 68

68 Figure 4 4. External temperatures and average internal temperature s of gopher tortoise burrows (n=8) from Ashton Biological Preserve, located well north of the current range of Burmese pythons in Florida. External temperatures obtained from Weather Underground climate station located near ABP. Python lethal limit set at 5 o C, the conservative limit described in Dorcas et al., (2011).

PAGE 69

69 Figure 4 5. External temperatures and average internal temperatures of gopher tortoise burrows (n=7) from Archbold Biological Station, located in close proximity to the current ra nge of Burmese pythons in Florida. External temperatures obtained from ABS climate station. External temperatures unavailable 3/10/2018 3/21/201 8. Python lethal limit set at 5 o C, the conservative limit desc ribed in Dorcas et al., (2011)

PAGE 70

70 Figu re 4 6. External temperatures and average internal temperatures of gopher tortoise burrows (n=7) from Rookery Bay NERR, located within the current range of Burmese pythons in Florida. External temperatures obtained from Weather Underground climate station located in close proximity to research site (Isles of Capri, FL ). Python lethal limit set at 5 o C, the conservative limit desc ribed in Dorcas et al., (2011)

PAGE 71

71 Table 4 1. Number of iButton temperature loggers used at each sampling site. Loc ation a Dates b # iButtons Used b Jones Center 01/02/2018 05/22/2018 9 Ashton Preserve 01/08/2018 06/01/2018 8 Archbold 01/15/2018 05/23/2018 7 Rookery Bay 01/15/2018 05/20/2018 7 a Locations include the Joseph Jones Ecological Research Center (Jones Center), Ashton Biological Preserve (Ashton Preserve), Archbold Biological Research Station (Archbold), and Rookery Bay National Estuarine Research Reserve (Rookery Bay). b Dates = the period that the iButton was recording temperature, # iButtons U sed = the number of iButtons that were used in analysis. Table 4 2. Results of analyses of variance (ANOVA) Variable F value P value Minimum Burrow Temp. 22.46 <0.01 Maximum Burrow Temp. 37.99 <0.01 Average Burrow Temp. 345.6 <0.01 Burrow Length 1.33 0.258 Burrow Entrance Width 8.58 <0.01 Orientation 1.67 0.206 Canopy Cover 126.3 <0.01 Table 4 3. Ranking of single season linear regression models used to describe Minimum Burrow Temperature. Model Name a K b AIC c b c b b (Site + L ength) 2 182.1304 0 0.7185 (Site + Cover + Length) 3 184.0314 1.901 0.2777 (Cover + Length) 2 192.7829 10.6525 0.0035 (Cover + Orientation) 2 199.0694 16.939 <0.001 (Site) 1 201.5777 19.4473 <0.001 (Site + Cover) 2 202.7301 20.5997 <0.001 (Length + O rientation) 2 213.1038 30.9734 <0.001 (Length) 1 214.5099 32.3795 <0.001 a Model covariates include sampling site (site), burrow length (Length), burrow entrance width (Width), burrow entrance orientation (Orientation), and canopy cover (Cover). b K = n umber of variables in model, AIC c = A ka i ke Information Criterion corrected for small sample size AIC c = difference between the AIC c value of each model and the lowest AIC c c weight.

PAGE 72

72 Table 4 4. Ranking of single season linear regression models used to describe Maximum B urrow Temperature. Model Name a K b AIC c b c b b (Cover + Length) 2 139.8935 0 0.569 (Site + Cover + Length) 3 141.863 1.9695 0.213 (Cover + Orientation) 2 142.7525 2.859 0.136 (Site + Cover) 2 144.5489 4.6554 0.056 (Site) 1 147.265 7.3715 0.014 (Site + Length) 2 147.6559 7.7624 0.012 (Length) 1 170.2474 30.3539 <0.001 (Length + Orientation) 2 172.2249 32.3314 <0.001 a Model covariates include sampling site (site), burrow length (Length), burrow entrance width (Width), burrow entrance orientat ion (Orientation), and canopy cover (Cover). b K = number of variables in model, AIC c = A ka i ke Information Criterion corrected for small sample size AIC c = difference between the AIC c value of each model and the lowest AIC c c weight. Table 4 5 Ranking of single season linear regression models used to describe Average Burrow Temperature. Model Name a K b AIC c b c b b (Site + Cov er) 2 106.3411 0 0.698 (Site + Cover + Length) 3 108.3282 1.9871 0.258 (Site) 1 112.7251 6.384 0.029 (Site + Length) 2 113.9555 7.6144 0.015 (Cover + Length) 2 132.1166 25.7755 <0.001 (Cover + Orientation) 2 134.157 27.8159 <0.001 (Length) 1 191.2541 84.913 <0.001 (Length + Orientation) 2 192.6038 86.2627 <0.001 a Model covariates include sampling site (site), burrow length (Length), burrow entrance width (Width), burrow entrance orientation (Orientation), and canopy cover (Cover). b K = number of variables in model, AIC c = A ka i ke Information Criterion corrected for small sample size AIC c = difference between the AIC c value of each model and the lowest AIC c c weight.

PAGE 73

73 CHAPTER 5 TRAPPING BURMESE PYTHONS FROM GOPHER TORTOISE BURROWS Synopsis Invasive species manag ement is among the most challenging conservation issues confronting the state of Florida this century with at least sixty three non native species of reptiles and amphibians currently established. Among these are Burmese pythons, which have been implicate d in the severe declines of mammal populations in southern Florida and are substantially impacting mammal diversity within the Florida Everglades. In southwestern Florida, Burmese pythons have been commonly observed occupying gopher tortoise burrows for va rious purposes When found in burrows, pythons are typically removed via excavation, a process that can sometimes damage the integrity of the burrow. Given the importance of burrows to native species, it is imperative that the structural integrity of torto ise burrows is maintained whenever possible. We developed and tested a new trap designed to remove pythons from burrows without damaging the burrow. While our design has limitations, it successfully captured pythons of various sizes in five of eight attemp ts. Background At least sixty three non native species of amphibians and reptiles are currently esta blished in Florida ( Krysko et al., 2016 ) including three species of large constricting snakes belonging to the families Pythonidae (Cope 1864) or Boidae ( Gray 1825) These species include the Burmese python ( Python bivittatus ), the Northern African python ( Python sebae ), and the boa constrictor ( Boa constrictor ; Snow et al., 2007; Reed et al., 2010 ). All three of these species use animal burrows to some ext ent in their native ranges ( Peirce, 1974; Montgomery & Rand 1978 ; Sharma & Kandel 2016 ) and are

PAGE 74

74 also likely using animal burrows wit hin their invaded ranges Two of the species, the boa constrictor and the Burmese python, have been observed using gopher tortoise ( Gopherus polyphemus ) burrows w ithin s outh eastern and s outhwest ern Florida, respectively ( per s. obs., M etzger, 2013; Bartoszek et al., 2018 ). Burmese pythons have been implicated in the severe declines of mammal populations in s outh ern Florida ( Do rcas et al., 2012; McCleery et al., 2015 ), making management of their expansion into new habitats of the utmost concern. Burmese pythons have been observed using gopher tortoise burrows regularly in s outhwest ern Florida, with the first record occurring in December 2010 ( Metzger, 2013 ). P ythons use burrows for refuge and reproductive purposes and will occasionally aggregate within active tortoise burrows ( Bartoszek et al., 2018 ). The eff ects that Burmese pythons may have on native burrow using species are c urrently unknown, but pythons could negatively affect native species though predation, competition, displacement, or the spread of disease. Due to the ecological damage attributed to pythons and their status as an invasive non native species, they should be removed when found. Removing pythons from burrows can be challenging, however, as gopher tortoise burrows commonly exceed 4.5 m in length (Hallinan 1923; Hansen 1963). Prior to 2017, burrow excavation was the only reliable method for removin g Bur mese pythons from burrows. Gopher tortoises are listed as a threatened species in Florida, and excavation of a burrow requires a permit. Additionally, g opher tortoise burrows are important features within the habit ats they are found, with over 36 0 native animal species using tortoise burrows for various purposes ( Hallinan, 1923; Young & Goff, 1939; Brode, 1959; Hutt,

PAGE 75

75 1967; Auffenburg, 1969; Landers & Speake, 1980; Woodruff, 1982; Eisenberg, 1983; Jackson & Milstrey, 1989 ) Given their importance to the tortoise and other native species, it is imperative that the structural integrity of tortoise burrows is maintained whenever possible. Our objective was to design and test a new trap to remove invasive pythons from gopher tortois e burrows without damaging the stru ctural integrity of the burrow via excavation. Our trap design is loosely inspired by the existing gopher tortoise flap trap described in Enge et al. ( 2012 ) We designed the python trap to safely and effectively capture pythons as they exit the burr ow. Study Area We conducted the study within Rookery Bay National Estuarine Research R eserve, located near Naples in s outhwest ern Florida. Situated near the edge of the Greater Everglades Ecosystem, Rookery Bay is comprised predominantly of mangrove fores ts, a habitat that covers over 110,000 acres of the reserve. Patches of upland hammock and scrub habitats can also be found intermittently within the mangrove forests. These are some of the southern most patches of upland scrub habitat in Florida, and are among the first upland areas that Burmese pythons have had access to over the course of their establishment in Florida. One such scrub patch located within the reserve has been monitored for pythons since 2014 with documented occurrences of pythons occupy ing gopher tortoise burrows every year since then. This patch served as the study area for this objective.

PAGE 76

76 Methods Trap Construction We designed the python trap with two components: a trap and an extension that allowed the python to fu lly enter the trap (Figure 5 1 ) The trap design was modeled off of Enge et al (2012). We constructed the trapping component using 16 gauge galvanized steel mesh wire, available at most hardware stores. We cut two sheets each measuring roughly 91 x 152 cm (one 91 x 15 2 cm sheet ma de a single trapping component) and made two 90 o folds in the 91 x 152 cm sheet of wire mesh approximately 56 cm from both ends (Figure 5 1) We then folded each long side inward at 90 o and used cable ties to securely attach the folded ends in the middle, enclosing the sides of the trap. This resulted in a rectangular cage measuring roughly 56 cm long x 51 cm wide x 41 cm high, with all sides enclosed except the front. We gently bent the wire mesh at the front of the trap to taper it slightly, allow ing the extension component to better attach. Next, we fitted a hinged door to the bottom of the trapping component to act as the trapping mechanism (Figure 5 1) To do this, we flipped the trapping component upside down and, u sing wire cutters, cut a 33 x 28 cm square shaped opening in the rear center of the bottom of the trap This would serve as the opening through which a python could enter the trap. A piece o f galvanized steel sheet metal was cut to approximately 36 x 30 cm to allow it to fit ove r the opening on the bottom of the trap. The sheet metal was then positioned overtop the opening on the inside of the trap and attached to the bottom using a stainless steel hinge Stainless steel screws and washers were used to attach the hin ge to both the sheet metal and the trap.

PAGE 77

77 We then constructed the extension component. The purpose of the extension component was to lengthen the trap, allowing room for the python to completely exit the burrow so that the hinged door could fall behind the python and trap it inside. Three types of extension components were tested: a canvas bag, a mesh bag, and a 16 gauge galvanized steel mesh wire cage. See Results for assessments of each of the extension types. Material for canvas bags and mesh bags were s ewn into 183 x 40 cm cylinders using a sewing machine We constructed the wire cage extension type using 16 gauge galvanized steel mesh wire. A piece of wire mesh measuring roughly 91 x 183 cm was folded length wise three times into a long rectangular cage with an entrance diameter of 40 cm (Figure 5 1) We then bent the wire mesh on the back of the cage folding it against itself so that all sides were enclosed but the front, where it would connect to the trapping component. We cut small pieces of the wire mesh measuring roughly 25 x 15 cm and used zip ties to loosely connect them to each side of the front of the extension component (Figure 5 1) These pieces allowed the extension component to connect to the trapping component with no space for the python t o escape or injure itself. Python Detection We located burrows occupied by pythons using random burrow surveys and telemetry tagged pythons. We randomly surveyed g opher tortoise burrows using a burrow scoping camera designed by Environmental Management Sy stems ( Canton, GA) and used the python trap to remove any python we detected in a burrow (Chapter 2) We also tested the python trap on three telemetry tagged pythons that needed to be removed from the field for equipment maintenance.

PAGE 78

78 Trappi ng We used the python trap to capture Burmese pythons during 2017 and early 20 18. All trapping took place in s outhwest ern Florida, within scrub and pine upland habitats where gopher tortoises were present. When a python was detected in a burrow veg etation around the burrow entrance was cleared so that the trap could lie flat on the ground without obstruction. We aligned the trap so that the hinged door on the bottom of the trapping component was directly above the entrance to the burrow. Alignment v aried slightly depending on the angle of the burrow entrance and surrounding habitat. We then secured the trapping component to the ground using four spiral shaped stakes, attaching one stake to each corner of the trap and twisting them firmly into the gro und. We packed sand in gaps between the trap and the ground, assuring that there was no space through which the python could escape or injure itself. We attached the extension component to the opening at the front of the trapping component using small ela stic cords or metal clips. This was done after the trapping component was secured to the ground so that the trap did not shift off the burrow entrance while attaching the extension component. To prevent prolonged sun exposure to any captured animals, we us ed palm fronds and snake bags to cover portio ns of the trap. Once the python trap was properly set we placed a trail camera within view of the trap to determine the exact moment the python was captured. Traps were checked every two hours during day light and first thing in the morning to assure the safety of any animals captured (Figures 5 3 & 5 4) Results During 2017 and 2018, we tested the python trap at eight tortoise burrows occupied by pythons Pythons were successfully captured in five of eight attempts

PAGE 79

79 (Table 5 1) T wo male pythons and three female pythons were successfully captured (Table 5 1) Captured pythons ranged in size from 1.74 m TL / 2.05 kg to 4.33 m TL / 34.6 kg. Trapping success was most affected by the type of extension component attached to the trap. The canvas bag extens ion was successful in one out of three attempts and the mesh bag extension was unsuccessful in one attempt. The mesh bag extension was tested only a single time because the python ripped through the bag. The wire mesh cage extension was the most successful ext ension type, capturing pythons four times in four attempts (Figure 5 2 5 3) The duration of successful trapping attempts varied from approximately 3.47 hrs to 73.03 hrs after trap placement with an ave rage of 31.72 hrs per attempt. F our of t he five trapped pythons entered the trap between 19:00 and 20:00 hrs (Table 5 1) Over the eight trapping attempts, we did not observe any signs that the traps disturbed native wildlife. The traps did not capture any non target species, although it has yet to be tested on a burrow containing both a python and a gopher tortoise. Trapping attempts were deemed unsuccessful if a python showed signs of activity ( via trai l camera) but was not captured within five days of detection. Attempts were limited to five days because, although there was no sign that the traps disturbed native wildlife, we wanted to minimize the amount of time that burrows were inaccessible to tortoi ses and other commensal species. When trapping attempts were deemed unsuccessful, we exchanged the extension component of the trap and began a new trapping attempt.

PAGE 80

80 Of the three unsuccessful trapping attempts, two occurred while using the canvas bag exten sion type and one occurred while using the mesh bag extension type. During the attempts using the canvas bag extension, pythons showed interest in exiting the burrow (seen via trail camera), but refused to enter the canvas bag. In both of these attempts, t he python was successfully captured after the canvas bag extension was exchanged for a wire mesh cage extension. During the attempt using the mesh bag extension, the python completely entered the trap but ripped through a corner seam of the bag and escaped It is worth noting, however, that trail camera footage showed the python was trapped overnight within the mesh bag for roughly five hrs before breaking free. Discussion We found the python trap to be a feasible and effective alternative to burro w excavation when an invasive python needs to be removed from a gopher tortoise burrow. The sample size is small due to the amount of effort needed to detect pythons, but the python trap shows promise The trap has successfully captured both juvenil e and adult pythons, showing that trapping pythons of all age classes is very possible. Juvenile pythons weighing only 2 3 kg, mature male pythons weighing 16 20 kg, and a mature female python weighing over 34 kg were all successfully captured (Table 5 1) The trap has yet to be tested on hatchling pythons. The python trap is strong and durable enough to withstand hours of pushing and prodding by large constricting snakes, while maintaining its structural integrity during long periods of exposure to direct sunlight or intense rain. Trapping duration was relatively short in most cases, ranging from 3.47 hrs to 73.03 hrs with an average of 31.72 hrs per attempt (Table 5 1). Reasons for this variation are not understood, but

PAGE 81

81 could include environmental c onditions, the duration of burrow occupancy before the python was detected, or if the python was undergoing ecdysis The trap design is simple, with few parts that can malfunction or affect trap success. Trap set up is easy and efficient typically taking less than 30 minutes once a python is located within a burrow Mos t importantly, using the python trap prevents any substantial damage to the burrow, which allows tortoises and other native animals to continue to use them for refuge after trapping The eight trapping attempts also shed light on several interesting aspects of Burmese python behavior. Of the pythons captured during winter, all four emerged from their respective burrows betwee n the hours of 19:00 and 20:00, roughly an hour after sunset The python captured in mid April a 34 kg female, emerged from her burrow at 12:23. Upon necropsy she was found to be gravid and was likely preparing to nest Her time of emergence reflects the fact that gravid female Burmese pythons typically bask for several hours during the heat of the day presumably to aid in physiological processes such as follicle development ( pers. obs. ). It is also worth noting that Burmese pythons seem to refrain from entering python traps that they cannot completely se e through. Only when the canvas bag extension type was being used did the python refuse to fully enter the trap. The canvas bags were large enough to hold pythons and were propped open in several places with wire framing. Trail camera footage shows the pyt hons in both circumstances partially exiting the burrow and tongue flicking towards the bags, but refusing to enter and ultimately retreating back down the burrow. All bags were washed prior to use, so it is unlikely that leftover scent on the bag caused t his behavior. Pythons did completely enter the traps

PAGE 82

82 when more transparent extensions like the wire cage extension type or the mesh bag extension type were being used. While the python trap has been successful, it has limitations The most successf ul version of the trap including the wire cage extension is fairly cumbersome making it difficult to transport to remote locations The mesh bag extension type shows promise, as it is light weight and successfully captu red a python for several hours, b ut a stronger material such as aviary cloth, is needed to prevent pythons from ripping through Trapping success is also dependent on the trap being flush with the ground, providing no space for the python to escape through. This can be difficult when tortoi se burrows are on uneven terrain or in dense vegetation. In these circumstances, burrow excavation may be the only option for python removal. If Burmese pythons continue to spread north, they will encounter more xeric habitats including Florida scrub and upland hammocks. Pythons use gopher tortoise burrows within these habitats for refuge, mating, and nesting ( Metzger, 2013; Bartoszek et al., 2018) While burrow use by Burmese pythons may be potentially harmful to native species, it can also be used as a m anagement tool. Management of pythons is notoriously difficult, as their probability of detection is <1% in most habitats (Dorcas & Willson, 2013) and there is currently n o reliable population estimate However, detection of pythons within tortoise burrows given a python is present, is very high ( Chapter 2 ). Using burrow surveys in areas near the python invasion front may be a practical method for managers to assess python presence and rates of spread in the near future. This trap design may be effective on other large burrow dwelling invasive snake species, as w ell, including boa constrictors and A frican pythons As invasive Burmese

PAGE 83

83 pythons gain access to more upland habitats within Florida, they will also gain increased access to gopher tortoise burrows. As this occurs, removing pythons without damaging the structural integrity of the burrow will be crucial for protecting native species that rely on tortoise burrows for refuge and protection The python trap captures the invasive python while prote cting the burrow and all the native species that rely on it.

PAGE 84

84 Figu re 5 1 The python trap designed to capture large snakes as they exit gopher tor toise burrows. a ) The trapping component is constructed by folding 16 gauge galvanized steel mesh wire roughly in to a squar e b) while the extension component is folded into rectangle c) B ungie cords are used to conn ect the two components together. d) A hinged door is used as the trapping mechanis m e) spiral stakes are used t o secure the trap to the ground, f) and vegetation is used to provid e shade to captured animals g) A Burmese python i n the process of being captured

PAGE 85

85 Figu re 5 2 The python trap in situ after having captured a 4.33 m / 34.6 kg python

PAGE 86

86 Figu re 5 3 The python trap in situ af ter having captured a 1.96 m / 2.95 kg python

PAGE 87

87 Table 5 1. Details of trapping attempts. Extension Type Date Success? Duration (Hrs) Time of Capture TL (m) / Weight (kg) Sex Canvas Bag 01/11/17 Yes 3.47 19:10 1.74 / 2.05 Canvas Bag 01 /21/17 No NA NA 2.95 / 14.7 Canvas Bag 01/25/17 No NA NA 1.96 / 2.95 Mesh Bag 03/24/17 No NA NA 3.60 / 22.45 Wire Cage 01/21/17 Yes 4.28 19:04 2.95 / 14.7 Wire Cage 01/26/17 Yes 25.58 19:17 1.96 / 2.95 Wire Cage 04/20/17 Yes 73.03 12:23 4.33 / 34.6 Wire Cage 02/18/18 Yes 52.25 19:47 3.20 / 15.06

PAGE 88

88 CHAPTER 6 CONCLUSION The primary objective of this project was to assess and describe burrow use by Burme se pythons in their invaded s outhwest ern Florida range. We conducted three complimentar y studies focused on measuring the frequency and drivers of python burrow use, as well as implications gopher tortoise burrow use could have on the geographic range of pythons in Florida. We assessed how frequently gopher tortoise burrows were occupied by pythons within our stu dy site in southwestern Florida, as well as which burrow and habitat characteristics pythons selected for in upland habitats. Further, we compared characteristics of the gopher tortoise burrow environment along a latitudinal gradient to assess the potential for pythons to use burrows as overwintering refugia north of their current range Finally, we developed and tested a burrow trap to remove and capture pythons from burrows without damaging the str uctural integrity of the burrow. Th rough the use of both repeated surveys and radiotelemetry, we observed Burmese pythons using gopher tortoise bur rows within upland habitats in s outhwest ern Florida. While the average proportion of burrow s occupied during the sampling season was extremely l ow (0.359% of burrows ), telemetry tagged animals were found using burrows fairly frequently, suggest ing that the population density of pythons in and around the site may be lower than expected Low population density in the area could be attributed to inva sion rate ( i.e., the research site is near the presumed northern margin of the current range of Burmese pythons in Florida) or python removals Python removal efforts in the area (see Chapter 1) have been high, resulting in the removal of more than twenty mature pythons and four python nests from the study site since 2014.

PAGE 89

89 Additionally, our radiotelemetry results suggest that pythons are selecting burrows nonrandomly based on characteristics of the burrows and the surrounding habitat. P ythons occurring with in upland habitats sel ect ed for burrows that were smaller possibly due to more efficient thermoregulation in the smaller volume of the compact chambers, or the avoidance of burrows that allow space for potential predators. They also selected for burrows l ocated within areas containing the highest percentage of canopy cover likely to provide suitable cover to better avoid predators and ambush potential prey. Our results suggest that gopher tortoise burrows can conceivably be used as overwintering refugia b y Burmese pythons north of their current range in Florida. Although internal burrow temperature showed latitudinal variation thirty of thirty one monitored burrows maintained temperatures above those described as lethal for Burmese pythons, even when exte rnal temperatures dropped well below the lethal limit (~5 o C, Dorcas et al., 2011). Although Burmese pythons are widely thought to be thermally restricted in Florida, t he importance of burrows and other subterranean ref ugia is often overlooked. Burrows pro vide protection against temperature extremes, fire, and predation ( Campbell & Clark, 1981; Alkon & S altz, 1988 ; Reichman & Smith, 1990; Mukherjee et al., 2017) and could potentially facilitate a range expansion by pythons substantially farther north than their current distribution. To maximize the success of management efforts, it is important to understand the ecology of Burmese pythons within all habitats they occupy including lowland habitats that comprise most of the Florida Everglades, and upland hab itats that comprise much of southwestern Florida T he data we have gathered for this study can

PAGE 90

90 be used to identify and predict areas within upland habitats that are most suitable for invasive pythons, thus improving python management efforts within Florida Our results suggest that r educing over / understory cover through management techniques such as prescribed fire may potentially make habitats less suitable and discourage pythons from establishing Additionally, as the detection probability of pythons wi thin burrows is extremely high, burrow surveys may be an effective tool for managing and monitoring python populations, especially along the range margins. I mplementing targeted searches during the python nesting season could help manage adult female pytho ns and python nests, potentially impacting Burmese pythons at a population level.

PAGE 91

91 LIST OF REFERENCES Akaike H. 2011. Encyclopedia of Statistical Science. Springer, Berlin, H eidelberg Alkon, P. U., & Saltz, D. 1988. Foraging time and the northern range limits of Indian crested porcupines ( Hystrix indica Kerr ). Journal of biogeography 403 408. Andreadis, P. 2011. Python molurus bivittatus (Burmese python). Reproducing populati on. Herpetological Review 42:302 303. Andreadis, P.T., Bartoszek, I.A., Prokop Ervin, C., & Pittman, S. 2018. Drymarchon kolpobasileus (Gulf Coast Indigo Snake) and Python bivittatus (Burmese Python). Predator Prey Interaction. Herpetological Review 49(2): 341 342. Auffenberg, W. 1969. Tortoise behavior and survival. Chicago: Rand McNally. Auffenberg, W., & Franz, R. 1982. The status and distribution of the gopher tortoise ( Gopherus polyphemus ). North American tortoises: conservation and ecology 95 126. Aver y, M. L., Engeman, R. M., Keacher, K. L., Humphrey, J. S., Bruce, W. E., Mathies, T. C., & Mauldin, R. E. 2010. Cold weather and the potential range of invasive Burmese pythons. Biological Invasions 12(11):3649 3652. Ball, L. C., Doherty Jr, P. F., & McDon ald, M. W. 2005. An occupancy modeling approach to evaluating a Palm Springs ground squirrel habitat model. The Journal of wildlife management 69(3):894 904. Barker, D. G., & Barker, T. M. 2008. The distribution of the Burmese python, Python molurus bivitt atus Bulletin of the Chicago Herpetological Society 43(3):33 38. Bartoszek, I.A., Andreadis, P.T., Prokop Ervin, C., Curry, G., & Reed, R.N. 2018. Python bivittatus (Burmese Python) and Gopherus polyphemus (Gopher Tortoise). Habitat Use, Breeding Aggregat ion, and Interspecific Interaction. Herpetological Review 49(2):353 354. Bhupathy, S. & Vijayan, V. S. 1989. Status, distribution and general ecology of the Indian python Python molurus molurus Linn. in Keoladeo National Park, Bharatpur, Rajasthan. Journal of the Bombay Natural History Society 86(3):381 387. Birkhead, R.D., Guyer, C., Hermann, S.M., & Michener, W.K. 2005. Patterns of Folivory and Seed Ingestion by Gopher Tortoises ( Gopherus polyphemus ) in a Southeastern Pine Savanna. American Midland Natura list, 143 151.

PAGE 92

92 Bradshaw, C. J., Boutin, S., Hebert, D. M., & Rippin, A. B. 1995. Winter peatland habitat selection by woodland caribou in northeastern Alberta. Canadian Journal of Zoology 73(8):1567 1574. Brode, W. E. 1959. Notes on behavior of Gopherus po lyphemus Herpetologica 15(2):101 102. Brown, J. H., St evens, G. C., & Kaufman, D. M. 1996 The geographic range: size, shape, boundaries, and internal structure. Annual review of ecology and systematics 27(1):597 623. Burnham, K. P., Ande rson, D.R., & Laak e, J.L. 1980 Estimation of density from line transect sampling of biological populations. Wildlife monographs, (72), 3 202. Butler, J. A. & Sowell, S. 1996. Survivorship and predation of hatchling and yearling gopher tortoises, Gopherus polyphemus Journa l of Herpetology 30(3):455 458. Campbell, H. W. & Christman, S. P. 1982. The herpetological components of Florida sandhill and sand pine scrub associations. Herpetological Communities 13:163 71. Campbell III, T. M. & Clark, T. W. 1981. Colony characteristi cs and vertebrate associates of white tailed and black tailed prairie dogs in Wyoming. American Midland Naturalist 269 276. Carr, A. F. 1952. Handbook of Turtles: The Turtles of the United States, Canada, and Baja California. Comstock Pub. Associates. Chri stman, S. P. 1988. Endemism in Florida's interior sand pine scrub. Florida Game and Fresh Water Fish Commission, Nongame Wildlife Program, Final Report, Tallahassee, Florida, USA. Clavero, M. & Garca Berthou, E. 2005. Invasive species are a leading cause of animal extinctions. Trends in ecology & evolution 20(3):110. Clout, M. N. & Williams, P. A. 2009. Invasive species management: a handbook of principles and techniques. Oxford University Press. Cox, G.W. 2004. Alien species and evolution: the evolutionar y ecology of exotic plants, animals, microbes, and interacting native species. Island Press. Cruz, F., Carrion, V., Campbell, K. J., Lavoie, C., & Donlan, C. J. 2009. Bio economics of large scale eradication of feral goats from Santiago Island, Galapagos. The Journal of Wildlife Management 73(2):191 200. Darling, E., Samis, K. E., & Eckert, C. G. 2008. Increased seed dispersal potential towards geographic range limits in a Pacific coast dune plant. New Phytologist 178(2):424 435.

PAGE 93

93 Davis, M. B. & Shaw, R. G. 2001. Range shifts and adaptive responses to Quaternary climate change. Science 292(5517):673 679. DeGregorio, B. A., Buhlmann, K. A., & Tuberville, T. D. 2012. Overwintering of gopher tortoises ( Gopherus polyphemus ) translocated to the northern limit of t heir geographic range: temperatures, timing, and survival. Chelonian conservation and biology 11(1):84 90. Diemer, J. E. 1986. The ecology and management of the gopher tortoise in the southeastern United States. Herpetologica 125 133. Diemer, J. E. & Speak e, D. W. 1981. The status of the eastern indigo snake in Georgia. In Proceedings Nongame and Endangered Wildlife Symposium, Georgia Department of Natural Resources Technical Bulletin WL 5 (Vol. 179, pp. 52 61). Dorcas, M. E. & Willson, J. D. 2013. Hidden g iants: problems associated with studying secretive invasive pythons. Reptiles in research: investigations of ecology, physiology, and behavior from desert to sea. Hauppage: Nova Science Publishers, Inc. Dorcas, M. E., Willson, J. D., & Gibbons, J. W. 2011. Can invasive Burmese pythons inhabit temperate regions of the southeastern United States? Biological Invasions 13(4):793 802. Dorcas, M. E., Willson, J. D., Reed, R. N., Snow, R. W., Rochford, M. R., Miller, M. A., ... & Hart, K. M. 2012. Severe mammal de clines coincide with proliferation of invasive Burmese pythons in Everglades National Park. Proceedings of the National Academy of Sciences 109(7):2418 2422. Douglass, J. F. & Winegarner, C. E. 1977. Predators of eggs and young of the gopher tortoise, Goph erus polyphemus (Reptilia, Testudines, Testudinidae) in southern Florida. Journal of Herpetology 11(2):236 238. Dove, C. J., Snow, R. W., Rochford, M. R., & Mazzotti, F. J. 2011. Birds consumed by the invasive Burmese python ( Python molurus bivittatus ) in Everglades National Park, Florida, USA. The Wilson Journal of Ornithology 123(1):126 131. Eisenberg, J. F. 1983. The gopher tortoise as a keystone species. In Proceedings of the Annual Meeting of the Gopher Tortoise Council (Vol. 4, pp. 1 4). Enge, K. M., Jernigan, A., Johnson, R., Berish, J. E. D., Moler, P. E., Thomas, T. M., & Kiltie, R. A. 2012. What the world needs is a better Gopher Tortoise trap. Herpetological Review 43(4):574 578. Engeman, R., Jacobson, E., Avery, M. L., & Meshaka Jr, W. E. 2011. T he aggressive invasion of exotic reptiles in Florida with a focus on prominent species: A review. Current zoology 57(5):599 612.

PAGE 94

94 Ernst, C. H. & Barbour, R. W. 1989. Turtles of the world. Washington. Smithsonian Institute. Press. Ernst, C. H., Lovich, J. E. & Barbour, R. W. 1994. Turtles of the United States and Canada. Washington, DC: Smithsonian Institute Press. Ernst, C.H. & Zug G.R 1996. Snakes in question. Smithsonian Institution Press, Washington, D.C, USA. Fernald, R. T. 1989. Coastal xeric scrub c ommunities of the Treasure Coast Region, Florida: a summary of their distribution and ecology, with guidelines for their preservation and management. Office of Environmental Services, Florida Game and Fresh Water Fish Commission. Gaston, K. J. 2003. The st ructure and dynamics of geographic ranges. Oxford University Press on Demand. Gillies, C. S., Hebblewhite, M., Nielsen, S. E., Krawchuk, M. A., Aldridge, C. L., Frair, J. L., Saher, J., Stevens, C.E., & Jerde, C. L. 2006. Application of random effects to t he study of resource selection by animals. Journal of Animal Ecology 75(4):887 898. Groombridge, B. & Luxmoore, R. A. 1991. Pythons in South East Asia: A Review of Distribution, Status, and Trade in Three Selected Species: a Report to the CITES Secretariat August 1990. Secretariat of the Convention on International Trade in Endangered Species of Wild Fauna and Flora. Gurevitch, J. & Padilla, D. K. 2004. Are invasive species a major cause of extinctions? Trends in ecology & evolution 19(9):470 474. Hallinan T. 1923. Observations made in Duval County, northern Florida, on the gopher tortoise ( Gopherus polyphemus ). Copeia 1923:11 20. Hart, K.M., Cherkiss, M.S., Smith, B.J., Mazzotti, F.J., Fujisaki, I., Snow, R.W., & Dorcas, M.E. 2015 Home range, habitat use and movement patterns of non native Burmese pythons in Everglades National Park, Florida, USA. Animal Biotelemetry, 3(1), 8. Hanley, J.A., & McNeil, B.J. 1 982 The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiolo gy, 143(1), 29 36. Hansen, K.L. 1963. The burrow of the gopher tortoise. Quarterly Journal of the Florida Academy of Sciences 26(4):353 360. dynamic perspective. Case studi es in mathematical modelling: ecology, physiology, and cell biology 25 50.

PAGE 95

95 Hubbard, H. G. 1893. The Florida land tortoise gopher, Gopherus polyphemus Science 548: 57 58. Hubbard, H. G. 1894. The insect guests of the Florida land tortoise. US Government Pr inting Office. Huey, R.B., Gilchrist, G.W., Carlson, M.L., Berrigan, D., & Serra, L. 2000. Rapid 309. Hutt, A. 1967. The gopher tortoise, a versatile vegetarian. Florida Wildl. 21 :20 24. Jackson, D. R. & Milstrey, E. G. 1989. The fauna of gopher tortoise burrows. In Gopher tortoise relocation symposium proceedings (pp. 86 98). Florida Game and Fresh Water Fish Commission Nongame Wildlife Program Tech. Rep. 5. Johnson, D. H. 1980. T he comparison of usage and availability measurements for evaluating resource preference. Ecology 61(1):65 71. Kalisz, P. J. & Stone, E. L. 1984. The longleaf pine islands of the Ocala National Forest, Florida: a soil study. Ecology 65(6):1743 1755. Kraus, F. 2009. Alien reptiles and amphibians: a scientific compendium and analysis. Springer, New York. Krebs, C.J. 1978. Ecology: the experimental analysis of distribution and abundance, 2nd edn. Harper & Row, United States. Krysko, K.L., Somma, L.A., Smith, D.C., Gillette, C.R., Cueva, D., Wasilew ski, J.A., Enge, K.M., Johnson, S.A., Campbell, T.S., Edwards, J.R., Rochford, M.R., Tompkins, R., Fobb, J.L., Mullin, S., Lechowicz, C.J., Hazelton, D., & Warren, A. 2016. New verified nonindigenous amphibians and reptiles in Florida through 2015, with a summary of over 152 years of introductions. Reptiles Amphib, 23, 110 143. Landers, J. L. 1980. Recent research on the gopher tortoise and its implications. Pp. 8 14. In R. Franz and R. J. Bryant (Eds.), The Dilemma of the Gopher Tortoise Is There a Solutio n? Proc. 1st Ann. Mtg., Gopher Tortoise Council. Landers, J. L. & J. A. Garner. 1981. Status and distribution of the gopher tortoise in Georgia. Pp. 45 51. In R. Odum and J. Guthrie (Eds.), Proc. Non Game End. Wildl. Symp., Georgia Dept. Nat. Resour., Game Fish Div. Tech. Bull. WL5. Landers, J. L. & D. W. Speake. 1980. Management needs of sandhill reptiles in southern Georgia. Proc. Ann. Conf. S.E. Assoc. Fish Wildl. Agen. 34:515 529. Lee, C.E. 1999. Rapid and repeated invasions of fresh water by the saltwa ter copepod Eurytemora affinis Evolution 53:1423 1434.

PAGE 96

96 Lee, C.E. 2002. Evolutionary genetics of invasive species. Trends Ecol. Evol. 17:386 91. Linkie, M., Dinata, Y., Nugroho, A., & Haidir, I. A. 2007. Estimating occupancy of a data deficient mammalian s pecies living in tropical rainforests: sun bears in the Kerinci Seblat region, Sumatra. Biological Conservation 137(1):20 27. Lips, K. R. 1991. Vertebrates associated with tortoise ( Gopherus polyphemus ) burrows in four habitats in south central Florida. Jo urnal of Herpetology 25(4):477 481. Lohoefener, R. & L. Lohmeier. 1984. The status of Gopherus polyphemus (Testudines, Testudinidae) west of the Tombigbee and Mobile Rivers. Rept. to U.S. Fish Wildl. Serv. in support of petition to list this population und er the Endangered Species Act of 1973:1 116. Lowe, S., Browne, M., Boudjelas, S., & De Poorter, M. 2000. 100 of the world's worst invasive alien species: a selection from the global invasive species database (Vol. 12). Auckland: Invasive Species Specialist Group. MacDonald, L.A. & Mushinsky, H.R. 1988. Foraging ecology of the gopher tortoise, Gopherus polyphemus in a sandhill habitat. Herpetologica 345 353. MacKenzie, D.I., Nichols, J.D., Lachman, G.B., Droege, S., Andrew Royle, J., & Langtimm, C.A. 2002. Estimating site occupancy rates when detection probabilities are less than one. Ecology 83(8):2248 2255. McCann, B.E. & Garcelon, D.K. 2009. Eradication of feral pigs from Pinnacles National Monument. Journal of Wildlife Management 72:1287 1295. McCleery, R.A., Sovie, A., Reed, R.N., Cunningham, M.W., Hunter, M.E., & Hart, K.M. 2015. Marsh rabbit mortalities tie pythons to the precipitous decline of mammals in the Everglades. Proc. R. Soc. B. 282(1805):20150120. Meshaka Jr, W. E. 2011. A runaway train in th e making: The exotic amphibians, reptiles, turtles, and crocodilians of Florida. Herpetological Conservation and Biology, 6 (Monograph 1). Meshaka Jr, W.E., Butterfield, B.P., & Hauge, J.B. 2004. The exotic amphibians and reptiles of Florida. Krieger Co., Malabar, Florida. Meshaka Jr, W.E., Loftus, W.F., & Steiner, T. 2000. The herpetofauna of Everglades National Park. Fla Sci 63:84 103. Metzger, C.j. 2013. Python molurus bivittatus (Burmese Python). Habitat Use / Occurrence Within Gopherus polyphemus Burro ws. Herpetological Review 44:333 334.

PAGE 97

97 Miller, M. A., Kinsella, J. M., Snow, R. W., Hayes, M. M., Falk, B. G., Reed, R. N., Mazzotti, F.J., Guyer, C., & Romagosa, C. M. 2018. Parasite spillover: indirect effects of invasive Burmese pythons. Ecology and evol ution 8(2):830 840. Minton Jr, S.A. 1966. A contribution to the herpetology of West Pakistan. Bull Am Mus Nat Hist 134:29 184. pp. 198 227. Montgomery, G. G. & Rand, A. S. 1978. Movements, body temperature and hunting strategy of a Boa constrictor Copeia 1978(3):532 533. Mukherjee, A., Pilakandy, R., Kumara, H. N., Manchi, S. S., & Bhupathy, S. 2017. Burrow characteristics and its importance in occupancy of burrow dwelli ng vertebrates in Semiarid area of Keoladeo National Park, Rajasthan, India. Journal of arid environments 141:7 15. Myers, R. L. 1990. Scrub and high pine. Ecosystems of Florida 150 193. Peirce, M. A. 1974. Distribution and ecology of Ornithodoros moubata porcinus Walton (Acarina) in animal burrows in East Africa. Bulletin of Entomological Research 64(4):605 619. Phillips, B.L., Brown, G.P., Webb, J.K., & Shine, R. 2006. Invasion and the evolution of speed in toads. Nature 439:803. Piaggio, A. J., Engeman, R. M., Hopken, M. W., Humphrey, J. S., Keacher, K. L., Bruce, W. E., & Avery, M. L. 2014. Detecting an elusive invasive species: a diagnostic PCR to detect Burmese python in Florida waters and an assessment of persistence of environmental DNA. Molecular e cology resources 14(2):374 380. Pimentel, D., Zuniga, R., & Morrison, D. 2005. Update on the environmental and economic costs associated with alien invasive species in the United States. Ecological economics 52(3):273 288. Potvin, C. & J.P. Simon. 1989. The evolution of cold temperature adaptation among populations of a widely distributed C4 weed: barnyard grass. Evol. Trends Plants 3:98 105. Pulliam, H. R. & Danielson, B. J. 1991. Sources, sinks, and habitat selection: a la ndscape perspective on population dynamics. The American Naturalist 137:S50 S66. Pyron, R. A., Burbrink, F. T., & Guiher, T. J. 2008. Claims of potential expansion throughout the US by invasive python species are contradicted by ecological niche models. Pl oS one 3(8), e2931.

PAGE 98

98 Reed, R.N., Krysko, K.L., Snow, R.W., & Rodda, G.H. 2010. Is the Northern African python ( Python sebae ) established in southern Florida? IRCF Reptile Amphib 17(1):52 54. Reed, R.N. & Rodda, G.H. 2009. Giant constrictors: biological and management profiles and an establishment risk assessment for nine large species of pythons, anacondas, and the boa constrictor (No. 2009 1202). US Geological Survey. Reichman, O. J. & Smith, S. C. 1990. Burrows and burrowing behavior by mammals. Current Ma mmalogy 2:197 244. Reznick, D.N. & Ghalambor, C.K. 2001. The population ecology of contemporary adaptations: what empirical studies reveal about the conditions that promote adaptive evolution. In Microevolution Rate, Pattern, Process (pp. 183 198). Springe r, Dordrecht. Rivas, J.A. 1999. The Life History of the Green Anaconda (Eunectes murinus). Rodda, G.H., Jarnevich, C.S., & Reed, R.N. 2009. What parts of the U.S. mainland are climatically suitable for invasive alien pythons spreading from Everglades Natio nal Park? Biological Invasions 11:241 252. Rodda, G.H., Jarnevich, C.S., & Reed, R.N. 2011. Challenges in identifying sites climatically matched to the native ranges of animal invaders. PloS one 6(2):e14670. Sexton, J.P., McIntyre, P.J., Angert, A.L., & Ri ce, K.J. 2009. Evolution and ecology of species range limits. Annual Review of Ecology, Evolution, and Systematics 40. Sharma, B. & Kandel, R. 2016. Assessment of python ( Python bivittatus Kuhl) habitats in Bardiya National Park, Nepal. Ecoprint: An Intern ational Journal of Ecology 22:85 90. Smith, B.J., Cherkiss, M.S., Hart, K.M., Rochford, M.R., Selby, T.H., Snow, R.W., & Mazzotti, F.J. 2016. Betrayal: radio tagged Burmese pythons reveal locations of conspecifics in Everglades National Park. Biological i nvasions 18(11):3239 3250. Smith, L.L., Linehan, J.M., Stober, J.M., Elliott, M.J., & Jensen, J.B. 2009. An evaluation of distance sampling for large scale gopher tortoise surveys in Georgia, USA. Applied Herpetology 6(4):355 368. Snow, R.W. 2006. Disposab le pets, unwanted giants: pythons in Everglades National Park. Proceedings from Florida Exotic Plant Council, 21st Annual Symposium, Gainesville, USA. Snow, R.W., Brien, M.L., & Cherkiss, M.S. 2007. Dietary habits of Burmese Python, Python molurus bivittat us from Everglades National Park, Florida. Herpetological Bulletin 101:5 7.

PAGE 99

99 Snow, R.W., Johnson, V.M., Brien, M.L., Cherkiss, M.S., & Mazzotti, F.J. 2007. Python molurus bivitatus (Burmese python) nesting. Herpetological Review 38:93. Snow, R.W., Krysko, K.L., Enge, K.M., Oberhofer, L., Walker Bradley, A., & Wilkins, L. 2007. Introduced populations of Boa constrictor (Boidae) and Python molurus bivittatus (Pythonidae) in southern Florida. In: The Biology of Boas and Pythons (eds Henderson RW and Powell R), pp. 416 438. Eagle Mountain Publishing, Eagle Mountain, Utah. Storey, K. B. 1990. Life in a frozen state: adaptive strategies for natural freeze tolerance in amphibians and reptiles. American Journal of Physiology Regulatory, Integrative and Comparative Physiology 258(3):R559 R568. Sugiura, N. 1978 Further analysts of the data by akaike's information criterion and the finite corrections: Further analysts of the data by akaike's. Communications in Statistics Theory and Methods, 7(1), 13 26. Taylor, D. & K atahira, L. 1988. Radio telemetry as an aid in eradicating remnant feral goats. Wildlife Society Bulletin (1973 2006) 16(3):297 299. Thomas, C.D., Bodsworth, E.J., Wilson, R.J., Simmons, A.D., & Davies, Z.G. 2001. Ecological and evolutionary processes at e xpanding range margins. Nature 411:577 581. Travis, J.M. & Dytham, C. 2002. Dispersal evolution during invasions. Evolutionary Ecology Research 4(8):1119 1129. Ultsch, G.R. 2006. The ecology of overwintering among turtles: where turtles overwinter and its consequences. Biological Review 1:1 29. Vitousek, P.M., D'antonio, C.M., Loope, L.L., Rejmanek, M., & Westbrooks, R. 1997. Introduced species: a significant component of human caused global change. New Zealand Journal of Ecology 1 16. Walters, T.M., Mazzot ti, F.J., & Fitz, H.C. 2016. Habitat selection by the invasive species Burmese python in Southern Florida. Journal of Herpetology 50(1):50 56. Whitaker, R., Captain, A., & Ahmed, F. 2004. Snakes of India. Draco Books. Williams, C.K. & Moore, R.J. 1989. Phe notypic adaptation and natural selection in the wild rabbit, Oryctolagus cuniculus, in Australia. The Journal of Animal Ecology 495 507. Woodruff, R. E. 1982. Arthropods of gopher burrows. In Gopher tortoise and its sandhill habitat. Proceedings of the thi rd annual meeting of the Gopher Tortoise Council, Tall Timbers Research Station, Tallahassee, FL (pp. 24 48).

PAGE 100

100 Woolnough, A.P., Lowe, T.J., & Rose, K. 2006. Can the Judas technique be applied to pest birds? Wildlife Research 33(6):449 455. Wright, J.S. 1982 Distribution and population biology of the gopher tortoise, Gopherus polyphemus South Carolina. Unpubl. MS thesis, Clemson University, Clemson, South Carolina. Young, F.N. & Goff, C.C. 1939. An annotated list of the arthropods found in the burrows of th e Florida gopher tortoise, Gopherus polyphemus (Daudin). The Florida Entomologist 22(4):53 62. Zeller, K.A., Nijhawan, S., Salom Prez, R., Potosme, S.H., & Hines, J.E. 2011. Integrating occupancy modeling and interview data for corridor identification: a case study for jaguars in Nicaragua. Biological Conservation 144(2):892 901. Zhao, E. M. & Adler, K. 1993. Herpetology of China. Contribution to Herpetology. No. 10. Society for the study of amphibians and reptiles, Athens, Ohio.

PAGE 101

101 BIOGRAPHICAL SKETCH Kodiak graduated from Northern Michigan University in 2014 with a Bachelor of Science in zoology. While at the University of Florida, Kodiak majored in wildlife ecology and conservation, earning his Master of Science in 2018.