<%BANNER%>

Distribution and Life History of Two Non-Indigenous Geckos (Hemidactylus garnotii and H. mabouia) in Southwest Florida

Permanent Link: http://ufdc.ufl.edu/UFE0021274/00001

Material Information

Title: Distribution and Life History of Two Non-Indigenous Geckos (Hemidactylus garnotii and H. mabouia) in Southwest Florida
Physical Description: 1 online resource (220 p.)
Language: english
Creator: Klowden, Gregg S
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: activity, annlaris, charlotte, county, demography, distribution, florida, fort, frenatus, garnotii, gecko, gekko, growth, hemidactylus, introduced, invasive, lee, lizard, longevity, mabouia, mark, maturity, myers, port, recapture, reproduction, reptile, sarasota, survival, tarentola, turcicus
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This research was directed toward discerning whether or not changes in the distribution and abundance of the non-indigenous gecko Hemidactylus garnotii in Charlotte and Lee Counties, Florida could be linked to the co-occurrence of the more recently introduced H. mabouia and if so, what role life history differences might play. I presented data on their distribution, co-occurrence, colonization and extinction rates, demographics, and other life history traits. Surveys indicated that, from 2003 to 2006 in Charlotte County and 2004 to 2006 in Lee County, the proportion of buildings on which H. garnotii occurred decreased whereas for H. mabouia it increased. The extinction rate for H. garnotii was high and the colonization rate low while for H. mabouia the extinction rate was low and colonization rate high. In Lee County, while both species were distributed randomly in 2004, by 2006 they were less likely to co-occur than would be predicted by chance. Though colonization by H. mabouia did not appear to be the primary cause of extinction of H. garnotii from a building, on buildings on which they co-occured, the extinction rate of H. garnotii was significantly higher than for H. mabouia. Further, the ability of H. garnotii to colonize buildings occupied by H. mabouia was significantly lower than for buildings that were unoccupied. I estimated life history traits including stage specific survival, fecundity, mean snout-vent length, weight, maturation size, growth rate, time to maturation, inter-clutch interval, mean and maximum longevity, and other parameters using mark-recapture techniques. In Charlotte County I captured 914 H. garnotii 2,200 times and in Lee County 546 H. mabouia 1,030 times. H. mabouia was nearly 13% longer and more than 90% heavier than H. garnotii. Both species were similar in that they matured rapidly and had high growth rates. Hemidactylus garnotii fecundity was twice that of H. mabouia. Adult survival was similar, however juvenile survival of H. mabouia was double that of H. garnotii. Data on the occurrence of other non-indigenous gecko species and the northward dispersal of H. mabouia into Sarasota County are also presented.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Gregg S Klowden.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Moulton, Michael P.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021274:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021274/00001

Material Information

Title: Distribution and Life History of Two Non-Indigenous Geckos (Hemidactylus garnotii and H. mabouia) in Southwest Florida
Physical Description: 1 online resource (220 p.)
Language: english
Creator: Klowden, Gregg S
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: activity, annlaris, charlotte, county, demography, distribution, florida, fort, frenatus, garnotii, gecko, gekko, growth, hemidactylus, introduced, invasive, lee, lizard, longevity, mabouia, mark, maturity, myers, port, recapture, reproduction, reptile, sarasota, survival, tarentola, turcicus
Wildlife Ecology and Conservation -- Dissertations, Academic -- UF
Genre: Wildlife Ecology and Conservation thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This research was directed toward discerning whether or not changes in the distribution and abundance of the non-indigenous gecko Hemidactylus garnotii in Charlotte and Lee Counties, Florida could be linked to the co-occurrence of the more recently introduced H. mabouia and if so, what role life history differences might play. I presented data on their distribution, co-occurrence, colonization and extinction rates, demographics, and other life history traits. Surveys indicated that, from 2003 to 2006 in Charlotte County and 2004 to 2006 in Lee County, the proportion of buildings on which H. garnotii occurred decreased whereas for H. mabouia it increased. The extinction rate for H. garnotii was high and the colonization rate low while for H. mabouia the extinction rate was low and colonization rate high. In Lee County, while both species were distributed randomly in 2004, by 2006 they were less likely to co-occur than would be predicted by chance. Though colonization by H. mabouia did not appear to be the primary cause of extinction of H. garnotii from a building, on buildings on which they co-occured, the extinction rate of H. garnotii was significantly higher than for H. mabouia. Further, the ability of H. garnotii to colonize buildings occupied by H. mabouia was significantly lower than for buildings that were unoccupied. I estimated life history traits including stage specific survival, fecundity, mean snout-vent length, weight, maturation size, growth rate, time to maturation, inter-clutch interval, mean and maximum longevity, and other parameters using mark-recapture techniques. In Charlotte County I captured 914 H. garnotii 2,200 times and in Lee County 546 H. mabouia 1,030 times. H. mabouia was nearly 13% longer and more than 90% heavier than H. garnotii. Both species were similar in that they matured rapidly and had high growth rates. Hemidactylus garnotii fecundity was twice that of H. mabouia. Adult survival was similar, however juvenile survival of H. mabouia was double that of H. garnotii. Data on the occurrence of other non-indigenous gecko species and the northward dispersal of H. mabouia into Sarasota County are also presented.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Gregg S Klowden.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Moulton, Michael P.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021274:00001


This item has the following downloads:


Full Text

PAGE 1

DISTRIBUTION AND LIFE HISTORY OF TWO NON-INDIGENOUS GECKOS Hemidactylus garnotii AND H. mabouia IN SOUTHWEST FLORIDA By GREGG STUART KLOWDEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007 1

PAGE 2

To Cathy Olson whose love, friendship, and patience made this possible. 2

PAGE 3

ACKNOWLEDGMENTS I would first like to thank the University of Florida Alumni Association, the University of Florida, the Institute of Food and Agricultural Sciences, and the Department of Wildlife Ecology for providing me with a Fellowship, tuition waiver, and Graduate Teaching Assistantship without which my Ph.D. would not have been possible. I would like to thank my committee, Drs. Michael Moulton, Alison Fox, Mark Hostetler, Max Nickerson, and Madan Oli, for their help, generosity, and patience. I am thankful to Dr. Moulton for offering me the opportunity to participate in the organization and teaching of his courses, guiding me through the many intricacies of running a large lecture course, and helping me to become a better teacher. In addition, as the chair of my supervisory committee, Dr. Moultons guidance over the years has been critical to my growth as a scientist, ecologist, and writer. His philosophical perspectives on ecology, wildlife conservation, and life are thoughtful, unique, and unmatched as are his humor, good nature, and kindness. His mentoring and friendship will undoubtedly be very influential as I continue to shape my career. I am very thankful to Dr. Hostetler for employing me as his Wildlife Extension Assistant. This position fostered my interest in public outreach and provided me with valuable skills and considerable knowledge that will allow me to continue teaching and encouraging the involvement of the public in conservation issues. Many thanks to Dr. Fox for launching my interest in non-indigenous species through her seminar course and for her very thorough editing, to Dr. Oli for his excellent modeling courses and encouragement, and to Dr. Nickerson for his kindness and being the one person on my committee who understands why someone would scramble around in the middle of the night in pursuit of a lizard. Many thanks also go out to the numerous other people who have helped me during my pursuit of this degree. For financial support I thank the Charlotte Harbor National Estuary Program. For designing the models, writing the software, writing the many brilliant papers, and answering my pestiferous emails I thank Drs. Jim Nichols, Jim Hines, Gary White, Darryl MacKenzie, Roger Pradel, and Rmi Choquet. I also thank Arpat Ozgul and Julien Martin for their consultations. For 3

PAGE 4

expediently fulfilling my specimen data requests I thank the numerous museum curators and collection managers. Thanks to Meghan Brennan, IFAS Statistics, for her assistance. For not arresting me for creeping around behind buildings in the middle of the night with a flashlight, I thank the Charlotte and Lee County Police Department. For employing me as his field biologist for the Nile monitor lizard project, his willingness to share his biological knowledge and perspectives on involving the public in conservation, and his friendship I thank Dr. Todd Campbell. For editing I thank Cathy Olson. For providing office space and encouragement, I thank Drs. Eleanor and Bill Marr. I owe a warm thank you to the entire Department of Wildlife Ecology and Conservation faculty for help along the way and for teaching me to be a better scientist. I want to thank the Wildlife Ecology and Conservation staff, past and present, for the many things they have done that made the completion of my Masters and Ph.D. possible and for their kindness and good humor. Many thanks are owed to Mom and Dad, other family, and friends for their love, encouragement, support, words of wisdom (especially Steve), good humor, and essential musical interludes. Huge thanks for keeping me sane, getting me out of the house to play, and making me laugh are owed to Denny, Darby, Grover, Ury, Gus, Bernie, Zydeco, B2G2, Pedro, and Lola. Lastly I would like to thank the many people I have known and camp counselors and teachers I have had who have fostered my love of animals, the outdoors, and music, and my desire to share my passion and knowledge with others. I am especially thankful to my best friend and wife Cathy Olson. Her willingness to share her knowledge and enthusiasm for natural history and ecology has been essential to my growth as an ecologist. Her kindness, selflessness, and uncompromising morals have been essential to my growth as a person. Her love has been essential to my life. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.........................................................................................................3 LIST OF TABLES.....................................................................................................................8 LIST OF FIGURES...................................................................................................................11 ABSTRACT...............................................................................................................................14 CHAPTER 1 INTRODUCTION..............................................................................................................16 2 BACKGROUND................................................................................................................19 Distribution.........................................................................................................................19 Natural History...................................................................................................................21 Size..................................................................................................................................22 Reproduction...................................................................................................................23 Activity............................................................................................................................25 Predators..........................................................................................................................26 Diet..................................................................................................................................27 Dispersal..........................................................................................................................28 County Descriptions...........................................................................................................28 3 DISTRIBUTION AND RELATIVE ABUNDANCE OF Hemidactylus garnotii AND H. mabouia IN TWO SOUTHWEST FLORIDA COUNTIES..................38 Introduction.........................................................................................................................38 Survey Methods..................................................................................................................39 Presurvey Methods..........................................................................................................40 Random generation of potential survey points.............................................................40 Selection of survey buildings.......................................................................................41 Data Collection................................................................................................................42 Data Analysis......................................................................................................................44 Distribution......................................................................................................................44 Abundance.......................................................................................................................44 Within a single season..................................................................................................44 Abundance changes over time......................................................................................48 Species Interactions.........................................................................................................49 Number of Geckos Per Building.....................................................................................51 Results.................................................................................................................................51 Charlotte County.............................................................................................................51 Single season occupancy estimates..............................................................................52 Colonization and extinction..........................................................................................52 Species co-occurrence..................................................................................................55 5

PAGE 6

Number of geckos per building....................................................................................56 Lee County......................................................................................................................57 Single year occupancy estimates..................................................................................58 Colonization and extinction..........................................................................................60 Detection probability in allopatry versus in sympatry................................................61 Co-occurrence patterns................................................................................................62 Number of geckos per building...................................................................................62 Discussion...........................................................................................................................64 Distributions....................................................................................................................64 Colonization and Extinction............................................................................................65 4 LIFE HISTORY OF Hemidactylus garnotii AND H. mabouia IN SOUTHWEST FLORIDA..................................................................................................91 Introduction.........................................................................................................................91 Timeline..............................................................................................................................91 Study Site Descriptions.......................................................................................................92 Methods..............................................................................................................................93 Data Collection................................................................................................................93 Data Analysis..................................................................................................................95 Survival........................................................................................................................95 Growth..........................................................................................................................98 Reproduction................................................................................................................99 Time until maturity..................................................................................................99 Maximum number of clutches per year...................................................................100 Breeding proportion.................................................................................................100 Mean fecundity and fertility.....................................................................................101 Maximum longevity.................................................................................................101 Abundance....................................................................................................................101 Results.................................................................................................................................102 General Summary............................................................................................................102 Size at Maturity...............................................................................................................103 Sex Ratio.........................................................................................................................103 Size Distribution..............................................................................................................103 Captures per Animal........................................................................................................105 Seasonal Effect of Capture Rate......................................................................................107 Survival probability estimation.......................................................................................108 Stages............................................................................................................................108 Model selection............................................................................................................108 Capture probability.......................................................................................................110 Transition probability...................................................................................................110 Survival probability......................................................................................................111 Growth.............................................................................................................................111 Time Until Maturity........................................................................................................112 Reproduction...................................................................................................................113 Reproductive season...................................................................................................113 Reproductive frequency..............................................................................................114 Maximum number of clutches per year......................................................................114 6

PAGE 7

Breeding proportion....................................................................................................115 Fecundity and fertility.................................................................................................115 Reproduction and Size: H. garnotii............................................................................116 Reproduction and Size: H. mabouia...........................................................................117 Maximum Longevity.......................................................................................................117 Tail Loss..........................................................................................................................118 Movement........................................................................................................................118 Abundance.......................................................................................................................119 Density.............................................................................................................................121 Discussion...........................................................................................................................121 5 ADDITIONAL DISTRIBUTIONAL OBSERVATIONS.................................................164 Hemidactylus turcicus........................................................................................................164 Hemidactylus frenatus........................................................................................................164 Gekko gecko.......................................................................................................................165 Tarentola annularis............................................................................................................165 Hemidactylus mabouia.......................................................................................................166 6 SUMMARY........................................................................................................................171 APPENDIX A DETAILED METHODOLOGY FOR GENERATION OF RANDOM POINTS.............175 B SURVEY BUILDING STATISTICS AND RAW DATA.................................................180 C GECKOS CAPTURED IN SARASOTA COUNTY.........................................................202 LITERATURE CITED..............................................................................................................207 BIOGRAPHICAL SKETCH.....................................................................................................220 7

PAGE 8

LIST OF TABLES Table page 2-1 Museums queried for Florida Hemidactylus specimens................................................31 3-1 Charlotte County: Number and proportion of buildings on which I observed Hemidactylus garnotii or H. mabouia each survey year and the inter-survey change............................................................................................................................69 3-2 Charlotte County: Number of buildings colonized by H. garnotii and H. mabouia from 2003 to 2004, 2004 to 2006, and overall from 2003 to 2006............70 3-3 Charlotte County: Number of buildings on which H. garnotii and H. mabouia went extinct from 2003 to 2004, 2004 to 2006, and overall from 2003 to 2006...........70 3-4 Charlotte County: Mean number of H. garnotii and H. mabouia per occupied building for 2003, 2004, and 2006 surveys....................................................................70 3-5 Lee County: Number and proportion of buildings on which I observed Hemidactylus garnotii and/or H. mabouia each survey and year and the proportional change between years ...............................................................................71 3-6 Lee County: Comparison of single and repeated surveys to estimate the proportion of buildings occupied by Hemidactylus garnotii and/ or H. mabouia.........72 3-7 Lee County: Hemidactylus garnotii model selection statistics and parameter estimates for one-year, one-species models...................................................................73 3-8 Lee County: Hemidactylus mabouia model selection statistics and parameter estimates for one-year, one-species models...................................................................74 3-9 Lee County: Goodness-of-fit test results as calculated in Program Presence (V. 2.0) for the one-species, one-year model in which detection probability was survey-specific........................................................................................................75 3-10 Lee County: Hemidactylus garnotii model selection statistics and parameter estimates for two-year, one-species models for 2004 and 2006....................................76 3-11 Lee County: Hemidactylus mabouia model selection statistics and parameter estimates for two-year, one-species models for 2004 and 2006....................................77 3-12 Lee County: 2004 and 2006 two-species, one-year model selection statistics for H. garnotii and H. mabouia evaluating if the occurrence of one species on a building affects the detection of the other species on that building............................78 8

PAGE 9

3-13 Lee County: 2004 and 2006 two-species, one-year model selection statistics evaluating if H. garnotii and H. mabouia co-occurred less often than would have been expected by chance.......................................................................................79 4-1 Description of buildings surveyed in Charlotte County................................................129 4-2 Description of buildings surveyed in Lee County.........................................................130 4-3 Summary of Charlotte County Hemidactylus garnotii surveys.....................................131 4-4 Summary of Lee County Hemidactylus mabouia surveys.............................................132 4-5 Proportion of total new captures for each stage.............................................................133 4-6 Number of each size class captured for each H. garnotii survey and the corresponding H. mabouia survey and the mean air temperatures for each survey......134 4-7 Multistate goodness-of-fit test results and estimated variance inflation factor ()for survival probability estimation models....................................................135 ^c 4-8 Survival probability model selection summary for H. garnotii.....................................135 4-9 Survival probability model selection summary for H. mabouia....................................135 4-10 H. garnotii survival, recapture, and transition probability estimates.............................136 4-11 H. mabouia survival, recapture, and transition probability estimates............................139 4-12 Mean temperature ( o C) and monthly growth rate (mm/day) for non-adult, adult, and all H. garnotii and H. mabouia...............................................................................141 4-13 Multistate goodness-of-fit test results and the estimated variance inflation factor () for breeding proportion estimation models...................................................141 ^c 4-14 Breeding proportion model selection summary for H. garnotii....................................141 4-15 Breeding proportion model selection summary for H. mabouia...................................141 4-16 Parameter estimates for breeding proportion models for H. garnotii............................142 4-17 Parameter estimates for breeding proportion models for H. mabouia...........................143 4-18 Number of gravid and nongravid adult female H. garnotii captured each survey and the nave and adjusted breeding proportion............................................................144 9

PAGE 10

4-19 Number of gravid and nongravid adult female H. mabouia captured each survey and the nave and adjusted breeding proportion............................................................145 4-20 Amount of original tail missing at the first capture of each gecko for H. garnotii and H. mabouia..............................................................................................................146 4-21 Amount of tail missing for male and female adult H. mabouia at the first capture for each gecko...................................................................................................147 4-22 Number of H. garnotii captured on each survey building.............................................147 4-23 Number of H. mabouia captured on each survey building............................................148 4-24 Estimated mean number of subadult and adult H. garnotii on each survey building during each survey period (SE).......................................................................148 4-25 Estimated mean number of subadult and adult H. mabouia on each survey building during each survey period (SE).......................................................................148 4-26 Estimated density of H. garnotii per linear meter of outer wall surface........................149 4-27 Estimated density of H. mabouia per linear meter of outer wall surface......................149 5-1 Gekko gecko, Hemidactylus frenatus, H. turcicus, and Tarentola annularis observed or captured from 2001 to 2006.......................................................................167 B-1 Charlotte County survey building locations and number of Hemidactylus garnotii and H. mabouia seen........................................................................................180 B-2 Lee County survey building locations and number of Hemidactylus garnotii and H. mabouia seen each survey..................................................................................189 C-1 Geckos captured in Sarasota..........................................................................................202 10

PAGE 11

LIST OF FIGURES Figures page 2-1 Number of Florida counties in which Hemidactylus spp. have been collected since 1963......................................................................................................................32 2-2 Number of Hemidactylus specimens collected in Florida and deposited in surveyed museums each decade.....................................................................................33 2-3 Number of Hemidactylus specimens per Florida County deposited in surveyed museums.........................................................................................................34 2-4 Charlotte and Lee Counties Florida...............................................................................35 2-5 Federal, state, and county numbered roads in Charlotte and Lee Counties Florida......36 2-6 Mean monthly high and low air temperatures ( o C) and rainfall (cm)...........................37 3-1 County, state and federal roads surveyed in Charlotte and Lee Counties Florida.........80 3-2 Charlotte County Number and proportion of buildings on which I observed Hemidactylus garnotii and/or H. mabouia each survey year.........................................81 3-3 Charlotte County: Gecko species captured at each survey building in 2003.................82 3-4 Charlotte County: Gecko species captured at each survey building in 2004.................83 3-5 Charlotte County: Gecko species captured at each survey building in 2006.................84 3-6 Charlotte County: Proportion of buildings occupied by Hemidactylus garnotii and H. mabouia each survey year..................................................................................85 3-7 Charlotte County: Building colonization and extinction rates for Hemidactylus garnotii and H. mabouia from 2003 to 2004, 2004 to 2006, and overall from 2003 to 2006..................................................................................................................86 3-8 Lee County: Gecko species captured at 176 survey buildings in 2004.........................87 3-9 Lee County: Gecko species captured at 71 survey buildings in 2004...........................88 3-10 Lee County: Gecko species captured at 68 survey buildings in 2006...........................89 3-11 Lee County: Proportion of buildings on which I observed Hemidactylus garnotii and/or H. mabouia each survey and year......................................................................90 4-1 Aerial view of Charlotte County study area..................................................................150 11

PAGE 12

4-2 Aerial views of Lee County study area..........................................................................151 4-3 Ground level views of Lee County study area...............................................................151 4-4 Peering behind gutters with helmet mounted light and holding extendable pole with squeegee mounted at the end.........................................................................152 4-5 Fold out rulers on end of extendable painters pole.......................................................152 4-6 Number of H. garnotii captures for each snout-vent length (mm) for new captures and all captures (new plus recaptures).............................................................153 4-7 Number of H. mabouia captures for each snout-vent length (mm) for new captures and all captures (new plus recaptures).............................................................153 4-8 Number of H. garnotii and H. mabouia captured throughout the study and the mean temperature ( o C).............................................................................................154 4-9 Number of juvenile H. garnotii and H. mabouia captured and the mean temperature throughout the study..................................................................................155 4-10 Number of subadult H. garnotii and H. mabouia captured and the mean temperature throughout the study..................................................................................156 4-11 Number of adult H. garnotii and H. mabouia captured and the mean temperature throughout the study..................................................................................157 4-12 Mean daily growth rate (mm) and temperature ( o C) for each month for all size classes combined....................................................................................................158 4-13 Mean non-adult daily growth rate (mm) and temperature ( o C) for each month............158 4-14 Mean adult daily growth rate (mm) and temperature ( o C) for each month...................158 4-15 Mean H. garnotii and H. mabouia maturation times, in days, for juveniles hatched in each month...................................................................................................159 4-16 Mean number of gravid H. garnotii and H. mabouia captured each month..................159 4-17 Number of gravid H. garnotii compared to the total number of adult females captured for each snout-vent length for surveys 19 and after..........................160 12

PAGE 13

4-18 Proportion of adult H. garnotii that were gravid for each snout-vent length (SVL) (mm) compared to the total number captured at each SVL (gravid at each SVL / (gravid + nongravid at each SVL)), the proportion that were gravid at each SVL as compared to the total number of gravid at all SVLs (gravid at each SVL / gravid at all SVLs), the proportion gravid for each SVL as compared to the total number of adults captured (gravid at each SVL / (gravid + nongravid adults at all SVLs), and the proportion of each SVL as compared to the total number of adults captured ((gravid + nongravid adults at each SVL / (gravid + nongravid adults at all SVLs)......................161 4-19 Number of gravid H. mabouia compared to the total number of adult females captured for each snout-vent length for surveys 3 and after............................162 4-20 Proportion of adult H. mabouia that were gravid for each snout-vent length (SVL)(mm) compared to the total number captured at each SVL (gravid at each SVL / (gravid + nongravid at each SVL)), the proportion that were gravid at each SVL as compared to the total number of gravid at all SVLs (gravid at each SVL / gravid at all SVLs), the proportion gravid for each SVL as compared to the total number of adults captured (gravid at each SVL / (gravid + nongravid adults at all SVLs), and the proportion of each SVL as compared to the total number of adults captured ((gravid + nongravid adults at each SVL / (gravid + nongravid adults at all SVL.........................163 5-1 Other geckos observed in Lee County...........................................................................169 5-2 Geckos observed in southern Sarasota County..............................................................170 13

PAGE 14

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DISTRIBUTION AND LIFE HISTORY OF TWO NON-INDIGENOUS GECKOS Hemidactylus garnotii AND H. mabouia IN SOUTHWEST FLORIDA By Gregg Stuart Klowden December 2007 Chair: Michael P. Moulton Major: Wildlife Ecology and Conservation This research was directed towards discerning whether or not changes in the distribution and abundance of the non-indigenous gecko Hemidactylus garnotii in Charlotte and Lee Counties, Florida could be linked to the co-occurrence of the more recently introduced H. mabouia and if so, what role life history differences might play. I presented data on their distribution, co-occurrence, colonization and extinction rates, demographics, and other life history traits. Surveys indicated that, from 2003 to 2006 in Charlotte County and 2004 to 2006 in Lee County, the proportion of buildings on which H. garnotii occurred decreased whereas for H. mabouia it increased. The extinction rate for H. garnotii was high and the colonization rate low while for H. mabouia the extinction rate was low and colonization rate high. In Lee County, while both species were distributed randomly in 2004, by 2006 they were less likely to co-occur than would be predicted by chance. Though colonization by H. mabouia did not appear to be the primary cause of extinction of H. garnotii from a building, on buildings on which they co-occured, the extinction rate of H. garnotii was significantly higher than for H. mabouia. Further, the ability of H. garnotii to colonize buildings occupied by H. mabouia was significantly lower than for buildings that were unoccupied. 14

PAGE 15

I estimated life history traits including stage specific survival, fecundity, mean snout-vent length, weight, maturation size, growth rate, time to maturation, inter-clutch interval, mean and maximum longevity, and other parameters using mark-recapture techniques. In Charlotte County I captured 914 H. garnotii 2,200 times and in Lee County 546 H. mabouia 1,030 times. H. mabouia was nearly 13% longer and more than 90% heavier than H. garnotii. Both species were similar in that they matured rapidly and had high growth rates. Hemidactylus garnotii fecundity was twice that of H. mabouia. Adult survival was similar, however juvenile survival of H. mabouia was double that of H. garnotii. Data on the occurrence of other non-indigenous gecko species and the northward dispersal of H. mabouia into Sarasota County are also presented. 15

PAGE 16

CHAPTER 1 INTRODUCTION Of the numerous species of the family Gekkonidae found in Florida in the early 2000s, only the Florida reef gecko, Sphaerodactylus notatus, is native to the state. This species occurs in Monroe County and coastal Broward and Miami-Dade Counties (Conant and Collins 1991, Bartlet and Bartlet 1999). The first nonindigenous gecko recorded in Florida, Hemidactylus turcicus, was collected in Key West by Hebard in 1910 (Fowler, 1915) and by 1922 was considered to be well established (Stejneger 1922). By 1936, this species occurred on mainland Florida (Barbour 1936). Since then, as many as 14 free-ranging nonindigenous gecko species have been reported in various parts of Florida (Bartlet and Bartlet 1999, Meshaka et al. 2004, Krysko and Daniels 2005) of which, Krysko and Daniels (2005) considered ten to be established. In addition to H. turcicus, only two other species, which also belong to the genus Hemidactylus, are commonly seen and widespread. Hemidactylus garnotii was first documented in Florida from the Miami area in 1963 (King and Krakauer 1966) and H. mabouia from Crawl Key, Monroe County in 1990 (Lawson et al. 1991). Since their introduction, both species have expanded their ranges considerably such that H. garnotii is now known from 41 counties and H. mabouia from 20 counties (Townsend and Krysko 2003, Meshaka et al. 2004, Krysko et al. 2005, FLMNH 2007, Klowden see Chapter 5). Another congener, H. frenatus, was first reported in 1993 in Monroe County (Meshaka et al. 1994c) and is now known from Broward, Lee, and Miami-Dade Counties (Krysko and Daniels 2005). Although this species does not appear to have spread as rapidly as the other Hemidactylus species (Meshaka et al. 1994c, Krysko et al. 2005), it warrants close attention as it can occur at extremely high densities (Case et al. 1994, Klowden see Chapter 5) and has been an extremely prolific invader in many places around the world (Bauer and Henle 1994). A fifth 16

PAGE 17

congener, H. platyurus, formerly Cosymbotus platyurus (Carranza and Arnold 2006), has been reported from several isolated locations since the mid 1980s but does not appear to have dispersed beyond its point of introduction (Meshaka and Lewis 1994, Hauge and Butterfield 2000, Krysko and Daniels 2005). In Florida and elsewhere, these Hemidactylus species have been reported to primarily occur on buildings (Broadley 1977, Meshaka 2000) however, they have also been found in more natural areas under bark, inside snags, and beneath rocks (Voss 1975, Schwartz and Henderson 1991, Meshaka 1999, Townsend et al. 2002, Enge et al. 2004). Several primary publications and numerous distributional notes have documented the gross distributions of nonindigenous geckos throughout the state (King and Krakaur, 1966; Wilson and Porras, 1983; Butterfield et al. 1997, Meshaka et al. 2004, Krysko and Daniels 2005, Krysko et al. 2005). Some authors have suggested the successive replacement of previously established widespread species by more recently invading species (i.e. H. turcicus by H. garnotii and H. mabouia and H. garnotii by H. mabouia) (Meshaka 1995, 2000, Meshaka et al. 2004). Despite literature references to distributional shifts and the purported role of species interactions in these shifts (Meshaka et al. 1994b, 1994c, 2005, Powell et al. 1998, Punzo 2005), this pattern has not been studied. There have been no detailed surveys of their distribution and abundance and except for H. turcicus, little is known of their life histories (Selcer 1986, Punzo 2001, Gomez-Zlatar and Moulton 2005, Gomez-Zlatar et al. 2006). The factors that influence the rates of spread and relative abundances of these species remain largely unknown. To investigate the hypothesized replacement of an already established non-indigenous species by a more recently arriving species I first conducted preliminary surveys of buildings throughout six counties in southwest Florida during 2001 and 2002. Results indicated that H. garnotii appeared to be well established throughout Charlotte and Lee Counties while H. 17

PAGE 18

mabouia appeared to be comparatively less common in Lee County and minimally present in Charlotte County. In surrounding counties, H. mabouia was either very abundant (Collier) or did not appear to occur at all (Hardee, Polk, Sarasota). Based on these observations, I decided to conduct my research in Charlotte and Lee Counties. My research had three primary goals. First was to define the distributions of H. garnotii and H. mabouia in Charlotte and Lee Counties and to clarify how they were distributed in relation to one another. Second was to investigate if changes appeared to be occurring in their distributional patterns over time. I accomplished these first two goals by developing a methodology to randomly select buildings and then repeatedly surveyed these buildings for the presence of geckos. I present the results of this study in chapter 3 in a form easily adaptable for submittal for publication. I present methodological details and complete survey results in Appendicies A and B. My third goal was to investigate demographic and other life history traits of H. garnotii and H. mabouia and explore if these could offer a preferential survival advantage to one species as compared to the other. Using mark-recapture techniques, I collected data on a population of H. garnotii in Charlotte County from 2001 to 2005 and on a population of H. mabouia in Lee County from 2003 to 2005. I present the results of this study in chapter 4 in a form easily adaptable for submittal for publication. In chapter 2, I include background information regarding H. garnotii and H. mabouia distributions in Florida, previously published life history data, and a brief description of Charlotte and Lee Counties. In chapter 5, I present miscellaneous distributional observations of these and other non-indigenous gecko species in southwest Florida. In chapter 6, I summarize my results. 18

PAGE 19

CHAPTER 2 BACKGROUND Distribution Hemidactylus garnotii is native to Southeast Asia (Carranza and Arnold 2006). It has been reported to be widespread from eastern India to Indonesia, on many of the Philippine Islands, and on numerous south Pacific islands, where it may have been accidentally transported by humans in ancient times (Kluge and Eckardt 1969, Brown and Alcala 1978). The exact timing of the introduction of Hemidactylus garnotii in Florida is unknown. Though the species was first reported in 1966 (King and Krakauer 1966) in Dade County, it was known to occur prior to 1960 in several locations elsewhere in the county (Wilson and Porras 1983). In Charlotte County, H. garnotii was first reported in 1991 by Conant and Collins (1991). Hemidactylus mabouia was first collected there in 2001 (Klowden 2002). In Lee County, H. garnotii was first recorded in 1971 on Sanibel Island (McCoy 1972) and was reported on the mainland in 1983 (Wilson and Poras 1983). Hemidactylus mabouia is native to much of the southern half of Africa from just north of the Equator to the Tropic of Capricorn and also occurs on Madagascar and other offshore islands (Loveridge 1947, Kluge 1969, Carranza and Arnold 2006). It was reported to have been brought to the New World either aboard slave ships around 500 years ago or possibly much earlier via transatlantic rafting on trees (Kluge 1969). Hemidactylus mabouia was first reported in Florida in 1990 on Crawl Key, Monroe County (Lawson et al. 1991). Based on its ubiquity, it was speculated to have been present since the early 1980s (Butterfield et al. 1993, Meshaka et al. 1994b, 2004). Hemidactylus mabouia was first collected in Lee County in 2000 (Townsend et al. 2002) on Gasparilla Island. It was first collected on the mainland in 2001 (Klowden 2002). 19

PAGE 20

More than 50 specific locality references, some separated by hundreds of miles, of the occurrence of Hemidactylus spp. in various Florida counties occur in the literature. Such references have led some authors to infer ubiquitous distributions of some species throughout counties and regions (Behler and King 1979, Conant and Collins 1991, 1998, Bartlett and Bartlett 1999, Meshaka et al. 2004, Punzo 2005). Additionally, limited surveys have led to conclusions regarding the replacement of one species by another (Meshaka 1995, 2000, Meshaka et al. 1994b, 2005). Though such records verify the presence of each species in a county, they generally have not helped to illustrate the complexities of their distributions or abundance throughout a county. To illustrate this, from February to May 2002, I queried 54 major museum collections (Table 2-1) for Hemidactylus specimens that were collected in Florida. In 2007, from their online database (FLMNH 2007), I updated my query of the Florida Museum of Natural History records, the principal curator of Florida Hemidactylus specimens. I have not verified the correct identification of these specimens and therefore do not presume these queries to be exact, and undoubtedly there are additional museum specimens I have overlooked and private collections I have not queried. Nonetheless, several gross patterns emerge. When viewed at the county level, a gross assessment of H. turcicus and H. garnotii distributions over time can be made. County distributional records accumulated since the introduction of each species indicate that until the 1990s there was a gradual increase in the number of counties in which H. turcicus and H. garnotii occurred. In the 1990s the count increased rapidly (Figure 2-1). However, this increase co-occurred with a dramatic increase in the number of specimens deposited in museums (Figure 2-2). Thus, the addition of new counties along this timeline might not accurately reflect the real rate of range expansion for these species as much as it reflects increased collecting effort. 20

PAGE 21

When viewed at a fine scale, these data reveal that substantial numbers of geckos have been collected in only a few counties and in most counties the number collected is fewer than five (Figure 2-3). Though such data show that a species occurred in a county on a particular building on the date of collection, they do not allow an assessment of how widespread or abundant the species was in the county. For instance, in 2002 I observed two and captured one H. turcicus on a building in Lee County (see Chapter 5). This was a county record and demonstrated its occurrence in Lee County, thereby adding another notch to the list of counties in which it occurred. However, with more extensive surveys, I determined that H. turcicus did not occur on any of the hundreds of other buildings I surveyed, making its existence in Lee County precarious at best. Systematic large-scale sampling is necessary to determine relative distributions, abundances, patterns of co-occurrence, and changes over time. This is not possible with the small number of specimens currently available for most counties. Opportunistic small-scale sampling may increase the likelihood that the absence of a record in a locality is due to insufficient sampling. Further, such sampling is likely to result in an inaccurate assessment of the true distributions and persistences since many opportunistic collectors may not collect specimens of species previously known to occur in a county, but only specimens thought to represent new county records. Thus to better elucidate these processes, I conducted detailed surveys throughout Charlotte and Lee Counties. Natural History In Florida and elsewhere, H. garnotii and H. mabouia are most commonly associated with buildings or human debris (Lando and Williams 1969, Broadley 1977, Vitt and Zani 1997, Meshaka 2000). Though primarily edificarian they are occasionally found in more natural areas 21

PAGE 22

under bark, inside snags, and beneath rocks (Voss 1975, Broadley 1977, Schwartz and Henderson 1991, Rodrigues 1996, Meshaka 1999, Townsend et al. 2002, Enge et al. 2004). Few data have been published regarding the demographics or other life-history attributes of H. garnotii and H. mabouia in Florida and elsewhere. Most inferences are based on very small sample sizes, often involving only a couple animals, and some are anecdotal offering little or no substantiating data. Size Hemidactylus garnotii has been reported up to 65 mm SVL in Central America (Savage 2002, Khler 2003), where they are also introduced. Similarly, the maximum size reported in Florida was 64 mm SVL (Butterfield 1996, Butterfield et al. 1997). The mean adult size, based on 52 individuals from Everglades National Park, was 55.0 mm SVL (Meshaka 1994). In a subsequent study at Everglades National Park, mean adult body size for 11 specimens was 56.2 3.0 mm SVL and, based on 12 specimens (unclear if sample was the same 11 adults plus one juvenile), mean body size for all size classes combined was 50.8 9.1 mm SVL (Meshaka 2000). Florida hatchling size was reported by Meshaka (1994) to be 22 to 26 mm with a mean of 23.7, based on 6 individuals. Voss (1975) anecdotally reports 24 to 26 mm SVL. Gibbons and Zug (1987) report 27 to 28 mm SVL for three hatchlings in Tonga. For H. mabouia, the maximum SVL reported in Florida is 68 mm (Butterfield et al. 1997). I have recorded individuals in Fort Myers, Florida up to 71 mm SVL (Klowden unpublished). Central American specimens have been reported to 75 mm SVL (Khler 2003), Peruvian specimens to 71 mm SVL (Dixon and Soini 1986), and Brazillian specimens to 68 mm SVL (Vitt 1986). Mean adult size for 20 Florida specimens was 58.7 4.3 mm SVL and, based on four males and 16 females, was not considered to differ between sexes (Meshaka 2000). 22

PAGE 23

Mean body size for all size classes combined (n=26) was 50.7 11.4 mm SVL and was not considered significantly different from H. garnotii. For 29 Brazilian specimens, the mean adult size was 58.3 mm SVL (Vitt 1986). In Africa, Tanzanian specimens have been reported to 86 mm SVL (Loveridge 1947). In Brazil, for a small sample of male and female adults, the SVL, head length, and head width were not sexually dimorphic (Vitt 1986). Fitch (1981) considered female SVL to be slightly larger than for males. However, sample size was not given and he acknowledged that these data may be from a small data set. The smallest hatchlings reported in Florida were 22 mm SVL (Meshaka et al. 2004). Hatchlings from Trinidad and Tobago were reported to be 20 mm SVL (Murphy 1997), from Iquitos, Peru to be 20 to 25 mm SVL (Dixon and Soini 1986), Venezuela to be 20 mm SVL (Fuenmayor et al. 2005), Carriacou, Grenada to be 21 to 25 mm SVL (Pendlebury 1972), and in Columbia to average 23.4 mm SVL (n = 18) (Bock 1996). Reproduction Hemidactylus garnotii is an all female, parthenogenetic species (Kluge and Eckardt 1969). Reproductive maturity in Florida has been reported to occur before one year of age at 49 mm SVL (Meshaka 1994). This estimate was based on an unspecified sample size of individuals with shelled eggs and enlarged follicles. Though a follicle size greater than one mm diameter was considered to indicate maturity, it was not stated if follicles were yolked or contained corpora lutea, standard criteria used by other authors. For example, Vitt (1986) considered H. mabouia to be reproductive if they contained vitellogenic follicles, Lin and Cheng (1984) considered H. frenatus mature if they contained yolked follicles greater than 2.5 mm diameter, and Shanbhag et al. (1998) considered H. brooki to be in the breeding phase if corpora lutea were present. Based on the presence of both shelled eggs and two sets of enlarged follicles, Meshaka 23

PAGE 24

(1994) also estimated that at least three clutches could be produced annually by H. garnotii. However, follicle count may not be indicative of the actual number of clutches since it is unclear if any of these follicles will develop further and result in ovulation. Based on similar criteria, Meshaka et al. (1994a) asserted that H. mabouia could produce up to seven clutches annually. This statement is misleading and has been misinterpreted as H. mabouia has been reported to oviposit up to seven clutches annually (Krysko et al. 2003b). Vitt (1986) considered the presence of shelled eggs and vitellogenic follicles to be evidence of multiple clutches in Brazilian H. mabouia. For H. mabouia, a sexually reproducing species, Meshaka et al. (1994a) reported maturity to occur at 46 mm SVL for males and 47 mm SVL for females though Meshaka et al. (2004) reported maturity to occur at 49 mm SVL. Meshaka et al. (1994a) stated that maturation size was similar to Brazilian H. mabouia specimens reported by Vitt (1986) who stated maturity was at 52 mm SVL. While No data on age at maturity are available for H. mabouia in Florida but Brazilian specimens are considered to mature at less than six months of age (Vitt 1992). Based on a sample of 17 adult H. mabouia, Howard et al. (2001) claim the sex ratio was far from 1:1. This sample size is too small to make such an inference. For 77 Brazilian H. mabouia captured, Zamprogno and Teixeira (1998) found a 1:1 sex ratio. Reproduction for H. garnotii was considered to be continuous in Florida (Meshaka 1994) based on adults containing shelled eggs in January, March, July, August, October, and November and enlarged follicles greater than 1 mm diameter in all other months except February. Again, enlarged follicles may not be a reliable estimator. Meshaka et al. (1994a) reported that H. mabouia also reproduced continuously in Florida but presented no data to support this conclusion. Meshaka (2000) considered both species to have continuous reproduction but again 24

PAGE 25

no data were presented. In Brazil, Vitt (1986) captured gravid H. mabouia females in many months of the year which he believed suggested nearly continuous breeding. This was misquoted by Bock (1996) as Vitt encountered gravid H. mabouia throughout the year. In the West Indies, Schwartz and Henderson (1991) collected eggs from January to September and hatchlings from February to November. This led me to investigate if this pattern occurred in Florida (Chapter 4). Vitt (1986) considered the clutch size of H. mabouia to be almost always two. Henkel and Schmidt (1995) also considered the clutch size to be two and also, presumably in an incubator, considered incubation time to be about 60 days at 26-30 o C. Punzo (2005) and Voss (1975) stated that egg incubation time in Florida was 50 and 60 days, respectively however, neither provided any data. In Columbia, the longest time between egg encounter and hatching, a possible maximum incubation period estimator, was 71 days (Bock 1996) and in Peru was 77 days (Dixon and Soini 1986). Dixon and Soini also reported that one set of eggs incubated 55 and 56 days before hatching and included hatch times of other eggs from one to three months. However, only dates of hatching, not oviposition, were given making the utility of these data somewhat limited. There are not any published data on interclutch intervals for these species in Florida. In Columbia, sets of eggs found in the wild and incubated, hatched 16 days apart giving an interclutch interval of 16 days (Bock 1996). Activity Though both H. garnotii and H. mabouia are considered to be primarily nocturnal (Beebe 1944, Hoge 1950, Vitt 1995, Vitt and Zani 1997, Meshaka et al. 2004), few data are available on the activity times of these species. H. garnotii was said to emerge before dark and remain active until 0200 hours but not after (Frankenberg 1982). Having collected specimens until 0430, my 25

PAGE 26

own observations indicate that both species forage actively throughout the night. Limited diurnal activity has been noted for both species. In Florida, H. garnotii occasionally forages on overcast and less often on clear days and H. mabouia has been seen basking (Meshaka et al. 2004). In Venezuela, H. mabouia was occasionally observed in the morning on overcast days (Fuenmayor et al. 2005). Reports of diurnal activity have also been made from Suriname, Puerto Rico, and Trinidad and Tobago (Hoogmoed 1973, Rivero 1978, Murphy 1997). Almost no data are available on seasonal variation in activity. Based on the number of specimens observed nightly, in Everglades National Park, Florida, both species were considered to respond similarly to variations in nightly air temperature, relative humidity, and volume of rainfall and were considered to be more active in the wet season (May to October) than the dry season (November to April) (Meshaka 2000, 2001). However, only minimal weather variation data were presented and no data were presented on nightly abundance. In Santa Cruz Bolivia, H. mabouia was said to have not been seen during cool months from May to October (Myers 1945). Predators Meshaka (2001) and Meshaka et al. (2004) reported the knight anole, Anolis equestris, the corn snake, Elaphe guttata, and the Cuban treefrog, Osteopilus septentrionalis to prey on H. garnotii. Steiner (1983) reported predation by the common web building house spider Achaearanea tepidariorum. Meshaka et al. (2004) stated that O. septentrionalis has been shown to reduce population size and increase mean individual body size in populations of H. garnotii by preying upon juveniles. To substantiate this pattern, they stated that on buildings on which O. septentrionalis was present, H. garnotii had a larger mean SVL (57.8 3.7 mm, n = 30) and lower proportion of juveniles (19%) than on buildings on which only H. garnotii was present (55.3 3.0 mm, n = 37 26

PAGE 27

and 34% juveniles). They did not state what months these surveys were conducted which, due to substantial monthly variation in juvenile abundance (see chapter 4), could greatly affect conclusions. Methodological considerations aside, based on the means and standard deviations they presented, the conclusion that SVLs differ appears unwarranted. Insufficient data were presented to determine if their conclusion that the proportion of juveniles differed on buildings with and without O. septentionalis was justifiable. Reported predators of H. mabouia in Florida include O. septentrionalis and the tokay gecko, Gekko gecko (Meshaka et al. 2004). In the British Virgin Islands it has been eaten by the lizard Anolis cristatellus wileyae and the snake Alsophis antillensis (Grant 1932b, Owen and Perry 2005), in Brazil by the collared lizard Tropidurus torquatus and the snakes Oxyrhopus rhombifer and Siphlophis pulcher (Araujo 1991, Sazima and Argolo 1994, Galdino and Van sluys 2004, Maschio et al. 2004), and in Puerto Rico by the snake Alsophis portoricensis (Henderson and Sajdak 1996). African varieties have been eaten by the bat Cardioderma cor in Kenya (Wojnowski and Selempo 2005), several lizards in South Africa (Egan 2005, Haagner 1997), the lizard Agama agama in Ghana (Gramentz 2000), and the kestrel Falco tinnnunculus carlo, monitor lizard Varanus niloticus, and various snakes in Tanzania (Loveridge 1941) where it was also attacked by a spider, Solpuga spp. (Loveridge 1947). To avoid predation, both species use crypsis and position themselves near previously used hiding spots where they can quickly retreat via known escape routes (Vitt 1995, Klowden, personal observation). Diet Both species are considered to be opportunistic sit-and wait predators (Vitt 1995, Meshaka 2000). Loveridge (1947) described the diet of H. mabouia in Tanzania as consisting of anything they are capable of overpowering. The diets of H. garnotii and H. mabouia in 27

PAGE 28

Florida have been reported to be broad and similar, and to include a variety of insects and spiders, especially flies, mosquitoes, and hymenoptera (Meshaka 2000, 2001). A cannibalistic subadult (43 mm SVL) H. mabouia was documented in Brazil preying on a 22 mm SVL juvenile (Zamprogno and Teixeira 1998). An attempted instance of saurophagy was seen in Tanzania on the gecko Lygodactylus picturatus mombasicus (Loveridge 1947). In Zimbabwe, the closely related H. m. tasmani had a juvenile flat-lizard, Platysaurus intermedius rhodesianus, in its stomach (Broadley 1977). Dispersal The primary mode of dispersal for these species is believed to be human mediated. Hemidactylus garnotii has been found in cultivated palms in Florida and the Bahamas and on construction materials (Meshaka 1996, 2001), was presumed to have been transported to Egmont Key in delivered supplies (Dodd and Griffey 2002), and was collected on a building next to a landscaping business 139 kilometers north of its closest population at the time (Myers 1978). I have similarly found disjunct populations of H. mabouia (see chapter 5). In New Zealand, South Africa, and Uruguay H. garnotii and H. mabouia, as well as H. turcicus and H. frenatus, have been found on boats and airplanes carrying sugar, produce, and other supplies (Bourquin 1987, Simo et al. 1988, Gil et al. 2001). In South Africa, H. mabouia has been found in a suitcase and observed on cargo trucks and automobiles (Bourquin 1987, Haagner and Branch 1996). County Descriptions Charlotte and Lee Counties are located in southwest, Florida (Figure 2-4). Charlotte County, located at approximately 26.9 o N/ -81.9 o W, encompasses 1,796 km 2 and had a human population of 141,627 in 2000. Lee County, located at approximately 26.6 o N/ -81.8 o W, is slightly larger at 2,082 km 2 but has a considerably larger human population of 440,888 in 2000 28

PAGE 29

(United States Census Bureau 2000). The human population in both counties is rapidly expanding. From 2000 to 2006 Charlotte Countys population grew 9.0% and Lee County 29.5% (United States Census Bureau 2006). Excluding I-75, Charlotte County contains approximately 278 and Lee County approximately 478 linear kilometers of federal, state, and county numbered roads (Figure 2-5). In Charlotte County, most development is clustered around US-41 and the western portion of US-17 in the central and northern part of the county in Port Charlotte and Punta Gorda and in the northwest portion of the county on western SR-776 and northern SR-775 in Englewood. The southern and eastern portions of the County are largely undeveloped. Two large protected areas comprise the bulk of eastern Charlotte County. The Fred C. Babcock/ Cecil B. Webb Wildlife Management Area is about 264 km 2 and approximately 274 km 2 of the recently protected Babcock Ranch Preserve lies in Charlotte County (Charlotte County 2007b). In Lee County, most of the western half of the county is developed except for some areas in the north. Though several large developed areas occur, many parts of eastern Lee County are undeveloped however, the eastern portion is rapidly developing. Rural areas along CR-850 and SR-82 have recently been developed. In the northeast, CR-78 is primarily undeveloped, however there are numerous developments proposed (personal communication Suzie Derheimer, Lee County Department of Community Development). High temperatures in both counties (Figure 2-6) are very similar, ranging from approximately 24 o to 33 o C. The hottest months are generally May through September with average high temperatures of 32 o to 33 o C and low temperatures of 19 o to 24 o C (MSN 2007). In Lee County, low temperatures range from 12 o to 24 o C but for Charlotte County are about 1 o C lower. The coldest months are generally December through February with average high temperatures of 24 o to 25 o C and low temperatures of 11 o to 14 o C (MSN 2007). 29

PAGE 30

In both counties, the relative humidity is fairly constant throughout the year. Daily averages normally range from about 75 to 80% throughout the year. The wet season for both counties extends from approximately June through September. Average rainfall in Charlotte County is about 20 cm per month while in Lee County is about 23 cm per month. The dry season extends from October through May with averages of approximately 6 cm per month in both counties (Weather Underground 2007). 30

PAGE 31

Table 2-1. Museums queried for Florida Hemidactylus specimens. Institution Abbreviation American Museum of Natural History AMNH Philadelphia Academy of Natural Sciences ANSP Auburn University Museum AUM University of Washington Burke Museum of Natural History and Culture BMUW California Academy of Sciences CAS Carnegie Museum CM Cincinnati Museum of Natural History & Science CMC Cornell University Museum of Vertebrates CUMV Delaware Museum of Natural History DMNH Fort Hayes State University, Kansas, Sternberg Museum FHSM Florida Museum of Natural History, University of Florida FLMNH Field Museum of Natural History FMNH Georgia Museum of Natural History, University of Georgia GMNH Naturhistoriska Museet, Gothenburg, Sweden GNM Illinois Natural History Survey INHS University of Kansas Museum of Natural History KU Natural History Museum of Los Angeles County LACM Louisiana Museum of Natural History, Louisiana State University LSUMZ Louisiana State University in Shreveport Museum of Life Sciences LSUS Harvard University Museum of Comparative Zoology MCZ Milwaukee Public Museum MPM The University of New Mexico Museum of Southwestern Biology MSB Michigan State University Museum MSU University of California, Berkley Museum of Vertebrate Zoology MVZ North Carolina Museum of Natural Sciences, NC State University NCSM Natural History Museum of London NHML Natural History Museum Vienna NHMW National Museum of Victoria, Melbourne, Australia NMVM Swedish Museum of Natural History NRM Sam Noble Oklahoma Museum of Natural History, University of Oklahoma OMNH (Norman) Oxford University Museum OUM Royal British Columbia Museum RBCM Royal Ontario Museum ROM San Diego Natural History Museum SDNHM Strecker Museum SMBU Texas Cooperative Wildlife Collection, Texas A&M University TCWC Transvaal Museum, South Africa TMP Texas Memorial Museum, University of Texas at Austin TNHC Tulane University Museum of Natural History TU University of Arkansas, Fayetteville, University Museum UA Alabama Museum of Natural History, University of Alabama UAIC University of Arizona Museum of Natural History UAZ University of Colorado Museum UCM University of Michigan Museum of Zoology UMMZ University of Nevada, Las Vegas Barrick Museum of Natural History UNLV University of Nevada, Reno UNR University of Nebraska State Museum UNSM Smithsonian Institution National Museum of Natural History USNM University of Texas at Arlington Amphibian and Reptile Diversity Research Center UTA University of Texas at El Paso Laboratory for Environmental Biology UTEP Uppsala University, Sweden Evolutionsmuseet, Zoologisektionen UUZM Virginia Museum of Natural History VMNH Yale University Peabody Museum of Natural History YPM Zoologisches Forschungsinstitut and Museum Alexander Koenig Bonn Germany ZFMK 31

PAGE 32

051015202530354045196319661969197219751978198119841987199019931996199920022005Year H. turcicus H. garnotii H. mabouia H. frenatus Number of counties Figure 2-1. Number of Florida counties in which Hemidactylus spp. have been collected since 1963. 32

PAGE 33

21141813781978110196520200400600800100012001910s1920s1930s1940s1950s1960s1970s1980s1990s2000sDecade Number of collected Figure 2-2. Number of Hemidactylus specimens collected in Florida and deposited in surveyed museums each decade. 33

PAGE 34

34 Number of counties Figure 2-3. Number of Hemidactylus specimens per Florida County deposited in 051015202530 1-56-1011-2526-5051+Number of specimens surveyed museums. H. turcicus H. garnotii H. mabouia H. frenatus

PAGE 35

Lee County Charlotte County 35 Figure 2-4. Charlotte and Lee Counties Florida.

PAGE 36

Charlotte County Lee County 36 ................................... .. Federal, state, and county numbered roads Figure 2-5. Federal, state, and county numbered roads in Charlotte and Lee Counties Florida (excluding Interstate 75).

PAGE 37

05101520253035Januar y FebruaryMarchAprilMayJuneJul y AugustSeptemberOctoberNovemberDecember0510152025 Charlotte low temperature (C) Lee low temperature (C) Charlotte high temperature (C) Lee high temperature (C) Charlotte Precipitation (cm) Lee Precipitation (cm) Air Tem p erature ( oC ) Amount of p reci p itation ( cm ) 37 Month Figure 2-6. Mean monthly high and low air temperatures ( o C) and rainfall (cm) (MSN 2007, Weather Underground 2007).

PAGE 38

CHAPTER 3 DISTRIBUTION AND RELATIVE ABUNDANCE OF Hemidactylus garnotii AND H. mabouia IN TWO SOUTHWEST FLORIDA COUNTIES Introduction Hemidactylus garnotii and H. mabouia are nonindigenous gecko species that were first collected in Florida in 1963 and 1990, respectively (King and Krakauer 1966, Lawson et al. 1991). Since the time of their introduction, both species have expanded their ranges considerably, such that H. garnotii is now known from 41 counties and H. mabouia from 20 counties (Townsend and Krysko 2003, Meshaka et al. 2004, Krysko et al. 2005, FLMNH 2007, Klowden unpublished). Some authors have speculated that H. mabouia is now the most abundant gecko in southern Florida (Krysko et al. 2003a). The comparatively rapid range expansion of H. mabouia into areas already occupied by H. garnotii has purportedly been accompanied by a widespread reduction in the population size or complete displacement of H. garnotii (Meshaka and Moody 1996, Meshaka 2000, Townsend and Krysko 2003, Enge et al. 2004, Meshaka et al. 2004). Though these assertions may reflect actual regional patterns, there have only been two studies providing any evidence of occupancy or abundance changes over time and both were highly limited in scope (Meshaka 2000, Enge et al. 2004). No extensive population surveys have occurred to help illuminate the complexities of these species distributions. Further, hypotheses regarding the causal nature of this relationship (Meshaka 2000, Meshaka et al. 2004) have not provided sufficient quantitative data to support the conclusion that H. mabouia has reduced H. garnotii populations. To examine H. garnotii and H. mabouia distributions, I conducted surveys from 2003 to 2006. The objectives of my study were to assess 1) the current distributions of H. garnotii and H. mabouia and how they were distributed in relation to each other; 2) if their relative 38

PAGE 39

distribution or proportion of buildings on which they occurred was changing over time; 3) and if changes were occurring, whether a causal relationship suggested. To address these questions, it was necessary to survey an area in which H. garnotii was established on most buildings and H. mabouia, while clearly present, was on substantially fewer buildings, possibly indicating a fairly recent and advancing invasion. Such a pattern would provide an opportunity to document the shift in patch occupancy rates, where patches are buildings, should it occur during the sampling period. Preliminary surveys throughout six counties in southwest Florida during 2001 and 2002 indicated that these conditions were best satisfied in Charlotte and Lee Counties. Charlotte County encompasses 1,797 km 2 and has a human population of 141,627. Lee County is slightly larger at 2,082 km 2 with a considerably larger human population of 440,888 (United States Census Bureau 2000). Hemidactylus garnotii was first reported in Charlotte County, in 1991 by Conant and Collins (1991). Hemidactylus mabouia was first collected there in 2001 (Klowden 2002). In Lee County, H. garnotii was first recorded in 1971 (McCoy 1972) on Sanibel Island and was reported on the mainland in 1983 (Wilson and Poras 1983). The first reported collection of H. mabouia in Lee County was in 2000 (Townsend et al. 2002) on Gasparilla Island. It was first collected on the mainland in 2001 (Klowden 2002). Survey Methods I surveyed Charlotte County in August 2003, August 2004, and September 2006 and Lee County in October 2004 and October 2006. On each survey I recorded the presence and number of H. garnotii and H. mabouia on randomly selected buildings. I limited surveys to all federal, state, and county numbered roads in Lee and Charlotte Counties excluding I-75 (Figure 3-1). This comprised approximately 278 and 463 linear kilometers in Charlotte and Lee Counties, 39

PAGE 40

respectively. Survey points consisted of small to medium sized accessible public buildings, businesses, and churches. I defined accessible as those where at least of a buildings walls, usually meaning three sides, were clear from major debris or fences. I chose not to include large shopping malls and supermarkets unless no other buildings were in the area. Additionally, I did not include businesses open all night such as convenience stores and gas stations since the detection of geckos may be more difficult due to very bright lights and constant human activity. I also excluded buildings where it was unwise to survey in the middle of the night with a flashlight, such as banks and jewelry stores. Finally, I excluded buildings on which I encountered other gecko species, including H. frenatus and Tarentola annularis, since the focus of this study was H. garnotii and H. mabouia and the occurrence of other species may have biased results. The sampling methodology I employed consisted of two stages: pre-survey and data collection. Pre-survey Methods In July 2003 and September 2004 I conducted a pre-survey in Charlotte and Lee Counties, respectively. The pre-survey consisted of two parts: ArcView GIS random generation of potential survey points and selection of actual survey buildings via pre-survey ground-truthing of randomly generated points. Random generation of potential survey points I randomly generated potential survey points using a two step procedure. First, I created maps of county and state roads by editing 2000 Florida Geographic Data Library v. 3.0 (FGDL 2000) road layers with ArcView GIS v. 3.2, to include only county and state numbered roads excluding I-75. Second, I used the Random Point Generator extension v.1.1 for ArcView GIS 40

PAGE 41

3.x (Jenness 2001) to randomly generate the potential survey points. I designated the number of potential survey points in each county as one point for every 2,500 people. Since the human population in a county seems likely to be roughly correlated with the total number of buildings, I felt this to be a reasonable basis for selecting an amount of survey points that could be efficiently surveyed, would accurately assess distributions, and allow for statistical rigor. According to the 2000 federal census, the populations of Charlotte and Lee were 141,627 and 440,888 respectively. The number of corresponding potential survey points was 57 and 176, respectively. I generated twice the needed number of potential survey points to compensate for rural areas containing large stretches without any suitable buildings. This resulted in a map containing the randomly generated potential survey points and a list of the latitude and longitude of each point. The details of this methodology are elaborate and are laid out in Appendix A. Selection of survey buildings I selected my survey buildings using the following methodology. I randomly chose, by coin toss, the end of each state and county road from which I would begin traveling. In Microsoft Excel 2000 I reordered and numbered the potential survey points so that all of the points on a particular road were grouped together and the points were ordered in the previously chosen direction of travel along each road. I uploaded the potential survey point latitude and longitudes into a Garmin Etrex Global Positioning System (GPS) using Waypoint + v. 1.8.03 (Hildebrand 2001). I began at the randomly selected end of a road and proceeded until arriving at the first potential survey point. If a building was present at that point on either side of the road I designated it as a survey building. If buildings were located on both sides of the road I selected the building with a coin toss. If no buildings were present on either side of the road I continued travel in the original direction and selected the first building I encountered. If I did not encounter 41

PAGE 42

any building before reaching the subsequent potential survey point, I discarded the original point and repeated the selection procedure with the subsequent point. If I did not encounter a building before the end of the road I discarded the point. After I had visited all of the potential survey points, I discarded excess survey buildings according to the order they were originally generated not the order in which they were visited. For example, in Lee County I needed 176 survey buildings so I generated 352 points. If after visiting all 352 points I had 200 survey buildings, I discarded 24. I discarded those survey buildings associated with the later randomly generated points not the last 24 points I visited since I determined the order of the points I visited by logistical considerations and not randomly. Data Collection Data collection consisted of two parts: surveys and resurveys. Initial surveys allowed me to determine both species current distributions, how they were distributed in relation to one another, and the relative proportion of buildings on which they occurred. Resurveys in subsequent years allowed me to address whether their relative distributions or proportions were changing over time and if so, whether a causal relationship was suggested. In Charlotte County I surveyed the 57 randomly selected buildings from August 3, 2003 to August 4, 2003. Initial occupancy rates from these surveys, especially for H. mabouia, were very low and so I decided to expand the number of survey buildings by including the two closest buildings to each of the original survey points. On four nights from August 12, 2003 to August 20, 2003, I surveyed these additional buildings. In some cases I did not encounter additional buildings prior to reaching the next survey building. This resulted in a total of 148 buildings surveyed in Charlotte County in 2003. In 2004, on four nights from August 5 to 8, I resurveyed 42

PAGE 43

the same buildings from the 2003 survey. On three nights from September 25-27, 2006 I again resurveyed these same Charlotte County buildings. In Lee County, on seven nights between October 1, 2004 and October 14, 2004, I surveyed the 176 randomly selected buildings. However, preliminary surveys also suggested that on many buildings a species would not be detected at every visit (i.e. detection probability was less than 1), and that non-detection could not be interpreted as absence. Therefore, occupancy estimates based on single surveys were likely to underestimate the true occupancy level. Thus, to obtain a measure of detectability to improve my occupancy estimations, in Lee County, I surveyed a subset of buildings a total of three times in 2004 and the same subset three times in 2006. For Lee County surveys, within each year, the time between surveys was brief enough, i.e. less than one month, that I considered the populations closed to changes in occupancy (i.e. no colonization or extinction). Thus, on four nights between October 18, 2004 and October 26, 2004 I resurveyed 71 of the 176 buildings twice. I resurveyed these same 71 buildings three times on five nights from October 8, 2006 to October 17, 2006. I began all surveys shortly after sunset and ended at approximately 2-3:00 a.m. Using a headlamp and a painters pole with an 18 inch floor squeegee attached to the end, I slowly walked around each building one time peering behind signs, gutters, objects leaning against the building, and other observable hiding spots. I recorded the number of geckos of each species identified as well those that were unidentified. Since gecko species are easily confused, I used the painters pole/ squeegee to slow down, redirect, and/or capture fleeing geckos or those in which it was difficult to make a quick species determination to be certain of accurate identification. Occasionally, I used a ruler or yardstick to help extract unidentified animals observed fleeing to hiding spots. 43

PAGE 44

Data Analysis Distribution To visually assess countywide distributions, I mapped the buildings on which I found no geckos, where each species occurred alone, and where both species occurred together using ArcView v. 3.2 and 2000 Florida Geographic Data Library (FGDL) layers (FGDL 2000). For Lee County data, in which some of the buildings were surveyed three times in each year, for mapping purposes, I considered a species present if it was observed during any of the three surveys. Abundance Within a single year I defined abundance as the proportion of buildings occupied by a species. I constructed presence-absence matrices for each species to determine the: proportion of buildings on which any gecko occurred; proportion of buildings on which each species occurred, regardless of the presence or absence of the other species; proportion of buildings on which each species occurred and the other species was absent; proportion of buildings on which both species occurred; proportion of buildings on which neither species occurred. If, prior to identification, a gecko retreated to a hiding spot and I was unable to extract it, this further yielded the proportion of buildings on which I was unable to identify all geckos seen. In Charlotte County, I collected data during a single visit to each building in each of three years. I compared proportional differences in abundance between species within each year using 95% confidence intervals determined using the Wilson method (Wilson 1927, Agresti and Coull 1998). 44

PAGE 45

For Lee County data, I used Program Presence version 2 (Hines and MacKenzie 2006) with presence / absence data from the within year repeated surveys to estimate the detection probability which was then used to adjust building occupancy estimates. I used a one-species, one-year model which is based on the model of Mackenzie et al. (2002) and is further described by Hines and MacKenzie (2006). This model consists of the following variables: i = probability that a species is present at site i p it = probability that a species will be detected at site i at time t, given it is present N = number of surveyed sites T = number of sampling occasions n t = number of sites where the species was detected at time t n. = number of sites at which the species was detected at least once Within Program Presence, these variables were used to derive maximum likelihood estimates of and p. Standard errors were estimated using a nonparametric bootstrap method (Buckland and Garthwaite 1991). Model assumptions were: the probability of species occupancy was homogeneous between sites (i.e. sites were similar such that all had an equal probability of being occupied); the species occupancy state of each site remained the same during each year (i.e. no colonization or extinction between surveys within a year) and; a species is never falsely detected when it is absent. For this model, the notation I use consists of the parameter symbol followed in parentheses by the factor for that parameter where (.) denotes a time-constant and (t) a survey-specific parameter. For each species and each year, I included in the candidate model set one model in which detection probabilities were constant among surveys, p(.), and one in which they were survey-specific, p(t). Consistent with model assumptions, in both models the occupancy probability, was held constant, (.). To test how well the models were supported by the data, or goodness-of-fit, within Program Presence, I calculated the variance inflation factor, ^c 45

PAGE 46

(Burnham et al. 1987, Lebreton et al. 1992), of the most parameterized, survey-specific model. I calculated (Equation 3-1) as the quotient of the Pearsons chi-square test statistic derived from the observed data and the chi-square test statistic derived from the mean of 10,000 parametric bootstrap simulations (MacKenzie and Bailey 2004, MacKenzie et al. 2006, Pradel et al. 2003, 2005, White et al. 2001). ^c 22^bootstrapobservedc (3-1) To derive the chi-square test statistic, I calculated observed and expected probabilities of obtaining each possible capture history combination over the three surveys (i.e. 100, 010, 001, 110, 011, 101, 111, 000). I followed Cooch and White (2006) in assuming that if the most parameterized model fit the data, then the less parameterized model, which is a simplified version constrained such that detection probability is constant, would also fit. Values of >1 indicated overdispersion, i.e. the actual sampling variance is greater than the estimated variance, and were used to adjust model selection statistics and standard errors. Parameter estimates did not require adjustment as they are generally considered to be robust even for overdispersed data (Burnham and Anderson 2002). I followed Cooch and White (2006) in assuming that values of = 1, indicating a perfect model fit, or of < 1 did not require similar adjustments. ^c ^c ^c I chose the most parsimonious model from the candidate model set using the Akaike Information Criterion (AIC) calculated in Program Presence (Equation 3-2). AIC = -2 ln(likelihood) + 2K (3-2) where ln(likelihood) is the maximized value of the log likelihood given the data and K is the number of model parameters (Burnham and Anderson 2002). The AIC value is based on the model likelihood but penalizes more parameterized models (Akaike 1973, Burnham and 46

PAGE 47

Anderson 1992, 1998, 2002, MacKenzie et al. 2006, Williams et al. 2002). AIC differences among models within a data set describe the relative support for that model compared to the most parsimonious model. Models with lower AIC values better represent the data compared to other models in that candidate set. Models with AIC values that differ by less than two would be considered similar in support to models in that data set (Burnham and Anderson 2002). AIC values are only relevant to a single data set to compare the models based on that data set. AIC values can not be compared among data sets. If the most parameterized model had a >1, indicating overdispersion and thus a poor variance estimate, I adjusted AIC and standard error (se) values using the quasi-likelihood based adjustment (Equation 3-4) and used the adjusted AIC value, QAIC, as a basis for model selection (Anderson et al. 1994, Burnham and Anderson 1998, 2002, Burnham et al. 1987, MacKenzie and Bailey 2004). ^c QAIC = Kclikelihood2)ln(*2^ (3-4) in which K includes the estimated parameter in addition to the number of model parameters used to calculate the original AIC value. I also calculated the adjusted SE, QSE, for occupancy estimates (Equation 3-5) and recalculated normalized Akaike model weights (Equation 3-6). ^c QSE = ^)(cSE (3-5) RrriiQAICQAICQw1)5.exp()5.exp( (3-6) 47

PAGE 48

Along with the AIC value, I also used normalized Akaike model weights, w, calculated for each model, i (Equation 3-3), to determine the probability that a particular model was the best of the candidate models for the data. RrriiAICAICw1)5.exp()5.exp( (3-3) To account for uncertainty in model selection, I used normalized AIC model weights to calculate model averaged estimates and SEs for and p (Burnham and Anderson 2002). Using the model averaged occupancy probability estimates I compared proportional differences between species within each year using 95% confidence intervals determined using the Wilson method (Wilson 1927, Agresti and Coull 1998). Abundance changes over time To assess changes in distribution and abundance over time, for each county I compared occupancy data among years by: comparing the proportion of buildings occupied in a year versus the subsequent year(s), and; analyzing the building colonization and extinction rates for each species in each county. I defined colonization as a change between years in the occupancy state from absent to present and extinction from present to absent. I compared proportional differences in seasonal occupancy level within each species and county using 95% confidence intervals determined using the Wilson method (Wilson 1927, Agresti and Coull 1998). For Charlotte County, I estimated colonization and extinction rates based on a single survey in each year whereas for Lee County, I calculated estimates from three surveys each year using Program Presence v. 2 in which the model of MacKenzie et al (2002) is extended by MacKenzie et al (2003) by adding two variables: [t] = the probability a species becomes locally extinct from year t to t+1 [t] = the probability a species colonizes a building from year t to t+1 48

PAGE 49

In addition to the assumptions of MacKenzie et al. (2002) this model further assumes that species colonization and extinction probabilities were also homogeneous between sites, and a site does not go extinct and then become recolonized between years. To select the model that best fit the data I used AIC to compare models in which the detection probabilities were constant between surveys within each year and constant between years, p(..), detection probabilities were constant between surveys within each year but varied between years p(.years), and detection probabilities varied between surveys within each year and varied between years, p(survey x year). Species Interactions For both counties, I calculated the proportion of buildings on which H. garnotii and H. mabouia co-occurred each year. I calculated Charlotte County proportions directly from single-survey presence/ absence data. I compared differences in co-occurrence proportions among years within each county using 95% confidence intervals determined using the Wilson method (Wilson 1927, Agresti and Coull 1998). I estimated Lee County proportions from repeated-survey data using Program Presence v.2 to estimate the proportion of Lee County buildings on which the two species co-occurred. I used a two-species, one-year model in which the model of MacKenzie et al. (2002) was extended by MacKenzie et al. (2004). Instead of only a single occupancy probability, this model differs by the inclusion of two occupancy probability parameters, one for each species. G = probability that a building is occupied by H. garnotii, regardless of the presence or absence of H. mabouia, M = probability that a building is occupied by H. mabouia, regardless of the presence or absence of H. garnotii, 49

PAGE 50

Additionally, an occupancy probability parameter for both species together is also included: GM = probability that a building is occupied by both species, Further, this model also conditions the probability of detection on a second species occurrence or absence: p G = probability of detecting H. garnotii, given that H. mabouia is not present, p M = probability of detecting H. mabouia, given that H. garnotii is not present, r Gm = probability of detecting H. garnotii but not H. mabouia, given that both are present, r gM = probability of detecting H. mabouia but not H. garnotii, given that both are present, r GM = probability of detecting both species, given that both species are present r gm = probability of detecting neither species, given that both species are present. For all two-species models I assumed that detection probabilities were equal (i.e. constant) among surveys and that the species were detected independently when both were present. In addition to estimating co-occurrence proportions, I also investigated yearly species co-occurrence patterns by addressing two questions. First, does the occurrence of H. mabouia reduce the detection of H. garnotii or vice-versa? Such a bias could be important when evaluating if H. mabouia has displaced H. garnotii from a building or region. By comparing a full model to one which is constrained such that p G = r Gm I investigated if the detection of H. garnotii was affected when H. mabouia also occurred on the same building (MacKenzie et al. 2004, 2006). Second, do these species co-occur less often than would be expected by chance? Nonrandom patterns in the species co-occurrence matrix may indicate that the occurrence of H. mabouia in some way affects the distribution of H. garnotii (Diamond 1975, Brown et al. 2002). To assess this question, I used a reparameterized model that incorporated an additional parameter (Equation 3-7). = GM / ( G M) (3-7) 50

PAGE 51

The parameter is described by MacKenzie et al. (2004, 2006) as a species interaction factor and is based on the statistical concept that if H. garnotii and H. mabouia co-occur independently then the product of G and M should equal GM (Equation 3-8). Equivalently, the ratio of GM to the product of G and M should equal 1. GM = G M (3-8) Values of equal to one indicate the species co-occur independently while values less than one indicate a nonrandom pattern and more than one that they co-occur more frequently than would be expected by chance. To directly estimate the model is reparameterized (Equation 3-9). GM = G M (3-9) This enables the comparison of a full model in which G M and are all estimated to a model which is constrained such that is fixed to 1 thereby only estimating G and M (Equation 3-10). GM = G M 1 = G M (3-10) Number of Geckos Per Building For each survey year I calculated the mean number of geckos per species on occupied buildings (i.e. containing one or more geckos of a species). For Lee County, I calculated the mean as the maximum number of geckos of each species seen during any of the 3 surveys on each building divided by total number of occupied buildings. To compare Lee County repeated surveys to Charlotte County single survey abundances, for Lee County I also calculated the mean number of geckos of each species seen during each of the 3 surveys each year. Results Charlotte County Results for all Charlotte County buildings surveyed are listed in Appendix B Table B-1 and summarized in Table 3-1 and Figure 3-2. I surveyed 148 buildings once in 2003 (Figure 351

PAGE 52

3), 139 of the original 148 buildings once in 2004 (Figure 3-4), and 130 of the original 148 buildings once in 2006 (Figure 3-5). Missing observations were due to buildings being substantially altered by renovation, demolished, destroyed by Hurricane Charley in August 2004, and other factors making once accessible buildings inaccessible. Single year occupancy estimates The proportion of buildings on which I observed any geckos did not differ from 2003 (99/148 = 0.67; 95% CI: 0.59 0.74) to 2004 (86/130 = 0.62; 95% CI: 0.54-0.70), from 2004 to 2006 (80/130 = 0.62; 95% CI: 0.53 0.69) or overall from 2003 to 2006. The proportion of buildings on which I observed H. garnotii was substantially greater than the proportion on which I observed H. mabouia in 2003 (H. garnotii = 81/148 = 0.55, 95% CI: 0.47 0.63; H. mabouia = 27/148 = 0.18, 95% CI: 0.13 0.25) and 2004 (H. garnotii = 68/139 = 0.49, 95% CI: 0.41 0.57; H. mabouia = 25/139 = 0.18, 95% CI: 0.12 0.25) but not in 2006 (H. garnotii = 41/130 = 0.32, 95% CI: 0.24 0.40; H. mabouia = 53/130 = 0.41, 95% CI: 0.33 0.49). The proportion of buildings on which I observed H. garnotii (Figure 3-6) was similar in 2003 (81/148 = 0.55, 95% CI: 0.47 0.63) and 2004 (68/139 = 0.49, 95% CI: 0.41 0.57). However, from 2004 to 2006 (41/130 = 0.32, 95% CI: 0.24 0.40) there was a 17% decrease. Overall from 2003 to 2006 there was a 23% decrease. The proportion of buildings on which I observed H. mabouia (Figure 3-6) in 2003 (27/148 = 0.18, 95% CI: 0.13 0.25) and in 2004 (25/139 = 0.18, 95% CI: 0.12 0.25) were similar. However, from 2004 to 2006 (53/130 = 0.41, 95% CI: 0.33 0.49) there was a 23% increase. Colonization and extinction From 2003 to 2004, the colonization rate (i.e. the number of buildings unoccupied in 2003 that were occupied in 2004) for H. garnotii (16/63 = 0.25, 95% CI: 0.16 0.37) was 52

PAGE 53

substantially greater than for H. mabouia (10/113 = 0.09, 95% CI: 0.05 0.16) whereas from 2004 to 2006, it was substantially greater for H. mabouia (34/104 = 0.33, 95% CI: 0.24 0.42) than for H. garnotii (12/62 = 0.19, 95% CI: 0.11 0.31) (Table 3-2, Figure 3-6). Overall in Charlotte County from 2003 to 2006 the colonization rate for H. mabouia (34/111 = 0.31, 95% CI: 0.23 0.40) was not substantially different than for H. garnotii (13/60 = 0.22, 95% CI: 0.13 0.34). For 2003 to 2004, the extinction rate (i.e. the number of buildings occupied in 2003 that were unoccupied in 2004) for H. garnotii was 0.32 (24/76, 95% CI: 0.22 0.43) and for H. mabouia was 0.42 (11/26 = 0.42, 95% CI: 0.26 0.61) (Table 3-3, Figure 3-6). However, on four of the 24 buildings from which I recorded H. garnotii as extinct and seven of the 11 buildings from which I recorded H. mabouia as extinct, it was again observed in 2006. If I make the assumption that these observations were not recolonizations but instead were missed observations, then the adjusted proportion of buildings from which H. garnotii went extinct from 2003 to 2004 was 0.26 (20/76, 95% CI: 0.18 0.37) and for H. mabouia was 0.15 (4/26, 95% CI: 0.06 0.34). While the extinction rates for both species were similar from 2003 to 2004, the extinction rate was substantially greater for H. garnotii from 2004 to 2006 (H. garnotii: 35/64 = 0.55, 95% CI: 0.43 0.66; H. mabouia: 3/22 = 0.14, 95% CI: 0.05 0.33) and overall from 2003 to 2006 (H. garnotii: 42/70 = 0.60, 95% CI: 0.48 0.71; H. mabouia: 1/20 = 0.05, 95% CI: 0.01 0.24) (Figure 3-7). In 2006 I was unable to survey six buildings that had contained H. garnotii in 2003 so these were not included in the extinction count for this period. For H. garnotii, the colonization rate from 2003 to 2004 (16/63 = 0.25, 95% CI: 0.16 0.37) was similar to that from 2004 to 2006 (12/62 = 0.19, 95% CI: 0.11 0.31) while for H. mabouia it was substantially higher from 2004 to 2006 (34/104 = 0.33, 95% CI: 0.24 0.42) 53

PAGE 54

than from 2003 to 2004 (10/113 = 0.09, 95% CI: 0.05 0.16). The extinction rate for H. garnotii was notably higher from 2004 to 2006 (35/64 = 0.55, 95% CI: 0.43 0.66) than from 2003 to 2004 (20/76 = 0.26, 95% CI: 0.18 0.37) while for H. mabouia was considerably lower from 2004 to 2006 (3/22 = 0.14, 95% CI: 0.05 0.33) than from 2003 to 2004 (11/26 = 0.42, 95% CI: 0.26 0.61). From 2003 to 2004, H. garnotii appeared to go extinct on 24 buildings. However, in 2006, I again found H. garnotii on four of these buildings. Assuming H. garnotii really did not go extinct from these four buildings but was instead missed in 2004, H. garnotii went extinct from 20 buildings. During this same period, H. mabouia colonized 5% (1/20, 95% CI: 1% 24%) of these buildings. From 2004 to 2006, H. mabouia colonized 17% (6/35, 95% CI: 8% 33%) of the buildings from which H. garnotii went extinct during this same period. Overall from 2003 to 2006, H. mabouia colonized 17% (7/42, 95% CI: 8% 31%) of the buildings from which H. garnotii went extinct in this same period. From 2003 to 2004, H. mabouia appeared to go extinct from 11 buildings. However, in 2006, I again found H. mabouia on seven of these buildings. Again, assuming H. mabouia really didnt go extinct from these seven buildings but was instead missed in 2004, H. mabouia went extinct from four buildings. During this same period, H. garnotii colonized 25% (1/4, 95% CI: 5% 70% ) of these four buildings. From 2004 to 2006, H. garnotii colonized 33% (1/3: 95% CI: 6% 79%) of the buildings from which H. mabouia went extinct during this same period. Overall from 2003 to 2006, H. mabouia went extinct from one building and this building was colonized by H. garnotii during this same period. In 2006 I was unable to survey six buildings that had contained H. mabouia in 2003 so these were not included in the extinction count for this period. 54

PAGE 55

Of the buildings colonized by H. garnotii in 2004, 81% (13/16, 95% CI: 57% 93%) were unoccupied in 2003, the previous survey year, and 19% (3/16, 7% 43%) were occupied by H. mabouia. Of those colonized in 2006, 75% (9/12, 95% CI: 47% 91%) were unoccupied in 2004 and 25% (3/12, 9% 53%) were occupied by H. mabouia. Overall from 2003 to 2006, of the buildings colonized by H. garnotii in 2006, 77% (10/13, 95% CI: 50% 92%) were unoccupied in 2003 and 23% (3/13, 8% 50%) were occupied by H. mabouia. Of the buildings colonized by H. mabouia in 2004, 40% (4/10; 95% CI: 17% 69%) were unoccupied in 2003, the previous survey year, and 60% (6/10, 95% CI: 31% 83%) were occupied by H. garnotii. Of those colonized in 2006, 62% (21/34, 95% CI: 45% 76%) were unoccupied in 2004 and 38% (13/34; 95% CI: 24% 55%) were occupied by H. garnotii. Overall from 2003 to 2006, of the buildings colonized by H. mabouia in 2006, 41% (14/34, 95% CI: 26% 58%) were unoccupied and 59% (20/34, 95% CI: 42% 74%) were occupied by H. garnotii. Species co-occurrence The proportion of buildings on which H. garnotii and H. mabouia co-occurred (Table 3-1) remained similar from 2003 (10/148 = 0.07, 95% CI: 0.04 0.12) to 2004 (11/139 = 0.08, 95% CI: 0.04 0.14) to 2006 (16/130 = 0.12, 95% CI: 0.08 0.19). Of the buildings on which H. garnotii and H. mabouia co-occurred in 2003, by 2004 H. garnotii went extinct from 50% (5/10, 95% CI: 24% 76%) and H. mabouia from 30% (3/10, 95% CI: 11% 60%). From 2004 to 2006 H. garnotii went extinct from 55% (6/11, 95% CI: 28% 79%) of the buildings on which they co-occurred and H. mabouia went extinct from 18% (2/11; 95% CI: 5% 48%). In 2006, I did not resurvey three buildings on which they co-occurred in 2003. Overall from 2003 to 2006 H. garnotii went extinct from 57% (4/7, 95% CI: 25% 84%) of the buildings on which they co55

PAGE 56

occurred and H. mabouia from 0 of 7. However, although H. garnotii went extinct from 57% (4/7, 95% CI: 25% 84%) of the buildings on which it co-occurred with H. mabouia, this was similar to the proportion of the buildings on which it went extinct when in allopatry (38/63 = 60.3%, 95% CI: 48% 71%). Number of geckos per building The estimated mean number of geckos per building for buildings occupied by H. garnotii (Table 3-4) was 2.22 (se = 0.20) for 2003, 2.68 (se = 0.26) for 2004, and 1.68 (se = 0.23) geckos per building for 2006. There was not a significant difference between the number of geckos per building in 2003 and 2004 (t = -1.400, df = 147, p = 0.164) but there was a difference between 2004 and 2006 (t = -2.163, df = 107, p = 0.010). However, overall for H. garnotii there was not a significant difference between the number of geckos per building from 2003 to 2006 (t = -1.635, df = 120, p = 0.105). The estimated mean number of geckos per building occupied by H. mabouia (Table 3-4) was 2.48 (se = 0.38) for 2003, 2.52 (se = 0.39) for 2004, and 3.25 (se = 0.43) geckos per building for 2006. There was not a significant difference between the mean number of geckos per building in 2003 and 2004 (t = -0.073, df = 50, p = 0.942), for 2004 and 2006 (t = -1.072, df = 76, p = 0.287), or overall for 2003 and 2006 (t = -1.167, df = 78, p = 0.247). Within years, there was not a significant difference between the number of geckos per building for H. garnotii and H. mabouia in 2003 (t = -0.625, df = 106, p = 0.534) or in 2004 (t = 0.323, df = 91, p = 0.747). However there was a significant difference for 2006 (t = -2.977, df = 92, p = 0.004). In 2003, H. garnotii and H. mabouia co-occurred on 10 buildings. In 2004, one of these buildings was not resurveyed and in 2006, three were not resurveyed. Of the 9 buildings on which H. garnotii and H. mabouia co-occurred in 2003 that were resurveyed in 2004, H. garnotii 56

PAGE 57

decreased in number or went extinct on 78% (7/9, 95% CI: 45% 94%) while H. mabouia on only 11% (1/9; 95% CI: 2% 44%). Conversely, H. mabouia increased in number on 33% (3/9; 95% CI: 12% 65%) while H. garnotii on 11% (1/9; 95% CI: 2% 44%). Of the seven buildings on which H. garnotii and H. mabouia co-occurred in 2003 that were resurveyed in 2006, H. garnotii decreased in number or went extinct (5/7 = 71%, 95% CI: 36% 92%) on substantially more buildings than H. mabouia (1/7 = 14%, 3% 51%). Conversely, H. mabouia increased in number (5/7 = 71%, 95% CI: 36% 92%) on substantially more buildings than H. garnotii (0/7 = 0%, 95% CI: 0% 35%). However, on buildings on which each species occurred alone in 2003 that were resurveyed in 2006, H. garnotii also decreased in number or went extinct (48/63 = 76%, 95% CI: 64% 85%) on substantially more buildings than H. mabouia (5/13 = 38%, 95% CI: 18% 64%). Lee County In Lee County, I surveyed 176 buildings in 2004 (Figure 3-8), 105 of which I surveyed once and 71 (Figure 3-9) I surveyed three times each. In 2006, I resurveyed 68 of the 71 buildings three times (Figure 3-10). The other three buildings were demolished (Appendix B Table B-2). In Figures 3-8 3-10, if a site was surveyed more than once, a species was considered present if it was observed in any of the three surveys. Relative proportions listed in Table 3-5 and Figure 3-11 are unadjusted nave estimates and reflect the number of sites where a species was or was not detected divided by the total number of sites surveyed. Comparisons of unadjusted occupancy estimates (Table 3-6) when all three repeated surveys were considered versus from survey one alone indicated a substantial improvement in nave estimates due to repeated surveys. In 2004, using data from only survey one, I estimated the proportion of buildings occupied by H. garnotii to be 0.42 (30/71, 95% CI: 0.31 0.54) 57

PAGE 58

whereas when I combine data from all three surveys this estimate was 0.59 (42/71, 95% CI: 0.48 0.70). Similarly, for 2006, using data from only survey one, I estimated the proportion of buildings occupied by H. garnotii to be 0.26 (18/68, 95% CI: 0.17 0.38) whereas when I combine data from all three surveys this estimate was 0.49 (33/68, 95% CI: 0.37 0.60). For H. mabouia, in 2004, using data from only survey one, I estimated the proportion of buildings occupied to be 0.38 (27/71, 95% CI: 0.28 0.50) whereas when I combine data from all three surveys this estimate was 0.56 (40/71, 95% CI: 0.45 0.67). Similarly, for 2006, using data from only survey one, I estimated the proportion of buildings occupied by H. mabouia to be 0.50 (34/68, 95% CI: 0.38 0.62) whereas when I combine data from all three surveys this estimate was 0.68 (46/68, 95% CI: 0.56 0.78). For H. garnotii, when I consider only data from survey one, the nave estimate of the proportion of buildings occupied by H. garnotii calculated using data from all 176 buildings surveyed (70/176 = 0.40, 95% CI: 0.33 0.47) was equal to the estimate when only 71 buildings were considered (30/71 = 0.42, 95% CI: 0.31 0.54). Similarly, for H. mabouia, the estimate using 176 buildings (72/176 = 0.41, 95% CI: 0.34 0.48) was nearly the same as the estimate when only 71 buildings were considered (27/71 = 0.38, 95% CI: 0.28 0.50). Single-year occupancy estimates Nave estimates (Table 3-5), based on the repeated survey buildings only, for the proportion of buildings occupied by any gecko were 0.85 (95% CI: 0.74 0.91) in 2004 and 0.93 (95% CI: 0.84 0.97) in 2006. For those containing H. garnotii, alone or with H. mabouia, estimates were 0.59 (95% CI: 0.48 0.70) in 2004 and 0.49 (95% CI: 0.37 0.60) in 2006 and for H. mabouia, alone or with H. garnotii, were 0.56 (95% CI: 0.45 0.67) in 2004 and 0.68 (95% CI: 0.56 0.78) in 2006. 58

PAGE 59

The candidate model sets used in Program Presence to obtain adjusted occupancy estimates and the model selection statistics obtained for H. garnotii are listed in Table 3-7 and for H. mabouia in Table 3-8. Models are abbreviated using the parameter symbol followed by a variable descriptor in parentheses where (.) and (t) indicate constant or survey-specific probabilities, respectively. Goodness-of-fit test results for each species and year for the most parameterized model, p(t), are presented in Table 3-9. For H. garnotii, the model supported the 2004 data set (comprised of all 176 survey buildings) ( 2 = 7.11, p = 0.529, = 0.882), the 2004 data set (comprised of only the 71 repeated survey buildings) ( ^c 2 = 6.85, p = 0.439, = 0.974), and while somewhat under-dispersed, also supported the 2006 repeated survey data set ( ^c 2 = 3.35, p = 0.854, = 0.479). For H. mabouia, the model showed slight over-dispersion for the 2004 data set (comprised of all 176 survey buildings) ( ^c 2 = 14.51, p = 0.070, = 1.814), slightly more over-dispersion for the 2004 data set (comprised of only the 71 repeated survey buildings) ( ^c 2 = 13.91, p = 0.057, = 1.99), and moderate over-dispersion for the 2006 repeated survey data set ( ^c 2 = 21.42, p = 0.008, = 3.05). H. mabouia AIC values and standard errors were thus adjusted accordingly. ^c In 2004, for both H. garnotii and H. mabouia, the model with constant detection probability {(.)p(.)} ranked highest when all 176 buildings were considered (H. garnotii AIC = 3.70; H. mabouia QAIC = 5.16) and when only the 71 repeated buildings were considered (H. garnotii AIC = 3.78; H. mabouia QAIC = 4.85). For H. garnotii, the occupancy estimate when all 176 buildings were considered (0.587, 95% CI: 0.49 0.69) was similar to when only the 71 repeated buildings were considered (0.609, 95% CI: 0.49 0.73). Similarly, for H. mabouia, the occupancy estimate when all 176 buildings were considered (0.577, 95% CI: 0.45 59

PAGE 60

0.70) was similar to when only the 71 repeated buildings were considered (0.574, 95% CI: 0.41 0.74). The occupancy probability for H. garnotii was similar to that of H. mabouia when all 176 buildings were considered or when only the 71 repeated buildings were included. In 2006, for H. garnotii, the model with survey-specific detection probability, (.)p(t), ranked highest (AIC = 3.11) whereas for H. mabouia, the model with constant detection probability, (.)p(.), ranked highest (QAIC = 4.28). While the occupancy estimate for H. garnotii, 0.502 (95% CI: 0.38 0.63), was substantially lower than for H. mabouia, 0.684 (0.49 0.88), model overdispersion increased the standard error thus making this difference inconclusive. However, when I compare the nave estimates (H. garnotii: 0.49, 95% CI: 0.37 0.60; H. mabouia: 0.68, 95% CI: 0.56 0.78) this difference is more apparent. When I only considered the repeated survey buildings, the occupancy probability was fairly similar for H. garnotii in 2004 (0.609, 95% CI: 0.49 0.73) and in 2006 (0.502, 95% CI: 0.38 0.63). Similarly, the occupancy probability was fairly similar for H. mabouia in 2004 (0.574, 95% CI: 0.41 0.74) as in 2006 (0.684, 95% CI: 0.49 0.88). Colonization and extinction Model statistics and occupancy, colonization, extinction, and detection estimates for the two-year single-species models are listed in Table 3-10 and 3-11. For both H. garnotii and H. mabouia, the model in which detection probability was constant had the best support. However, the model in which detection probability was year-specific was similarly supported (H. garnotii AIC = 1.56; H. mabouia AIC = 1.58). I therefore calculated model-averaged estimates. As anticipated, occupancy estimates were similar to estimates from the one-year, single-species models. The model-averaged estimate for the proportion of buildings colonized in 2006 by H. garnotii (0.088, 95% CI: 0.02 0.34) differed considerably from H. mabouia (0.381, 95% CI: 60

PAGE 61

0.23 0.57). Model-averaged estimates for the proportion of buildings on which H. garnotii went extinct in 2006 (0.259, 95% CI: 0.14 0.42) also differed considerably from H. mabouia (0.069, 95% CI: 0.02 0.23). For both species there was a substantial difference between the proportion of buildings it colonized and from which it went extinct. The nave estimate of the proportion of buildings from which H. garnotii went extinct from 2004 to 2006 that were colonized by H. mabouia in that same period (1/12 = 0.08, 95% CI: 0.01 0.35) was similar to the proportion of buildings from which H. mabouia went extinct from 2004 to 2006 that were colonized by H. garnotii in that same period (zero of three). The nave estimate of the proportion of buildings colonized by H. mabouia in 2006 that were unoccupied in 2004 (6/12 = 0.50) was the same as the proportion of buildings colonized by H. mabouia in 2006 that were occupied by H. garnotii in 2004 (6 of 12 = 0.50). The proportion of buildings colonized by H. garnotii in 2006 that were unoccupied in 2004 was 100% (three of three) while the proportion of buildings colonized by H. garnotii in 2006 that were occupied by H. mabouia in 2004 was zero (zero of three). Of the buildings on which H. garnotii and H. mabouia co-occurred in 2004, by 2006 H. garnotii went extinct from substantially more (10/22 = 46%, 95% CI: 27% 65%) than H. mabouia (3/22 = 14%, 95% CI: 5% 33%). Further, H. garnotii went extinct from a considerably greater proportion of buildings when it co-occurred with H. mabouia than when it was in allopatry (2/20 = 10%, 95% CI: 3% 30%). In 2006, both species continued to co-occur on 41% (9/22, 95% CI: 23% 61%) of the buildings on which they co-occurred in 2004. Detection probability in allopatry versus in sympatry To test if detection probability differed in allopatry than in sympatry, I used a one-year, two-species model in Program Presence. For 2004, the model in which all parameters were 61

PAGE 62

estimated separately was similar in support to both of the constrained models in my candidate model set (Table 3-12) (p G = r Gm : AIC = 0.67; p M = r gM : AIC = 1.81). Likewise, in 2006 these models were also similar in support (p G = r Gm : AIC = 1.64; p M = r gM : AIC = 1.28). This indicates that the detection probability for each species was not affected when the other species also occupied the same building. Co-occurrence patterns I used a one-year, two-species model to test if H. garnotii and H. mabouia co-occurred less often than would be expected by random chance. For 2004 data (Table 3-13) the full model and the constrained model were similar in support (AIC = 1.84) and had a species interaction factor, of 0.965 (se = 0.153) suggesting that the two species were distributed randomly. However, in 2006 the full model was substantially better supported than the constrained model (AIC = 4.13) and had a species interaction factor, of 0.768 (se = 0.069). This implies that the species were no longer randomly distributed and had a decreased likelihood of co-occurring than would be predicted. The proportion of buildings on which H. garnotii and H. mabouia co-occurred was estimated from the full model to be 34.2% in 2004 and 27.0% in 2006. Number of geckos per building I calculated nave estimates based on repeated surveys of the number of geckos per building for each species from repeated survey data in 2004 and 2006. The maximum number of H. garnotii counted on a single building during a single survey was seven in 2004 and six in 2006. The maximum number of H. mabouia counted on a single building during a single survey was 25 in 2004 and 24 in 2006. The number of individuals of H. garnotii on a building was positively correlated to the number of H. mabouia in 2004 (r = 0.386, p = .001, n = 71) and in 2006 (r = 0.955, p < 0.001, n 62

PAGE 63

= 71). In 2004, on the 42 buildings that contained H. garnotii, there was a total of 104 geckos resulting in a mean of 2.48 (se = 0.24) geckos per occupied building. The mean number of geckos for each survey was 2.13 (se= 0.28, n = 30), 2.62 (se = 0.30, n = 29), and 1.94 (se = 0.20, n = 31). In 2006, on the 33 buildings that contained H. garnotii, there was a total of 68 geckos for a mean of 2.06 (se = 0.25) geckos per occupied building. The mean number of geckos for each survey was 1.72 (se = 0.24, n = 18), 1.71 (se = 0.23, n = 21), and 1.86 (se = 0.29, n = 28) In 2004, on the 40 buildings that contained H. mabouia there were 169 total geckos resulting in a mean of 4.23 (se = 0.79) geckos per occupied building. The mean number of geckos for each survey was 4.56 (se = 1.02, n = 27), 4.36 (se = 0.91, n = 33), and 3.73 (se = 0.93, n = 30). In 2006, on the 46 buildings that contained H. mabouia there were a total of 198 geckos for a mean of 4.30 (se = 0.68) geckos per occupied building. The mean number of geckos for each survey was 3.03 (se = 0.50, n = 34), 3.88 (se = 0.68, n = 40), and 4.77 (se = 0.84, n = 34). There was not a significant difference between the 2004 and 2006 means for either species (H. garnotii: t = 1.18, df = 73, p = 0.24; H. mabouia: t = -0.067, df = 84, p = 0.95). There was a significant difference between the 2004 means for H. garnotii and H. mabouia (t = 2.16, df = 80, p = 0.03) as well as the 2006 means (t = 2.69, df = 77, p = 0.01). Of the buildings on which H. garnotii and H. mabouia co-occurred in 2004, by 2006, H. garnotii decreased in number or went extinct on substantially more buildings (14/22 = 64%, 95% CI: 43% 80%) than H. mabouia (6/22 = 27%, 95% CI: 13% 48%). Conversely, H. mabouia increased in number on 55% (12/22, 95% CI: 35% 73%) of the buildings while the number of H. garnotii only increased on 14% (3/22, 95% CI: 5% 33%) of the buildings. Of the buildings on which each species occurred alone in 2004 that were resurveyed in 2006, the number of buildings on which H. garnotii decreased in number or went extinct (14/20 = 70.0%, 95% CI: 63

PAGE 64

40% 85%) was slightly greater than the number of buildings on which H. mabouia decreased in number or went extinct (6/15 = 40%, 95% CI: 20% 64%). However, for H. garnotii, there was not a difference in the number of buildings on which it decreased in abundance or went extinct when it co-occurred with H. mabouia (14/22 = 64%, 95% CI: 43% 80%) versus when it was alone (14/20 = 70.0%, 95% CI: 40% 85%). Discussion Distributions H. garnotii was first reported in Charlotte County in 1991 (Conant and Collins 1991). In 2003, my surveys indicated that it occurred throughout much of the county. In contrast, H. mabouia, first collected in 2001 (Klowden 2002), was primarily limited to the central most urbanized areas with a few isolated records elsewhere in the county including on the southern part of County Road 765 near the Lee County border and in the northwest part of the county (Englewood area). By 2006, the proportion of buildings on which H. mabouia was found in Charlotte County increased considerably (+23%) and I found individuals throughout the county. Though H. garnotii could also still be found throughout the county, it occurred on appreciably fewer buildings (-23%) by 2006, especially in the central part of the county and in the south part of the county near Lee County. H. garnotii was first recorded from Lee County in 1972 (McCoy 1972) while H. mabouia was first recorded in 2000 (Townsend et al. 2002). In 2004, my surveys revealed that both species occurred in a mosaic pattern throughout the county and occurred on buildings at nearly the same proportions (61% and 57%, respectively). This likely indicates that H. mabouia arrived earlier than first reported. In 2006, a similar type of mosaic pattern persisted though the proportion of buildings on which H. garnotii was found had decreased (-11%) while for H. mabouia had increased (+11%). 64

PAGE 65

The use of repeated surveys to assess presence/ absence is important. Comparisons of Lee County single survey to repeated survey data increased nave occupancy estimates by more than 17%. This indicates that detection probability for these species was substantially less than one. In Charlotte County where I only surveyed each building one time each year, occupancy estimates are likely biased low. Despite this, within Charlotte County, since methodology was similar among years, between year comparisons are still useful for evaluating trends. I chose not to use Lee County model estimated detection probabilities to adjust Charlotte County single survey data since detection probabilities may vary between survey areas due to differences in abundances or other site specific attributes. An increase in the number of surveys per building improved estimates substantially. However, an increase in the number of buildings surveyed was less important. Comparison of single survey estimates for 176 versus 71 buildings did not show a difference in occupancy estimates for either species. This suggests that rather than searching a large number of buildings once, increased effort should be allocated towards an increased number of surveys of each building. It appears that three surveys were sufficient since with three surveys, the nave estimates were similar to the model estimates. Colonization and Extinction In Charlotte County, I estimated the colonization rate for H. garnotii to be 25% from 2003 to 2004, 19% from 2004 to 2006, and overall from 2003 to 2006 to be 22%. The extinction rate was 32% from 2003 to 2004, 55% from 2004 to 2006 and overall from 2003 to 2006 was 60%. The colonization and extinction rates were fairly similar for 2003 to 2004, but from 2004 to 2006 and overall from 2003 to 2006, the estimated colonization rate was much lower than the extinction rate. For H. mabouia the colonization rate was 9% from 2003 to 2004, 33% from 65

PAGE 66

2004 to 2006, and overall from 2003 to 2006 was 31%. The extinction rate was 42% from 2003 to 2004, 14% from 2004 to 2006, and overall from 2003 to 2006 was 5%. From 2003 to 2004, the colonization rate for H. mabouia was very low and the extinction rate substantially higher. As this appears to have been early in its invasion of Charlotte County, a low colonization rate would be expected. A high extinction rate may be an artifact of H. mabouia occurring on so few buildings during these years. In contrast, from 2004 to 2006 and overall from 2003 to 2006, the colonization rate for H. mabouia was much higher than the extinction rate. In Lee County, from 2004 to 2006 I estimated the colonization rate for H. garnotii to be 9% and the extinction rate to be 26%. For H. mabouia, I estimated the colonization rate during this same period to be 38% and the extinction rate to be 7%. I made these estimates using repeated survey data with a model that included a detection probability parameter. Though nave colonization estimates, which did not account for detection probability, were slightly lower and extinction rates slightly higher than model estimates, for these data, nave estimates were fairly close to model estimates (Tables 3-10 and 3-11). Since Charlotte County estimates did not include detection probability, and were based on single, as opposed to multiple, survey data, they may substantially underestimate colonization and overestimate extinction. Because of this bias, Charlotte County estimates should not be directly compared with Lee County results, however, the overall pattern within Charlotte County and its comparison with that of Lee County is still informative. That is, a similar pattern is seen in both counties of low colonization and high extinction rates for H. garnotii and high colonization and low extinction rates for H. mabouia. In Charlotte County, for all survey years, the ability of H. mabouia to colonize buildings that were occupied by H. garnotii the previous survey year was similar to those that were unoccupied (2004: 60% vs. 40%; 2006: 42% vs. 58%). In contrast, H. garnotii colonized 66

PAGE 67

significantly fewer buildings with H. mabouia than unoccupied buildings (2004: 20% vs. 80%; 2006: 27% vs. 73%). A similar trend was seen in Lee County however, sample sizes were small making this conclusion more tenuous for Lee County. This may indicate a reduced ability of H. garnotii to colonize buildings already occupied by H. mabouia but not vice versa. In Charlotte County from 2003 to 2004, I estimated that 5% of the buildings from which H. garnotii went extinct were colonized by H. mabouia in this same period. From 2004 to 2006, 17% of the buildings and overall from 2003 to 2006, 17% of the buildings from which H. garnotii went extinct were colonized by H. mabouia in this same period. For H. mabouia from 2003 to 2004, I estimated that 25% (one of four) of the buildings on which it went extinct was colonized by H. garnotii in this same period. From 2004 to 2006, 33% (one of three) of the buildings and overall from 2003 to 2006, the one building on which H. mabouia went extinct was colonized by H. garnotii in this same period. In Lee County from 2004 to 2006, of the 12 buildings on which H. garnotii went extinct, one was colonized by H. mabouia in that same period. Of the three buildings on which H. mabouia went extinct, none was colonized by H. garnotii in that same period. In Lee County, from 2004 to 2006, there was not a difference in the number of buildings on which H. garnotii decreased in abundance or went extinct when it co-occurred with H. mabouia (64%, 95% CI: 43% 80%) versus when it was alone (70.0%, 95% CI: 40% 85%). Though a decrease in abundance is not necessarily a precursor to extinction, it may increase the risk of extinction due to a catastrophic event. When I consider the extinction rate alone, H. garnotii went extinct from a considerably greater proportion of buildings when it co-occurred with H. mabouia (46%, 95% CI: 27% 65%) than when it was in allopatry (10%, 95% CI: 3% 67

PAGE 68

68 30%). In Charlotte County, I did not find simila r differences, however sample sizes were very small. While it appears that colonization of a building by H. mabouia is not linked with the immediate extinction of H. garnotii, longer term associations may have different consequences. However, these species seem to be able to co-exi st for at least some period of time as indicated by their continued co-occurrence in 2006 on 41% of the buildings on which they co-occurred in 2004. Longer term sampling is requ ired to determine if they can persist together and under what conditions. Changes in their dist ributions are further highlighted by assessment of co-occurrence patterns. In Lee County model calculated co-occurrence estimat es of 34% in 2004 and 27% in 2006 were fairly similar. However, while in 2004, H. garnotii and H. mabouia co-occurred as often as would be expected by chance, by 2006, they were no longer randomly distributed and had a decreased likelihood of co-occurring than would be predicted. Though colonization by H. mabouia did not appear to be the primary cause of extinction of H. garnotii from a building, on buildings on which th ey co-occured, the extinction rate of H. garnotii was significantly higher than for H. mabouia. Further, the successful colonization of a building by H. garnotii may have been affected by the presence of H. mabouia. As further exhibited by yearly decreases in the occupancy probability for H. garnotii in both counties and simultaneous increases in the occupancy probability for H. mabouia, by the low colonization rate and high extinction rate of H. garnotii as opposed to the high colonization rate and low extinction rate of H. mabouia, and their shift to a non-random distribution pattern, interactions between these species may be causing changes in H. garnotii distribution and abundance which could lead to its displacement. However, while the evidence is comp elling, further study is necessary to conclude causation.

PAGE 69

Table 3-1. Charlotte County: Number and proportion of buildings on which I observed Hemidactylus garnotii or H. mabouia each survey year and the inter-survey change. Absolute count Absolute change Relative proportion a Relative change 2003 2004 2006 2003-4 2004-6 2003-6 2003 2004 2006 2003-4 2004-6 2003-6 Number of buildings surveyed 148 139 130 -9 -9 -18 148 139 130 -0.06 -0.06 -0.12 None 49 50 53 +4 -3 +1 0.33 0.38 0.38 +0.05 0.00 +0.05 (0.26 0.41) (0.30 0.46) (0.31 0.47) Any gecko 99 86 80 -13 -6 -19 0.67 0.62 0.62 -0.05 0.00 -0.05 (0.59 0.74) (0.54 0.70) (0.53 0.69) H. garnotii only 71 57 25 -14 -32 -46 0.48 0.41 0.19 -0.07 -0.22 -0.29 (0.40 0.56) (0.33 0.49) (0.13 0.27) H. mabouia only 17 14 37 -3 +23 +20 0.11 0.10 0.28 -0.01 +0.18 +0.17 (0.07 0.18) (0.06 0.16) (0.21 0.37) H. garnotii and H. mabouia 10 11 16 +1 +5 +6 0.07 0.08 0.12 +0.01 +0.04 +0.05 (0.04 0.12) (0.04 0.14) (0.08 0.19) H. garnotii (alone + with H. mabouia) 81 68 41 -13 -27 -40 0.55 0.49 0.32 -0.06 -0.17 -0.23 (0.47 0.63) (0.41 0.57) (0.24 0.40) H. mabouia (alone + with H. garnotii) 27 25 53 -2 +28 +26 0.18 0.18 0.41 0.00 +0.23 +0.23 (0.13 0.25) (0.12 0.25) (0.33 0.49) Only UI geckos 1 4 2 +3 -2 +1 0.01 0.03 0.02 +0.02 -0.01 +0.01 (0.00 0.04) (0.01 0.07) (0.00 0.05) 69 (a) The 95% confidence interval is in parentheses below each estimate.

PAGE 70

Table 3-2. Charlotte County: Number of buildings colonized by H. garnotii and H. mabouia from 2003 to 2004, 2004 to 2006, and overall from 2003 to 2006. Number of buildings unoccupied a Number of buildings colonized b Proportion of buildings colonized c 2003-4 2004-6 2003-6 2003-4 2004-6 2003-6 2003-4 2004-6 2003-6 H. garnotii 63/139 62/130 60/130 16/63 12/62 13/60 0.254 0.194 0.217 H. mabouia 113/139 104/130 111/130 10/113 34/104 34/111 0.088 0.327 0.306 (a) Number of buildings surveyed in last year of range that were unoccupied in first year of range, (b) Number of buildings unoccupied in first year of range that were occupied in last year of range, (c) Proportion of buildings unoccupied in first year of range that were occupied in last year of range. Table 3-3. Charlotte County: Number of buildings on which H. garnotii and H. mabouia went extinct from 2003 to 2004, 2004 to 2006, and overall from 2003 to 2006. Number of buildings occupied a Number of buildings extinct b Proportion of buildings extinct c 2003-4 2004-6 2003-6 2003-4 2004-6 2003-6 2003-4 2004-6 2003-6 H. garnotii 76/139 64/130 70/130 24/76 35/64 42/70 0.316 0.547 0.600 H. mabouia 26/139 22/130 20/130 11/26 3/22 1/20 0.423 0.136 0.050 70 (a) Number of buildings surveyed in last year of range that were occupied in first year of range, (b) Number of buildings occupied in first year of range that were unoccupied in last year of range, (c) Proportion of buildings occupied in first year of range that were unoccupied in last year of range. Table 3-4. Charlotte County: Mean number of H. garnotii and H. mabouia per occupied building for 2003, 2004, and 2006 surveys. H. garnotii H. mabouia 2003 2004 2006 2003 2004 2006 Number of buildings occupied 81 68 41 27 25 53 Total number of geckos 180 182 69 67 63 172 Mean 2.22 2.68 1.68 2.48 2.52 3.25 Standard error 0.20 0.26 0.23 0.38 0.39 0.43 Standard deviation m 1.84 2.171 1.461 1.97 1.961 3.121 Minimu 1 1 Maximum 12 11 8 9 9 12

PAGE 71

Table 3-5. Lee County: Number and proportion of buildings on which I observed Hemidactylus garnotii and/or H. mabouia each survey and year and the proportional change between years. Absolute count Relative proportion a Proportional change 2004 2006 2004 2006 2004-6 Survey # 1 2 3 All 1 2 3 All 1 2 3 All 1 2 3 All 1 All # of buildings 71 71 71 71 68 68 68 68 surveyed No geckos 21 18 18 11 20 17 15 5 0.30 0.25 0.25 0.15 0.29 0.25 0.22 0.07 -0.01 -0.08 (0.20.41) (0.17-0.37 ) (0.17-0.37 ) (0.09.26) (0.20.41) ( 0.16-0.36) (0.14-0.33 ) (0.03.16) Any geckos 50 53 53 60 48 51 53 63 0.70 0.75 0.75 0.85 0.71 0.75 0.78 0.93 -0.01 +0.08 (0.59.80) ( 0.63-0.83) ( 0.63-0.83) (0.74.91) (0.59.80) ( 0.64-0.84) ( 0.67-0.86) 0.84.97) H garnotii 30 29 31 42 18 21 28 33 0.42 0.41 0.44 0.59 0.26 0.31 0.41 0.49 -0.16 -0.10 (alone + both) (0.31.54) ( 0.30-0.52) ( 0.33-0.55) (0.48.70) (0.17.38) ( 0.21-0.43) ( 0.30-0.53) (0.37.60) H. mabouia 27 33 30 40 34 40 34 46 0.38 0.46 0.42 0.56 0.50 0.59 0.50 0.68 +0.12 +0.12 (alone + both) (0.28.50) ( 0.35-0.58) ( 0.31-0.54) (0.45.67) (0.38.62) ( 0.47-0.70) ( 0.38-0.62) (0.56.78) H. garnotii 8 9 8 22 7 10 9 17 0.11 0.13 0.11 0.31 0.10 0.15 0.13 0.25 -0.01 -0.06 & H. mabouia (0.06.21) ( 0.07-0.22) ( 0.06-0.21) (0.21.42) (0.05.20) ( 0.08-0.25) ( 0.07-0.23) (0.16.36) H. garnotii only 22 20 23 20 11 11 19 16 0.31 0.28 0.32 0.28 0.16 0.16 0.28 0.24 -0.15 -0.04 (0.21.42) ( 0.19-0.40) ( 0.23-0.44) (0.19.40) (0.09.27) ( 0.09-0.27) ( 0.19-0.40) (0.15.35) H. mabouia only 19 24 22 18 27 30 25 29 0.27 0.34 0.31 0.25 0.40 0.44 0.37 0.43 +0.13 +0.18 (0.18.38) ( 0.24-0.45) ( 0.21-0.42) (0.17.37) (0.29.52) ( 0.33-0.56) ( 0.26-0.49) (0.32.54) Unidentified 1 0 0 0 3 2 1 1 0.01 0.00 0.00 0.00 0.04 0.03 0.01 0.02 +0.03 +0.01 geckos only (0.00.08) ( 0.00) ( 0.00) (0.00) (0.02.12) ( 0.01-0.10) ( 0.00-0.08) (0.00.08) 71 (a) The 95% confidence interval is in parentheses below each estimate.

PAGE 72

Table 3-6. Lee County: Comparison of single and repeated surveys to estimate the proportion of buildings occupied by Hemidactylus garnotii and/ or H. mabouia. Absolute count Relative proportion of occupied buildings c 2004 2006 2004 2006 Number of buildings surveyed 176 a 71 b 68 b 176 71 68 Survey number 1 all 1 all 1 all 1 all 1 all 1 all No geckos 53 43 21 11 20 5 0.30 0.24 0.30 0.15 0.29 0.07 1031 (0.24 0.37) (0.19 0.31) (0.20 0.41) (0.09 0.26) (0.20 0.41) (0.03 0.16) Any geckos 123 133 50 60 48 63 0.70 0.76 0.70 0.85 0.71 0.93 (0.63 0.76) (0.69 0.81) (0.59 0.80) (0.74 0.91) (0.59 0.80) 0.84 0.97) H garnotii 70 82 30 42 18 33 0.40 0.47 0.42 0.59 0.26 0.49 (alone + both) (0.33 0.47) (0.39 0.54) (0.31 0.54) (0.48 0.70) (0.17 0.38) (0.37 0.60) H. mabouia 72 85 27 40 34 46 0.41 0.48 0.38 0.56 0.50 0.68 (alone + both) (0.34 0.48) (0.41 0.56) (0.28 0.50) (0.45 0.67) (0.38 0.62) (0.56 0.78) H. garnotii 23 37 8 22 7 17 0.13 0.21 0.11 0.31 0.10 0.25 and H. mabouia (0.09 0.19) (0.16 0.28) (0.06 0.21) (0.21 0.42) (0.05 0.20) (0.16 0.36) H. garnotii only 47 45 22 20 11 16 0.27 0.26 0.31 0.28 0.16 0.24 (0.21 0.34) (0.20 0.32) (0.21 0.42) (0.19 0.40) (0.09 0.27) (0.15 0.35) H. mabouia only 49 48 19 18 27 29 0.28 0.27 0.27 0.25 0.40 0.43 (0.22 0.35) 2 (0.21 0.34) 2 (0.18 0.38) 1 (0.17 0.37) 0 (0.29 0.52) 4 (0.32 0.54) 2 Only unidentified 4 3 0.0 0.0 0.0 0.0 0.0 0.0 geckos (0.01 0.06) (0.01 0.05) (0.00 0.08) (0.00) (0.02 0.12) (0.00 0.08) 72 (a) -1 includes 176 buildings surveyed once in survey 1 only. -all includes 71 buildings surveyed three times and 105 buildings surveyed once. (b) -1 and -1 include only the 1 st of three surveys and -all and -all includes all three surveys. is a resurvey of the same buildings from the previous year minus three buildings. (c) 95% confidence intervals in parentheses below each estimate.

PAGE 73

Table 3-7. Lee County: Hemidactylus garnotii model selection statistics and parameter estimates for one-year, one-species models. Detection probability (p) estimate Occupancy probability () estimate Data Set / Model a AIC b w c K d -2*ln(like) e Survey f ^p SE 95% Confidence Interval Nave g ^ SE 95% Confidence Interval H. garnotii (2004-71) (.)(.)p 0 0.869 2 244.60 All 0.694 0.045 0.61 0.78 0.592 0.609 0.061 0.49 0.73 )((.)tp 3.78 7 0.131 4 244.38 1 0.694 0.073 0.55 0.84 0.592 0.609 0.061 0.49 0.73 2 0.671 0.074 0.53 0.82 3 0.717 0.072 0.58 0.86 H. garnotii (2004-1(.)(.) 6) p 0 0.864 2 384.54 All 0.685 0.044 0.60 0.77 0.466 0.587 0.051 0.49 0.69 )((.)tp 3.70 06) 0.136 4 384.24 1 0.672 0.064 0.55 0.80 0.466 0.592 0.054 0.49 0.70 2 0.671 0.074 0.53 0.82 3 0.717 0.072 0.58 0.86 H. garnotii (20 )((.)tp 0 0.826 4 209.12 1 0.528 0.088 0.36 0.70 0.485 0.502 0.063 0.38 0.63 2 0.616 0.086 0.45 0.79 3 0.821 0.072 0.68 0.96 (.)(.)p 3.11 0.174 2 216.23 All 0.647 0.054 0.54 0.75 0.485 0.508 0.064 0.38 0.63 73 (a) For 2004, I considered 2 data sets; 2004-176 comprised 176 points of which 105 were surveyed once and 71 were surveyed three times, 2004-71 only comprised the 71 sites surveyed three times; = probability that a species is present at a site, p= probability that a species will be detected at a site given it is present, (.) = probability is constant, (t) = probability is survey-specific, (b) difference in AIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model, (e) like is the maximum likelihood of each model, (f) All indicates that detection probability was constant making estimates equivalent for all three surveys,(g) nave occupancy estimate = number of sites at which species was seen / total number of sites.

PAGE 74

Table 3-8. Lee County: Hemidactylus mabouia model selection statistics and parameter estimates for one-year, one-species models. Detection probability (p) estimate Occupancy probability () estimate Data Set / Model a QAIC b w c K d -2*ln(like) e Survey f ^p SE g 95% Confidence Interval f Nave h ^ SE g 95% Confidence Interval f H. mabouia (2004-71) (.)(.)p 0.00 0.919 3 230.88 All 0.736 0.097 0.55 0.93 0.563 0.574 0.085 0.41 0.74 )((.)tp 4.85 0.081 5 228.58 1 0.664 0.169 0.33 1.00 0.563 0.573 0.085 0.41 0.74 2 0.811 0.143 0.53 1.00 3 0.738 0.159 0.43 1.00 H. mabouia (2004-176) (.)(.) p 0.00 0.929 3 374.30 All 0.737 0.092 0.56 0.92 0.483 0.577 0.064 0.45 0.70 )((.)tp 5.16 0.071 5 372.40 1 0.690 0.134 0.43 0.95 0.483 0.593 0.067 0.46 0.73 2 0.811 0.143 0.53 1.00 3 0.738 0.159 0.43 1.00 H. mabouia (2006) (.)(.)p 0.00 0.895 3 229.11 All 0.774 0.084 0.61 0.94 0.676 0.684 0.100 0.49 0.88 )((.)tp 4.28 0.105 5 225.97 1 0.732 0.148 0.44 1.00 0.676 0.683 0.100 0.49 0.88 2 0.861 0.117 0.63 1.00 3 0.732 0.148 0.44 1.00 (c) Akaike model weight adjusted for overdispersion, (d) number of parameters in each model, including, (e) like is the maximum likelihood of each model, (f) All indicates that detection probability was constant making estimates equivalent for all three surveys, (g) standard errors, SE, and confidence intervals have been adjusted for overdispersion, (h) nave occupancy estimate = number of sites at which species was seen / total number of sites. ^c (a) For 2004, I considered 2 data sets; 2004-176 comprised 176 points of which 105 were surveyed once and 71 were surveyed three times, 2004-71 only comprised the 71 sites surveyed three times; = probability that a species is present at a site, p= probability that a species will be detected at a site given it is present, (.) = probability is constant, (t) = probability is survey-specific, (b) QAIC is the AIC value adjusted for overdispersion, QAIC = difference in QAIC value from the highest ranking model, 74

PAGE 75

75 Table 3-9. Lee County: Goodness-of-fit test results as calculated in Program Presence (V. 2.0) for the one-species, one-year model in which detection probability was surveyspecific. Data Set a 2observed2bootstrap^c b c p-value d e H. garnotii (2004-176) 7.11 8.06 0.529 0.88 H. garnotii (2004-71) 6.85 7.03 0.439 0.97 H. garnotii (2006) 3.35 6.99 0.854 0.48 H. mabouia (2004-176) 14.51 8.00 0.070 1.81 H. mabouia (2004-71) 13.91 6.98 0.057 1.99 H. mabouia (2006) 21.42 7.02 0.008 3.05 (a) For 2004, I considered 2 data sets; 2004-176 comprised 176 points of which 105 were surveyed once and 71 were surveyed three times. 2004-71 only comprised the 71 sites surveyed three times, (b) test statistics derived from the observed survey data and the expected probabilities of obtaining each possible capture history combination, (c) Mean of thetest statistics obtained from 10,000 bootstrap simulations, (d) the probability of observing a test statistic greater than or equal to the one observed from 10,000 parametric bootstraps, (e) the variance inflation factor, c, where: 22^22^/bootstrapobservedc

PAGE 76

Table 3-10. Lee County: Hemidactylus garnotii model selection statistics and parameter estimates for two-year, one-species models for 2004 and 2006. Model a AIC b w c K d ^2004e ^2006e ^2006e ^12004pe ^22004pe ^32004pe ^12006pe ^22006pe ^32006pe p(..) 0 0.560 4 0.613 0.085 0.260 0.674 0.674 0.674 0.674 0.674 0.674 (0.49-0.72) (0.02-0.35) (0.14-0.42) (0.60-0.74) (0.60-0.74) (0.60-0.74) (0.60-0.74) (0.60-0.74) (0.60-0.74) p(.year) 1.56 0.257 5 0.609 0.093 0.253 0.694 0.694 0.694 0.647 0.647 0.647 (0.49-0.72) (0.02-0.34) (0.14-0.42) (0.60-0.77) (0.60-0.77) (0.60-0.77) (0.53-0.75) (0.53-0.75) (0.53-0.75) p(survey x year) 2.23 0.184 9 0.609 0.092 0.262 0.694 0.671 0.717 0.528 0.616 0.821 (0.49-0.72) (0.02-0.34) (0.14-0.43) (0.54-0.82) (0.51-0.80) (0.56-0.84) (0.36-0.69) (0.44-0.77) (0.64-0.92) Model-averaged estimates 0.088 0.259 (0.02-0.34) (0.14-0.42) Nave estimates 0.044 0.176 (a) = probability that a species is present at a site, = probability a species colonizes a building in 2006, = probability a species becomes locally extinct in 2006, p = probability that a species will be detected at a site given it is present, (.) = probability is constant within and between years, (.year) = probability is constant within each year but can vary between years, (survey year) = probability can vary between surveys within each year and between years, (b) difference in AIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model, (e) Years are designated by the subscripts 2004 and 2006. Within each year, survey-specific detection estimates are further designated by 1, 2, and 3. The 95% confidence interval is in parentheses below each estimate. 76

PAGE 77

Table 3-11. Lee County: Hemidactylus mabouia model selection statistics and parameter estimates for two-year, one-species models for 2004 and 2006. Model a AIC b w c K d ^2004e ^2006e ^2006e ^12004pe ^22004pe ^32004pe ^12006pe ^22006pe ^32006pe p(..) 0 0.633 4 0.572 0.382 0.068 0.756 0.756 0.756 0.756 0.756 0.756 (0.45-0.68) (0.23-0.57) (0.02-0.23) (0.70-0.81) (0.70-0.81) (0.70-0.81) (0.70-0.81) (0.70-0.81) (0.70-0.81) p(.year) 1.58 0.287 5 0.574 0.378 0.070 0.736 0.736 0.736 0.774 0.774 0.774 (0.45-0.69) (0.22-0.57) (0.02-0.23) (0.64-0.81) (0.64-0.81) (0.64-0.81) (0.69-0.84) (0.69-0.84) (0.69-0.84) p(survey x year) 4.13 0.080 9 0.573 0.379 0.072 0.664 0.811 0.738 0.732 0.861 0.732 (0.45-0.68) (0.22-0.56) (0.02-0.23) (0.50-0.79) (0.65-0.91) (0.58-0.85) (0.58-0.84) (0.72-0.94) (0.59-0.84) Model-averaged estimates 0.381 0.069 (0.23-0.57) (0.02-0.23) Nave estimates 0.176 0.044 (a) = probability that a species is present at a site, = probability a species colonizes a building in 2006, = probability a species becomes locally extinct in 2006, p = probability that a species will be detected at a site given it is present, (.) = probability is constant within and between years, (.year) = probability is constant within each year but can vary between years, (survey year) = probability can vary between surveys within each year and between years, (b) difference in AIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model, (e) Years are designated by the subscripts 2004 and 2006. Within each year, survey-specific detection estimates are further designated by 1, 2, and 3. The 95% confidence interval is in parentheses below each estimate. 77

PAGE 78

Table 3-12. Lee County: 2004 and 2006 two-species, one-year model selection statistics for H. garnotii and H. mabouia evaluating if the occurrence of one species on a building affects the detection of the other species on that building. Model a AIC b w c K d -2*ln(like) e 2004 Constrained (p G =r Gm ) 0.00 0.4988 7 462.61 Full 0.67 0.3568 8 461.28 Constrained (p M =r gM ) 2.48 0.1444 7 465.09 2006 Constrained (p G =r Gm ) 0.00 0.598 7 434.00 Full 1.64 0.263 8 433.64 Constrained (p M =r gM ) 2.92 0.139 7 436.92 (a) Full describes the model G M p G p M r Gm r gM r GM in which all of the parameters are estimated separately. Constrained describes a similar model with the added constraint following in parentheses ( G = probability that a building is occupied by H. garnotii, regardless of the presence or absence of H. mabouia, M = probability that a building is occupied by H. mabouia, regardless of the presence or absence of H. garnotii, = the likelihood of the species co-occurring on a building compared to co-occurrence by chance, p G = probability of detecting H. garnotii, given that H. mabouia is not present, p M = probability of detecting species H. mabouia, given that H. garnotii is not present, r Gm = probability of detecting H. garnotii but not H. mabouia, given that both are present, r gM = probability of detecting H. mabouia but not H. garnotii, given that both are present, r GM = probability of detecting both species, given that both species are present), (b) AIC = difference in AIC value from the highest ranking model, (c) Akaike model weight, 78 (d) Number of parameters in each model, (e) ln(like) is the natural log of the maximum likelihood of the model.

PAGE 79

Table 3-13. Lee County: 2004 and 2006 two-species, one-year model selection statistics evaluating if H. garnotii and H. mabouia co-occurred less often than would have been expected by chance. Model a AIC b w c K d -2*ln(like) e 2004 Constrained ( = 1) 0.00 0.715 7 461.44 1 Full 1.84 0.285 8 461.28 0.965(0.153) 2006 Full 0.00 0.8875 8 433.64 0.768(0.069) Constrained ( = 1) 4.13 0.1125 7 439.77 1 (a) Full describes the model G M p G p M r Gm r gM r GM in which all of the parameters are estimated separately. Constrained describes a similar model with the added constraint of = 1 ( G = probability that a building is occupied by H. garnotii, regardless of the presence or absence of H. mabouia, M = probability that a building is occupied by H. mabouia, regardless of the presence or absence of H. garnotii, = the likelihood of the species co-occurring on a building compared to co-occurrence by chance, p G = probability of detecting H. garnotii, given that H. mabouia is not present, p M = probability of detecting species H. mabouia, given that H. garnotii is not present, r Gm = probability of detecting H. garnotii but not H. mabouia, given that both are present, r gM = probability of detecting H. mabouia but not H. garnotii, given that both are present, r GM = probability of detecting both species, given that both species are present), (b) AIC = difference in AIC value from the highest ranking model, (c) Akaike model weight, 79 (d) Number of parameters in each model, (e) ln(like) is the natural log of the maximum likelihood of the model.

PAGE 80

80 Figure 3-1. County, state and federal roads surveyed in Charlotte and Lee Counties Florida.

PAGE 81

.7 .669 .6 .600 .590 .547 .5 Proportion of Buildings .489 .4 2003 .408 .385 .381 2004 .331 .3 2006 .315 .2 .182 .180 .1 .123 .079 .068 .0 81 None A ny gecko H. garnotii H. mabouia H. g arnotii & H. mabouia (alone+both) (alone+both) Gecko Species Present Figure 3-2. Charlotte County: Number and proportion of buildings on which I observed Hemidactylus garnotii and/or H. mabouia each survey year.

PAGE 82

82 Figure 3-3. Charlotte County: Gecko species captured at each survey building in 2003.

PAGE 83

83 Figure 3-4. Charlotte County: Gecko species captured at each survey building in 2004.

PAGE 84

Figure 3-5. Charlotte County: Gecko species captured at each survey building in 2006. 84

PAGE 85

00.10.20.30.40.50.6200320042006Year H. garnotii H. mabouia Pr opo r t i o n Occup i ed Figure 3-6. Charlotte County: Proportion of buildings occupied by Hemidactylus garnotii and H. mabouia each survey year. 85

PAGE 86

86 .254.316.423.327.547.136.217.306.600.050.200.300.400.500.600.700H. mabouiaxtinctionoportion of Buildings .088.194.000.100H. garnotiiH. mabouiaH. garnotiiColonization EPr 2003-4 2004-6 2003-6 Figure 3-7. Charlotte County: Building colonization and extinction rates for Hemidactylus garnotii and H. mabouia from 2003 to 2004, 2004 to 2006, and overall from 2003 to 2006.

PAGE 87

87 Figure 3-8. Lee County: Gecko species captured at 176 survey buildings in 2004.

PAGE 88

88 Figure 3-9. Lee County: Gecko species captured at 71 survey buildings in 2004.

PAGE 89

89 Figure 3-10. Lee County: Gecko species captured at 68 survey buildings in 2006.

PAGE 90

.845.926.592.485.563.676.310.250.074.15500.10.20.30.40.50.60.70.80.912004200620042006200420062004200620042006 NoneAnyBoth 90 ( ( Gecko Species PresentProportion of Buildings Survey 1 Survey 2 Survey 3 All Surveys H garnotii H mabouia ( alone or w/ H. mabouia ) (alone or w/ H. garnotii) Figure 3-11. Lee County: Proportion of buildings on which I observed Hemidactylus garnotii and/or H. mabouia each survey and year.

PAGE 91

CHAPTER 4 LIFE HISTORY OF Hemidactylus garnotii AND H. mabouia IN SOUTHWEST FLORIDA Introduction Hemidactylus garnotii and H. mabouia are two of the most widespread and commonly encountered non-indigenous geckos in southwest Florida and have been frequently introduced around the world (e.g. Oliver and Shaw 1953, Dinsmore 1969, Bourquin 1987, Branch 1987, Araujo 1991). In Florida, H. garnotii has been implicated in the displacement of the earlier introduced congener H. turcicus, and H. mabouia in the displacement of both of these species (Meshaka 1994; Meshaka and Moody 1996; Meshaka 2000; Meshaka et al 2005). Despite this, little has been written of their natural history (see chapter 2) in either their natural range or, prior to this study, in Florida. The few data that exist on length, weight, size at maturity, and clutch frequency are based on small sample sizes, often fewer than 20 individuals (Vitt 1986, Meshaka 1994, Meshaka et al. 1994a, Meshaka and Moody 1996, Meshaka 2000). The objective for this chapter was to investigate aspects of the natural history of H. garnotii and H. mabouia that may have led to the overall success of these species and to elucidate any possible differences that could influence the differential survival of either species. To this end, I conducted a four year mark-recapture study in southwest Florida to compare inter-specific differences in the demography and other life history characteristics of H. garnotii and H. mabouia. Timeline I conducted a mark-recapture study on a population of Hemidactylus garnotii in Charlotte County Florida from August 2001 to August 2005. Simultaneously, from March 2003 to August 2005, I conducted a similar study on a population of H. mabouia in Lee County Florida. 91

PAGE 92

The initial four H. garnotii surveys from August 2001 to March 2002 were part of a pilot study to determine feasibility. These were low intensity one night surveys in which I only captured and marked exposed adult geckos. I did not include these initial surveys in most analyses. Beginning in April 2002, I increased the survey intensity to a three-day survey approximately every three to four weeks. Moreover, I redesigned my capture equipment, and formalized other aspects of my methodology such that it remained nearly identical for the rest of the surveys. I conducted no surveys in August 2004, December 2004, and January 2005. Study Site Descriptions In Charlotte County I surveyed sixteen neighboring buildings on the west side of US 41, Tamiami Trail, a distance of approximately 1/3 miles from just south of West Tarpon Boulevard NW to just north of Orange Drive (Figure 4-1). The average building area was 408 m 2 with an average perimeter of 93 linear meters (Charlotte County 2007a). In some cases, some of the perimeter inaccessible due to fences, sheds, or when buildings abutted one another (Table 4-1). In Lee County, I surveyed three nearly identical buildings in an office park located at 1342 Colonial Boulevard, Fort Myers, bounded by Colonial Boulevard, McGregor Boulevard, Royal Palm Square Boulevard, and Via Royale (Figures 4-2 and 4-3). Including the covered porch and overhang, the outer dimensions of each building were approximately 27.7 meters per side, or 111 linear meters total, with an area of approximately 767 m 2 The perimeter without the porch is approximately 23 meters per side, or 92 linear meters total, with an area of approximately 529 m 2 Each building also has a 5.5 x 22 meter carport on one side (Table 4-2) (Lee County Property Appraiser 2007). 92

PAGE 93

Methods Data Collection Beginning shortly after dark, wearing a helmet-mounted Western Rivers 100,000 candle power light, and carrying an extendable painters pole with an 18 floor squeegee with fold out rulers at the end (Figures 4-4 and 4-5), I slowly walked around each building looking for exposed geckos and peering behind refugia such as gutters and signs to detect and drive out hidden animals for hand capture. After completing an entire pass around all buildings I resurveyed the most active buildings a second time each evening in Charlotte County and all buildings two more times in Lee County. Multiple passes around many buildings were necessary to increase detection probability especially in Lee County where buildings had many refugia and geckos more easily evaded capture. I waited a minimum of 15 minutes between surveys to give hiding geckos more time to emerge from refugia. In Charlotte County one building had many large overhead signs for which I used a metal yardstick attached to an extendable pole to drive out geckos. I only surveyed these signs once per evening since I was likely to detect all hiding geckos the first time. At each survey period I hand-captured as many geckos of all sizes as possible. Once captured, I marked each animal with a unique toe clip or recorded the toe clip number if it had been previously captured. Since geckos rely on their ability to cling to vertical surfaces, I did not clip more than one toe per foot. This permitted 1,170 unique combinations. This manner of clipping has been previously shown, in another closely related Florida invasive species H. turcicus, not to affect clinging ability or sprint speed (Paulissen and Meyer 2000). For H. mabouia I then determined the sex using a Peak lighted 15x magnification lupe to detect the presence or absence of precloacal and/or femoral pores which only occur in adult males (Grant 93

PAGE 94

1932a, Loveridge 1947). This was not necessary for H. garnotii since they are an all female parthenogenetic species (Kluge and Eckardt 1969). Additionally I recorded the time of capture, measured snout-vent length (SVL) and tail length (TL) to the nearest millimeter using a flat plastic ruler, qualitatively assessed the amount of tail that was missing (none, tip, < 1 / 3 2 / 3 > 2 / 3 ), qualitatively assessed how much of the tail had regrown (none, tip, < 1 / 3 2 / 3 > 2 / 3 ), if female checked for the presence of eggs via candling, and placed a survey-specific temporary mark on the dorsum with a water-based marker to prevent recapture during each survey period thereby decreasing unnecessary stress on the animal. After January 2003, I also began recording the air temperature to the nearest 0.1 degree Celsius and percent relative humidity to the nearest 0.1 percent using an Extech Humidity/ Temperature Pen and the weight to the nearest 0.1 gram using a hanging Pesola scale. In April 2003 I began to qualitatively assess the size of visible eggs as small, medium, or large. After I recorded all data, I released each animal immediately at the capture site. To reduce the likelihood of a gecko jumping from the building upon release, I placed the gecko behind nearby refugia when available. If none was nearby or could not be reached, I turned off my light, placed the gecko on a window sill or on the wall, covered it with my hand, applied a gentle pressure, and waited for about 1 minute to calm it. I then stepped away trying not to startle it and watched for a short time to make sure it did not jump off the building. In Charlotte County, H. mabouia was occasionally encountered in the study site and was removed as quickly as possible to eliminate the possibility of demographic effects due to species interactions. Similarly, in Lee County, H. garnotii and H. frenatus were encountered a few times in the study site and removed. 94

PAGE 95

Data Analysis Survival I used Program Mark version 4.3 (White and Burnham 1999) to estimate monthly survival rate from mark-recapture data. Since each survey period lasted three days and survey periods were not evenly spaced, in Program Mark, I adjusted monthly time intervals between each survey period based on the first day in each three day period. I used a multistate model (Arnason 1972, 1973, Brownie et al. 1993, Hestbeck et al. 1991, Nichols et al. 1992, Schwarz et al. 1993, Wood et al. 1998) in which the typical Cormack-Jolly-Seber (CJS) (Cormack 1964, Jolly 1965, Seber 1965) open population mark-recapture model is extended to account for stages. Both the CJS and multi-state model incorporate the parameter, the probability a marked gecko in stage g is recaptured in any month i, given it is alive. However, for the multi-state model, the parameter the probability that an animal survived from time i to i+1 and remained in the sampling area, used in the CJS model, is subdivided into the two parameters:, the probability that a gecko in stage r survives one month, and giP riS rgi the probability a gecko in stage r one month is in stage g the next month, given it is alive. As opposed to the CJS model, in which age-classes have deterministic intervals, the multi-state model allows for stages in which intervals are stochastic by including the probability of transitioning to the next stage or of remaining in the same stage. Modified CJS assumptions for this model are: all marked geckos in stage r during survey i have equal probabilities of being recaptured, and; immediately following survey i, all marked geckos in stage r have an equal probability of surviving until survey i+1 and moving to stage s by survey i+1. This model additionally assumes that: survival from time i to i+1 does not depend on stage at time i+1, and; all individuals make transitions to the next stage at the same 95

PAGE 96

moment during the month (Cooch and White 2006, Joe and Pollock 2002, Williams et al. 2002). Standard CJS assumptions (Cormack 1964, Jolly 1965, Pollock et al. 1990, Seber 1965, 1982) include: marks are not lost (i.e. toes do not regrow) throughout the study; marks are read and recorded correctly; sampling is instantaneous (i.e. sampling period length is short compared to the amount of time between sampling periods); marked and unmarked geckos have homogeneous survival rates between survey periods; marked and unmarked animals have equal probability of being captured at each survey; the characteristics of the survey area do not change throughout the study; losses due to emigration are permanent (i.e. geckos do not leave the study area and then return). I designated three stages based on snout-vent length (SVL) as juvenile, ( 30 mm SVL), subadult, (31 to 1 mm less than maturation size), and adult, ( maturation size). I based the specific cutoff point between the subadult and adult stages on the smallest gravid individual I detected during this study. For juvenile and subadult, the division was somewhat arbitrary but based upon my preliminary survey observations that few hatchlings appeared to survive. To test the hypothesis that juvenile and subadult survival differed, I included models in my candidate set where: juvenile and subadult survival were the same, and; juvenile and subadult survival differed. For H. mabouia, I also tested the hypotheses that adult male and female survival and/or recapture rates differed by dividing the adult stage into separate male and female stages (Wood et al., 1998). This permited subadults of unknown sex to transition into either the adult male or females stage. Since male adults could not transition to female adults and vice versa I fixed 0mffm Due to the growth rate of juvenile geckos, I assumed that juveniles would transition to the subadult stage within one month and therefore fixed transition 1js (thereby also fixing 96

PAGE 97

0jj since their sum must equal 1). I also fixed and excluded from these analyses all records in which geckos were captured for the second time in the juvenile state. Additionally, since juveniles could not skip the subadult stage and transition directly to the adult stage and subadult and adult stages could not transition to previous stages, I also fixed for H. garnotii 0jp 0asajsjja and for H. mabouia 0fjfsmjmssjjfjm For H. garnotii, this resulted in the direct estimation of only a single transition probability, sa since ss and aa were determined by subtraction. (i.e. ss = 1sa sj =1sa and aa =1as aj =1) (Cooch and White 2006). Similarly, for H. mabouia, only sm and sf were directly estimated and ss =1sm sf and ff = mm =1. For comparative purposes, I estimated means for time-specific probabilities using the variance components procedure in Program Mark (White et al. 2001). To test how well the models were supported by the data, I used a multistate goodness-of-fit test in program U-CARE version 2.2.5 (Choquet et al. 2005) to estimate the variance inflation factor, (Burnham et al. 1987, Lebreton et al.1992) of the most parameterized fully time-dependent model in which the survival probability varied between stages and between surveys (Pradel et al. 2003, 2005) where [=/df] and ^c ^c 2 2 is the Pearson goodness-of-fit statistic and df is the model degrees of freedom. A value of [ = 1] indicates a perfect model fit and [>1] indicates overdispersion, meaning that the actual sampling variance is greater than the estimated variance. There are many possible reasons for overdispersion including lack of independence among individuals and heterogeneity in survival and/or capture probabilities (Anderson et al. 1994). The multistate goodness-of-fit test implemented using U-CARE is based on what Pradel et al. (2003) call the Jolly Move model (Brownie et al. 1993) which is more general than, and ^c ^c 97

PAGE 98

encompasses, the Arnason-Schwarz model (Arnason 1972, 1973, Brownie et al. 1993, Hestbeck et al. 1991, Schwarz et al. 1993) used when analyzing multistage data in Program Mark. However, their similarity is such that, in most cases, the goodness-of-fit test for the Jolly Move model is also considered to be a valid goodness-of-fit test for the Arnason-Schwarz model (Cooch and White 2006). To test hypotheses involving recapture and survival probabilities, I used the Akaike Information Criterion (AIC) to determine the most parsimonious model and compare it with models in which I varied a single recapture or survival time or stage parameter. The AIC value is based on the model likelihood but penalizes more parameterized models (Akaike 1973, Burnham and Anderson 1992, 1998, 2002, MacKenzie et al. 2006, Williams et al. 2002). AIC differences among models within a data set describe the relative support for that model compared to the most parsimonious model. Models with lower AIC values better represent the data compared to other models in that candidate set. Models with AIC values that differ by less than 2 would be considered similar in support to models in that data set. Growth I calculated growth rates for all size classes combined and separately for adults and non-adults (i.e. juveniles plus subadults). For growth rate analyses, I only included geckos that were captured in two consecutive sampling periods. If a gecko was captured in two consecutive sampling periods more than once, I only considered the first two consecutive captures. An exception was when the first consecutive captures spanned the non-adult/adult cutoff point in which case I used the next two consecutive captures. If a geckos only consecutive captures spanned the non-adult/adult cutoff point then this was not used for the non-adult or adult size 98

PAGE 99

class estimations but only for the combined size-class estimate. Since growth appeared to slow substantially or cease, I excluded the largest 10% of adults from these analyses. To estimate the overall growth rates throughout the study, I determined the growth in millimeters between each geckos capture and recapture in a consecutive sampling period and divided this by the number of days in this interval. To estimate if the growth rate varied by month of the year, I assigned each growth rate estimate to the first of the two consecutive months in which the gecko was captured. Reproduction For logistical reasons I only included reproductive data from 23 April 2003 and after (survey 19 and after for H. garnotii and survey 3 and after for H. mabouia) in data analyses. Time until maturity I calculated the time until maturity using my estimate of growth rate to estimate the time to grow from juvenile size to the minimum size at maturity. I included only geckos captured as a juvenile and recaptured within 5 mm SVL of the minimum size of maturity determined in this study. First, I divided the number of days to grow from juvenile size to a size within 5 mm SVL of the minimum size of maturity by the mean daily growth rate determined in this study. I then calculated a correction for each estimate. For juveniles larger than the mean juvenile capture size, I added the estimated number of days to grow from the mean juvenile size to this larger initial juvenile capture size. For subadults up to five millimeters smaller than the minimum size at maturity, I added the estimated number of days to grow from this size to the minimum size at maturity. For adults up to five millimeters larger than the minimum size at maturity, I subtracted the number of days to grow from the minimum size at maturity to this capture size. I chose a 99

PAGE 100

cutoff of 5 mm to minimize potential inaccuracies of growth rate estimations over large intervals. Maximum number of clutches per year I determined the maximum number of times any gecko was captured when gravid during any one year interval throughout this study. A one year interval consisted of the 365 days following the first day in any of the survey periods (Table 4-3 and 4-4). Since survey periods were not spaced evenly, all one year intervals were actually slightly less than one year (e.g. for H. mabouia, one year periods occurred from surveys 3 to 18, 4 to 19, 5 to19, 6 to 20 etc.). I also estimated the theoretical maximum annual number of clutches by dividing the estimated length of the reproductive season by the combined time to produce a clutch of eggs and the interclutch interval. I estimated the combined time to produce a clutch of eggs and the interclutch interval as the shortest time recorded between two consecutive gravid captures of similar sized eggs (e.g. capture one egg size = large, capture 2 egg size = large or capture 1 egg size = medium, capture 2 egg size = medium). Breeding proportion I estimated the monthly breeding proportion of adult females using a multi-state model in Program Mark (Nichols et al. 1994). This model was similar to the multi-state model I described previously to estimate survival of juvenile, subadult, and adult size classes except that instead of size classes, I used two breeding states: gravid and non-gravid. Instead of transitioning from one size class to the next (e.g. juvenile to subadult or subadult to adult), in this case, geckos could transition between gravid and nongravid reproductive states. Unlike size class transitions, in which transitions could only occur in one direction (e.g. from juvenile to subadult but not from subadult to juvenile), reproductive state transitions could occur both from gravid to nongravid 100

PAGE 101

and from nongravid to gravid. In this case, the parameter of interest was the capture probability and if it differed for gravid and non-gravid geckos. I then used the estimated capture probabilities to adjust the nave monthly breeding and non-breeding counts to obtain an adjusted estimate of the proportion of breeding adult females per month (Equation 4-1). Breeding proportion = (N r i / p r i ) / [(N r i / p r i ) + (N n i / p n i )] (4-1) where: N r i = Number of gravid geckos in month i N n i = Number of nongravid geckos in month i p r i = probability a marked gravid gecko is recaptured in any month i p n i = probability a marked non-gravid gecko is recaptured in any month i To test how well the models were supported by the data, I used a multistate goodness-of-fit test in Program U-CARE version 2.2.5 (Choquet et al. 2005) as described previously. Mean fecundity and fertility To estimate the mean monthly fecundity I multiplied the breeding proportion by two, the fixed clutch size (Fitch 1981), and by the proportion of females in the population (i.e. sex ratio), under the assumption that the adult sex ratio was representative of the hatchling sex ratio. I estimated the mean fertility by multiplying the mean fecundity by the adult survival probability. Maximum longevity. To estimate maximum longevity, I calculated the longest time interval between any geckos first and last captures. I then adjusted this estimate by adding an estimate of this geckos age at first capture. I determined its age at first capture by multiplying the difference between the mean juvenile size and its size at first capture by the mean daily growth rate from this study. Abundance I estimated the mean number of subadult and adult geckos on each survey building at each survey period, by dividing the number of geckos captured on each building for each ^giN 101

PAGE 102

size class in each time period, by the capture probability for each size class in each time period, (Equation 4-2). gin gip^ ^^/gigigipnN (4-2) I estimated the standard error of as the product of the number of geckos captured on each building for each size class in each time period and the estimated standard error for each size class and time-specific capture probability, divided by the square of the size class and time-specific capture probability, (Equation 4-3) (Wood et al. 1998). ^giN gin )(^^gipSE 2^)(gip 2^^^)()]([^^gigigigippSEnNSE (4-3) Capture probabilities were estimated from the combined data from all buildings and do not reflect individual probabilities for each building. Results General Summary In Charlotte County during 47 survey periods I captured 914 individual H. garnotii and made 1,286 recaptures for a total of 2,200 captures (Table 4-3). I did not record SVL data for five captures so was unable to include these in most analyses. In Lee County during 31 survey periods I captured 546 individual H. mabouia and made 484 recaptures for a total of 1,030 captures (Table 4-4). 102

PAGE 103

Size at Maturity Of the 914 H. garnotii captured, 198 (21.7%) were gravid at least once during the study. The smallest gravid individual was 49 mm SVL (Figure 4-17). Two 46 mm and two 47 mm SVL geckos appeared to contain very small eggs but I was not confident in this assessment and chose to exclude these. Of the 546 H. mabouia captured, 92 (16.8%) were gravid at least once. The smallest size of any gravid individual was 51 mm SVL (Figure 4-19). For this research, I considered the minimum size at maturity to be the minimum adult SVL. However, when considering only the first gravid capture for each individual, the modal SVL was 53 mm for H. garnotii and 55 mm for H. mabouia. This indicated that most H. garnotii and H. mabouia likely matured at a somewhat larger size than the minima presented. I discuss reproductive output and size in a subsequent section. Sex Ratio I identified the sex of 261 H. mabouia adults of which 49% (127) were female and 51% (134) male. This observed adult sex ratio is not significantly different from 1:1 ( 2 = 0.137, df = 1, p = 0.711). These adult ratios do not preclude the possibility of a differential sex ratio during the egg, hatchling, juvenile, or subadult stages. Size Distribution H. garnotii length ranged from 23 to 63 mm SVL (Figure 4-6) and H. mabouia from 20 to 69 mm SVL (Figure 4-7). When only the first capture for each gecko was considered, the overall mean SVL of 37.7 mm (se = 0.326, 95% CI: 37.1 38.3, n = 913) for H. garnotii was significantly smaller than for H. mabouia (41.3 mm (se = 0.551, 95% CI: 40.2 42.4, n = 546)) (t = 5.99, df = 1457, p < 0.001). When juveniles were excluded, the mean SVL for H. garnotii of 42.3 mm (se = 0.340, 95% CI: 41.6 43.0, n = 644) was also significantly smaller than for H. 103

PAGE 104

mabouia (47.7 mm (se = 0.481, 95% CI: 46.8 48.6, n = 387)) (t = 9.35, df = 1029, p < 0.001), a 12.8% length differential. When size classes were considered separately, the adult mean SVL of H. garnotii of 53.9 mm (se = 0.243, 95% CI: 53.4 54.4, n = 175) was significantly smaller than that of H. mabouia (57.2 mm (se = 0.341, 95% CI: 56.5 57.9, n = 162)) (t = 7.84, df = 335, p < 0.001). For H. mabouia adults, the mean SVLs of 56.6 mm (se = 0.407, 95% CI: 55.8 57.4, n = 76) for females and 57.7 mm (se = 0.537, 95% CI: 56.6 58.8, n = 84) for males were not significantly different (t = 1.57, df = 158, p = 0.119). For subadults, the mean SVL for H. garnotii of 37.7 mm (se = 0.241, 95% CI: 37.2 38.2, n = 454) was significantly smaller than for H. mabouia (40.9 mm (se = 0.367, 95% CI: 40.2 41.6, n = 225)) (t = 7.62, df = 677, p < 0.001). On the other hand, the mean juvenile size of 27.8 mm (se = 0.102, 95% CI: 27.6 28.0, n = 284) for H. garnotii was significantly greater than for H. mabouia at 25.7 mm (se = 0.177, 95% CI: 25.4 26.0, n = 159) (t = -11.28, df = 441, p < 0.001). H. garnotii weight ranged from 0.2 to 5.1 grams. When juveniles were excluded and only the first capture for each non-gravid gecko was considered, the overall mean weight was 1.1 grams (se = 0.05, 95% CI: 1.0 1.2, n = 197, range = 0.4 to 4.0). When size classes were considered separately, the mean subadult weight was 0.9 grams (se = 0.03, 95% CI: 0.8 1.0, n = 176, range = 0.4 to 2.4) and adult was 2.6 grams (se = 0.13, 95% CI: 2.3 2.9, n = 21, range = 1.8 to 4.0). For H. mabouia, weight ranged from 0.2 to 6.9 grams. When juveniles were excluded and only the first capture for each non-gravid gecko was considered, the overall mean weight was 2.0 grams (se = 0.08, 95% CI: 1.8 2.2, n = 249, range = 0.4 to 6.8), for subadults was 1.3 grams (se = 0.04, 95% CI: 1.2 1.4, n = 177, range = 0.4 to 3.1), and for adults was 3.6 grams 104

PAGE 105

(se = 0.12, 95% CI: 3.4 3.8, n = 72, range = 2.1 to 6.8). When the weight at first capture for male and nongravid female adults were considered separately, the mean of 3.8 grams (se = 0.16, 95% CI: 3.5 4.1, n = 47, range = 2.1 to 6.8) for males was not significantly different (t = -1.81, df = 70, p = 0.075) than the mean of 3.4 grams for nongravid females (se = 0.15, 95% CI: 3.1 3.7, n = 25, range = 2.2 to 5.2). The overall mean weight for H. garnotii was significantly different than from H. mabouia (t = -9.4, df = 444, p < 0.001), as was the mean for subadults (t = -8.9, df = 351, p < 0.001), and for adults (t = -4.52, df = 91, p < 0.001). Captures per Animal I captured 914 H. garnotii a total of 2,200 times for a mean of 2.41 captures per animal. Of the 914 individuals captured, 536 (58.6%) were not recaptured and 378 (41.4%) were recaptured one or more times. Of those recaptured, the mean number of captures was 4.40 captures per gecko. I captured 546 H. mabouia a total of 1,030 times for a mean of 1.89 captures per animal. Of the 546 individuals captured, 308 (56.4%) were not recaptured and 238 (43.6%) were recaptured one or more times. Of those recaptured, the mean number of captures was 3.03 captures per gecko. The maximum number of captures for an individual H. garnotii was 28 and for H. mabouia was 7. The maximum number of individuals captured during one survey period was 92 for H. garnotii and 100 for H. mabouia. The maximum number of newly captured geckos in one survey was 67 for H. garnotii and 74 for H. mabouia. The maximum number of recaptured geckos in one survey was 59 for H. garnotii and 31 for H. mabouia (Tables 4-3 and 4-4). When only the first capture for each gecko was considered, the proportion of the total H. garnotii captured in each size class was 0.31, 0.50, and 0.19 for juvenile, subadult, and adult stages, respectively and for H. mabouia was 0.29, 0.41, and 0.30, respectively (Table 4-5). Thus, 105

PAGE 106

of those geckos captured, approximately 70% of both species avoided capture as juveniles. However, I was able to capture 70 to 81% of all individuals prior to the adult stage. Whereas, the mean number of captures for all H. garnotii was 2.41 captures per animal and for all H. mabouia was 1.89 captures per animal, the number of captures when only adults were considered was higher. I captured 336 H. garnotii at least one time at adult size and made 1,045 total adult captures for a mean of 3.11 captures per adult. Of the 336 adults captured, 136 (40.5%) were not recaptured and 200 (59.5%) were recaptured one or more times. Of those recaptured, the mean number of captures was 4.55 captures per gecko. I captured 264 H. mabouia at least one time at adult size and made 528 total adult captures for a mean of 2.00 captures per adult. Of the 264 adults captured, 135 (51.1%) were not recaptured and 129 (48.9%) were recaptured one or more times. Of those recaptured, the mean number of captures was 3.05 captures per gecko. There was a significant difference between the number of captures per animal for the two species (t = -6.78, df = 860, p < 0.001) as well as in the number of geckos captured only once (t = 2.522, df = 596, p = 0.012). The higher number of adult recaptures, as compared to the overall number of recaptures when all size classes were considered for both species, likely indicates a higher survival rate for adults. I consider survival rates more thoroughly in a subsequent section. I captured 134 H. mabouia adult males a total of 261 times for a mean of 1.95 captures per adult male and 128 adult females 265 times for a mean of 2.07 captures per adult female. The mean adult male and female capture rates were not significantly different (t = 0.786, df = 258, p = 0.433). Overall 51.1% of the adults were only captured one time while 47.7% (61) of the females and 53.7% (72) of the males were only captured one time. This difference was also not significant (t = -0.985, df = 260, p = 0.326). That these capture rates are not significantly 106

PAGE 107

different for males and females further reinforces the notion of a 1:1 adult sex ratio. In a subsequent section I will also test if the recapture probabilities differed among sexes using a multistate model. Seasonal Effects of Capture Rate The number of each size class captured at each survey period (i.e. capture rate) and the mean air temperatures are presented in Table 4-6, and Figures 4-8 to 4-11. Total number of captures for both species are strongly correlated with the mean air temperature throughout the study (H. garnotii: r = 0.775, p < 0.001; H. mabouia r = 0.925, p < 0.001). Both species exhibited clear seasonal highs and lows in capture rate. Peak capture rates for H. garnotii occurred in July and August and for H. mabouia in September. The fewest captures occurred from mid December through mid February for H. garnotii and from November through April for H. mabouia (Figure 4-8). Hemidactylus mabouia was more sensitive to low temperatures with capture rates dropping to just a few animals as temperatures fell to 21 o C and below while for H. garnotii such a reduction in activity was not seen until temperatures fell to around 12 o C. When capture rates are viewed separately for each size class, clear seasonal patterns are seen for juvenile (Figure 4-9) and subadults (Figure 4-10) of both species. Adult patterns (Figure 4-11) are more convoluted but still show seasonality. For H. garnotii, juvenile capture rates were highly correlated with mean air temperature (r = 0.755, p < 0.001) and were highest in late June in 2002, 2003, and 2005, and in July in 2004. Hemidactylus mabouia juvenile capture rates and mean air temperatures were significantly positively correlated (r = 0.548, p = 0.001) and were highest in mid-September in both 2003 and 2004 (Figure 4-9). Subadult capture rates for both species were positively correlated with mean air temperature (H. garnotii: r = 0.478, p = 0.001; H. mabouia: r = 0.533, p = 0.002) and were highest for H. garnotii in late summer to early 107

PAGE 108

fall and for H. mabouia in mid to late September (Figure 4-10). Adult capture rates and mean air temperature were weakly correlated for H. garnotii (r = 0.347, p = 0.023) but were highly correlated for H. mabouia (r = 0.920, p < 0.001). For H. garnotii adult capture rates are somewhat erratic but generally have a peak month or two in the spring, summer, and fall and decrease in the winter. While H. mabouia adults have distinct peak capture rates each year, the timing is less specific than for other size classes (Figure 4-11). Minimum capture rates were fairly similar for all size classes. Survival probability estimation Stages I considered the juvenile stage for both species to be 30 mm SVL and less. I determined the cutoff between the subadult and adult stages from the minimum size at maturity. This resulted in a subadult size class of 31 to 48 mm and 31 to 50 mm SVL for H. garnotii and H. mabouia, respectively, and an adult stage from 49 mm SVL and more for H. garnotii and 51 mm SVL and more for both male and female H. mabouia. Model selection Since I fixed transition 1js and p j = 0, I removed the 9 H. garnotii and 12 H. mabouia capture records which were repeat juvenile captures of individuals previously captured as a juvenile. Results from multistate goodness-of-fit tests (Table 4-7) indicate slight underdispersion, = 0.819, for H. garnotii and moderate underdispersion, = 0.451 for H. mabouia so did not necessitate any adjustments to (Cooch and White, 2006). ^c ^c ^c For H. garnotii, the model that was best supported by the data was S(g)p(g*t)(g) where survival probability (S) is estimated separately for each stage (g), recapture probability is estimated separately for each stage and for each capture period (g*t), and transition probability 108

PAGE 109

() is estimated separately for each stage (Table 4-8). Comparisons of models in which I only varied survival probability indicated that the survival probability showed stage differences but not survey specific differences (Table 4-8). The model in which only stage was considered had an AIC value of 100.69 lower than models in which both stage and time were considered. Further, there was good evidence that survival rates differed for all three stages since two-stage models, in which the survival rates for the subadult and adult stages were considered equal and in which the juvenile and subadult stages were considered equal, had substantially higher AIC values differing by 30.56 and 86.30, respectively. Comparisons of models in which I only varied recapture probability indicated that the recapture probability differed both among stages and over time (i.e. among surveys). Models in which stage or time was not considered were not as well supported by the data with AIC values of 21.40 and 97.41 greater, respectively. For H. mabouia, the best supported model was S(g J, S=M=F )p(t)(.) where survival probability was estimated separately for the juvenile stage (g J ) and the other stages, which were estimated together (g S=M=F ), recapture probability was estimated jointly for all stages but separately for each capture period (t), and transition probability was estimated jointly for all stages and constant among capture periods (.) (Table 4-9). There did not appear to be any survey specific survival differences (Table 4-9) however, the model in which time was a factor had some convergence problems so this conclusion is not certain. Male and female adult survival probabilities did not differ (AIC = 3.27) and juvenile survival probability differed from the other stages (AIC = 11.08). However, while subadult survival probability appeared to differ from adult probabilities (AIC = 2.10) this small AIC value makes this inference somewhat less conclusive. For H. mabouia recapture probabilities, there were only differences among time 109

PAGE 110

but not among stage. The best supported model in which stage was considered had an AIC value 42.27 greater than with time alone. For transition probability, models in which time was a factor did not converge properly for either species so were not considered further. For H. mabouia, models in which subadult to adult transition probabilities differed among sexes or were constant had somewhat similar support but leaned towards suggesting similar transition probabilities from the subadult stage for both sexes (AIC = 2.03) (Table 4-9). Capture probability For H. garnotii capture probabilities differed for subadults and adults and were time-specific. For subadults, the probability of recapturing a marked gecko in any given month, given that it was alive, ranged from an estimate of about 0 to 1 with an overall mean estimate of 0.48 (= 0.04). For adults, estimated probabilities ranged from about 0.06 to 0.79 with an estimated mean of 0.36 (= 0.02) (Table 4-10). For H. mabouia, capture probabilities were time-specific but did not differ for subadults and both sex adults. Estimated probabilities ranged from about 0.01 to 0.87 with a mean estimate of 0.238 (mean = 0.045) (Table 4-11). That these capture probabilities did not differ for male and female adults and that capture rates presented previously were not significantly different for males and female adults, again is evidence of a 1:1 adult sex ratio. ^se ^se ^se Transition probability For both species, due to the growth rate, I fixed the probability that a gecko captured as a juvenile in one month would be captured as a subadult the next month, js to 1 thereby fixing jj to 0. Likewise for adults, since as = aj =0, aa =1 for H. garnotii and similarly for H. 110

PAGE 111

mabouia, since ms = fs = mj = fj =0, mm = ff =1. For H. garnotii subadults, I estimated sa to be 0.218 (= 0.017) and therefore ^se ss = 0.782 (Table 4-10). For H. mabouia subadults, sm = sf and were estimated to be 0.130 (= 0.012) and therefore ^se mm = ss is estimated to be 0.870 (Table 4-11). Survival probability I estimated the probability that a juvenile H. garnotii survived one month was 0.283 (= 0.035), a subadult was 0.749 (= 0.016), and adult was 0.857 (= 0.008) (Table 4-10). For H. mabouia, I estimated the probability that a juvenile survived one month was 0.570 (= 0.074) and for subadults, male adults, and female adults to be 0.866 (= 0.009) (Table 4-11). ^se ^se ^se ^se ^se Growth When all size classes are combined, the mean daily growth rates of 0.147 mm per day for H. garnotii and 0.135 mm per day for H. mabouia are not significantly different (t = 1.125, n = 409, p = 0.261) (Table 4-12). When adults were analyzed separately, the mean daily growth rates of 0.043 mm per day for H. garnotii and 0.068 mm per day for H. mabouia were significantly different (t = -3.214, n = 125, p = 0.002). When non-adults were analyzed separately, the mean daily growth rates of 0.183 per day for H. garnotii and 0.191 mm per day for H. mabouia were not significantly different (t = -0.639, n = 259, p = 0.524) (Table 4-12). In general, the highest growth rates occurred in May and June for both species and all size classes (Table 4-12 and Figures 4-12 to 4-14). Maximum non-adult growth rates of 0.267 mm per day (8.01 mm per month) and 0.302 mm per day (9.06 mm per month) for H. garnotii and H. mabouia, respectively were significantly higher than for adults which were 0.089 mm per 111

PAGE 112

day (2.67 mm per month) and 0.110 mm (3.30 mm per month), respectively (H. garnotii: t = 3.644, df = 35, p < 0.001; H. mabouia: t = -4.591, df = 6, p = 0.005) (Table 4-12 and Figures 4-13 and 4-14). Higher than average daily growth rates occurred for non-adults May through August for both species. For adult H. garnotii, above average rates occurred April to June and in September and October and for adult H. mabouia from April through July however sample sizes for both were very small. Markedly slower growth occurred during the winter for both species. From November to March with growth rates ranging from 0.026 to 0.049 mm per day for H. garnotii and from 0.033 to 0.045 mm per day for H. mabouia. H. garnotii sample size in December and all winter H. mabouia samples are very small, with no samples for December. This likely reflects decreased movement during these periods resulting in low capture and recapture rates. When the overall log growth rates were compared to log mean temperatures there was a highly positive correlation (H. garnotii: r = 0.882, p < 0.001; H. mabouia; r = 0.768, p = 0.006). For adults there was a significant correlation for H. garnotii (r = 0.751, p = 0.005) but not for H. mabouia (r = -0.075, p = 0.861). For non-adults there was a highly significant correlation for both species (H. garnotii: r = 0.795, p = 0.002; H. mabouia: r = 0.765, p = 0.006). Time until maturity I captured 31 H. garnotii and 17 H. mabouia individuals both as juveniles and within 5mm SVL of the minimum size of maturity. Once mean growth rate correction factors were applied to juveniles larger than the mean juvenile size (i.e. H. garnotii: 28 < x 30 mm SVL; H. mabouia: 26 < y 30 mm SVL) and for subadults and adults mm from the minimum maturation size (i.e. H. garnotii: 44 x 48 and 50 x 54 mm SVL; H. mabouia: 46 x 50 and 52 x 56 mm SVL), the time until maturity ranged from 67 to 283 days with a mean of 112

PAGE 113

147 days (se = 12) for H. garnotii and from 100 to 295 days with a mean of 226 days (se = 15) for H. mabouia. As a comparison, maturation times calculated using subadult yearly survival rates derived from mark-recapture data using Program Mark (H. garnotii: = 0.032; H. mabouia: = 0.178) were 136 days (0.372 years) for H. garnotii and 242 days (0.663 years) for H. mabouia. These estimates agree with those calculated using mean daily growth rates. subadultyearS subadultyearS For H. garnotii maturation time was significantly correlated with hatch month such that maturation time increased from a mean of 78 days for those hatched in April up to 213 days for August hatchlings (r = 0.465, p = 0.008). Hemidactylus mabouia maturation times were similarly significantly correlated with month with May-June hatchlings maturating after 118-137 days and August to November hatchlings in 240-264 days (r = 0.775, p < 0.001) (Figure 4-15). Reproduction Reproductive season I captured gravid H. garnotii in every month and gravid H. mabouia in all months except December. However, just as activity was reduced in certain months for both species, so was reproductive output. I characterized the reproductive season in relation to the mean of the monthly mean number of gravid captures. The primary reproductive season included those months in which the number of gravid captures exceeded the mean of the monthly mean number of gravid captures (H. garnotii: 8.54 (se = 1.16); H. mabouia: 3.56 (se = 0.8)) (Figure 4-16). For H. garnotii this occurred in six months from February through June and November and for H. mabouia in six months from April through September. A more liberally defined reproductive season in which the mean number of gravid captures exceeded half the mean of the monthly mean number of gravid captures occurred for H. garnotii for eleven months from January 113

PAGE 114

through November and for H. mabouia for seven months from April through October. Overall, H. garnotii had a bimodal reproductive season with a primary peak occurring in March and a secondary peak in November and lows in July and December. Hemidactylus mabouia had a single peak in June, decreasing until December and then gradually increasing until June. Reproductive frequency The minimum time observed between two consecutive gravid captures was fifteen days for H. garnotii and sixteen days for H. mabouia, indicating that the interclutch interval is relatively short. For H. mabouia I classified the eggs in both of these captures as similar in size (large). This indicates that the interclutch interval is substantially shorter than sixteen days since this period encompassed both the production of a clutch of eggs as well as the interclutch interval. For H. garnotii, the shortest time recorded between two consecutive gravid captures of similar sized eggs was 21 days. One H. garnotii individual was captured gravid three times consecutively over a 48 day span, or once every 24 days. Another individual was captured gravid four times in 79 days in March, April, May, and June 2004, once every 26 days, and then again in July 2004 for a total of five times in a 132 day span (once every 33 days). Four H. mabouia individuals were captured gravid three times consecutively, one of which was over a 33 day span, or once every 16.5 days. Two others I captured were gravid in four of five periods or four times in 79 and 89 days (once every 26 and 30 days). Maximum number of clutches per year Based on the shortest time recorded between two consecutive gravid captures of similar sized eggs and the estimated length of the reproductive season, assuming a constant reproductive rate, H. garnotii could theoretically produce 16 clutches per year and H. mabouia 13 clutches per 114

PAGE 115

year. The maximum number of gravid captures within any year period I recorded for any adult H. garnotii was eight and for any adult female H. mabouia was four. Breeding proportion Multistate goodness-of-fit test results for the most parameterized fully time-dependent model indicated good model fit for both species (H. garnotii: = 0.653; H. mabouia: =0.499) (Table 4-13). The most parsimonious model for both species was S(.)p(g)(.) in which survival and transition probabilities were constant but capture probabilities were group but not time dependent (Tables 4-14 and 4-15). However, for H. garnotii model S(g)p(g)(.) was similarly supported (AIC = 0.28) and for H. mabouia also similarly supported were models S(.)p(.)(.) (AIC = 0.92) and S(.)p(g)(g) (AIC = 1.80). I therefore calculated model averaged estimates (Tables-4-16 and 4-17). For both species, capture probabilities are group specific however, for H. garnotii the model averaged capture probability for gravid adults is lower (0.345, 95% CI: 0.265 0.434) than for non-gravid adults (0.531, 95% CI: 0.432 0.628) (Table 4-16) whereas for H. mabouia it is higher for gravid adults (0.265, 95% CI: 0.116 0.497) than for non-gravid adults (0.176, 95% CI: 0.105 0.281) (Table 4-17). The estimated mean monthly proportion of adult females that were gravid was 0.452 for H. garnotii (Table 4-18) and 0.378 for H. mabouia (Table 4-19). ^c ^c Fecundity and fertility Multiplying the breeding proportion by a fixed clutch size of two and, for H. mabouia, a female population proportion (i.e. sex ratio) of the mean monthly fecundity was estimated to be 0.452 2 *1 = 0.904 for H. garnotii and for H. mabouia to be 0.378 2 0.5 = 0.378. Multiplying by the adult survival probability, fertility is estimated to be 0.904 0.857 = 0.774 and 0.378 0.866 = 0.327. 115

PAGE 116

Reproduction and size H. garnotii The proportion of gravid H. garnotii at each SVL (number gravid at each SVL / total number captured at each SVL) increased gradually from a low of 3.3% at 49 mm SVL to 49% at 56 mm SVL and then, except, for a couple of variations generally remained steady for larger SVLs through 62 mm (Figures 4-17 and 4-18). The substantial decrease at 63 mm SVL was due to a single individual being captured at this size that was not gravid. The proportion of the total population that was gravid (number gravid at each SVL / total number of adults captured) ranged from 2.9 to 6% for geckos between 53 and 60 mm SVL but was 2% or less for the smallest (49-52 mm SVL) and largest (61-63 mm SVL) adults. Further, analysis of the contribution of each SVL to the overall reproductive output of the population (i.e. number gravid at each SVL / total number gravid at all SVLs) revealed a similar pattern (Figure 4-18). While individuals were collected gravid from 49 to 62 mm SVL, only those between 53 and 60 mm SVL exceeded the mean reproductive contribution of 6.7% with approximately 85% of all reproduction from this size range. Overall, 95% of all reproduction was from individuals with SVLs between 51 and 60 mm which comprised 86.3% of the total population. Generally, the reproductive contribution of geckos with SVLs between 53 and 60 mm was greater than their population proportion. For geckos with SVLs greater than 60 mm, the reproductive contribution was similar to their population proportion and for those with SVLs less than or equal to 52 mm, the reproductive contribution was less than their population proportion. Though the minimum reproductive size was 49 mm SVL, this size class only contributed 0.4% to total reproduction despite comprising 4.3% of the total population. At 16.4%, geckos with SVLs of 58 mm had the greatest reproductive contribution. 116

PAGE 117

Reproduction and size H. mabouia The proportion of gravid H. mabouia individuals at each SVL (number gravid at each SVL / total number captured at each SVL) increased gradually from a low of 8.3% at 51 mm SVL to 40% at 54 mm SVL and then remained between 40 and 100% for larger SVLs (Figures 4-19 and 4-20). No 68 mm SVL females were captured and the one 69 mm female was not gravid. The proportion of the total population that was gravid (number gravid at each SVL / total number of adults captured) was 2.2% or less for the smallest (51-54 mm SVL) and largest (63-69 mm SVL) adults and ranged from 4.4 to 7.9% for those between 55 and 62 mm SVL. Although H. mabouia individuals were collected gravid with SVLs between 51 and 67 mm, only those with SVLs between 55 and 62 mm exceeded the mean reproductive contribution of 5.3% with approximately 83% of all reproduction from this size range (Figure 4-20). Overall, 93.5% of all reproduction was from individuals with SVLs between 53 and 63 mm. This size group comprised 85.5% of the total population. Though the minimum reproductive size was 51 mm SVL, this size class only contributed 0.8% to total reproduction despite comprising 5.3% of the total population. At 14.6%, geckos with a SVL of 58 mm had the greatest reproductive contribution. Maximum Longevity The longest time interval between the first and last capture was 1,197 days for H. garnotii and 837 for H. mabouia. At first capture these individuals had SVLs of 57 mm and 65 mm SVL, respectively. At a mean daily growth rate of 0.147 mm for H. garnotii and 0.135 mm for H. mabouia this makes their estimated maximum longevity to be 1,401 and 1,126 days, respectively, or 3.84 and 3.08 years, respectively. This may be an underestimate since the longest 117

PAGE 118

time interval I saw between first and last capture was only slightly shorter than the entire sampling period. Tail Loss On the first capture for 893 H. garnotii and 532 H. mabouia, the proportion for which any portion of the tail was missing or regrown (Table 4-20) was 34.6% (309) for H. garnotii and 19.2% (102) for H. mabouia and was significantly different between species (t = 6.29, df = 1423, p < 0.001). The proportion of juveniles, subadults, and adults for which any portion of the tail was missing or regrown was 13.5% (38 of 281), 36.2% (160 442), and 65.3% (111 of 170), respectively, for H. garnotii and 1.3% (2 of 157), 20.4% (44 of 216), and 35.2% (56 of 159), respectively, for H. mabouia. For both species, there was a significant difference among size classes (H. garnotii: 2 = 126.440, df = 2, p < 0.001; H. mabouia: 2 = 59.078, df = 2, p < 0.001). For all size classes, there were highly significant differences between species (Juveniles: t = 4.357, df = 436, p < 0.001; Subadults: t = 4.176, df = 656, p < 0.001; Adults: t = 5.717, df = 327, p < 0.001). For H. mabouia adults, 44.4% (36 of 81) of males and 26.3% (20 of 76) of females captured were missing a portion of the tail or had a portion regrown (Table 4-21). This difference was significant (t = 2.41, df = 155, p = 0.017). Movement Most H. garnotii that were recaptured, were on the same building as the original capture. Only 28 of 378 (7.4%) individuals were recaptured on a different building than the original capture. One individual switched buildings twice. Of the 27 individuals captured on two buildings, 20 (74.1%) switched to a neighboring building, ten of which touched or nearly touched the neighboring building. Five of the 27 (18.5%) returned to the original building and 118

PAGE 119

one of these switched back to the second building again. Overall, only eight of 378 (2.1%) moved two or more buildings away from their original point of capture. For H. mabouia six of 237 (2.5%) individuals were recaptured on a different building than the original capture. One of these moved back to the original capture building. Though this movement rate is substantially lower than for H. garnotii, any comparison should also consider two factors. First, unlike for H. garnotii in which I surveyed most buildings in close proximity, I only surveyed three buildings for H. mabouia and did not survey several other surrounding buildings to which movement could have occurred. Second, the sampling period for H. mabouia was 11 months shorter than for H. garnotii. Abundance The total number of H. garnotii, including juvenile, subadult, and adult size classes, captured on each building ranged from 14 to 428. The mean number of captures per individual for each building ranged from 1.2 to 2.7 captures per gecko with an overall mean of 1.9 (se = 0.108). The maximum number of individuals captured on each building in a single survey period ranged from 2 to 14 with a mean of 5.1 (se = 0.861) for all buildings except 2811 Tamiami Trail which had a high of 55 (Table 4-22). This building was much larger than the others and had an abundance of excellent quality hide spots. For H. mabouia, the total number of individuals, including juvenile, subadult, and adult size classes, captured was 176 on Building D, 217 on Building E, and 160 on Building F. The mean number of captures per individual for each building ranged from 1.8 to 2.0 with an overall mean of 1.9 (se = 0.058). The maximum number of individuals captured on each building in a single survey period ranged from 28 to 41 with a mean of 33.3 (se = 3.930) (Table 4-23). 119

PAGE 120

The above counts do not however, reflect capture probabilities and are likely biased low. To more accurately estimate abundance, I used previously estimated class-specific capture probabilities (Tables 4-11 and 4-12), to estimate the monthly mean number of subadults and adults per building for each species. For H. garnotii, I estimated the number of subadults on each building to range from 0.4 to 4.2 subadults/ month with a mean of 1.3 ( = 0.413) for all buildings except the most populated building (2811 Tamiami Trail) which had an estimated mean of 24.2 ( = 7.83) (Table 4-24). I estimated the number of H. mabouia subadults on each building to range from 18.3 to 26.5 subadults/ month with a mean of 22.2 ( = 7.3) (Table 4-25). For adults, I estimated the number of H. garnotii to range from 0.5 to 5.2 with a mean of 2.0 ( = 0.391) for all buildings except the most populated which had a mean of 35.7 adults/ month ( = 7.38) and for H. mabouia from 20.3 to 32.0 with an estimated mean of 26.0. ( = 7.4). I estimated the total subadult and adult population for H. garnotii to range from a mean of 1.0 to 9.4 geckos/ building/month with an overall mean of 3.3 (pooled = 0.413) for all buildings except 2811 Tamiami Trail which had a mean of 59.9 ( = 7.61) geckos per month. For H. mabouia, I estimated the total subadult and adult population to range from a mean of 38.6 to 53.8 geckos/ building/month with an overall mean of 48.2 (pooled = 7.5). ^se ^se ^se ^se ^se ^se ^se ^se ^se Adjusted abundance estimates of juveniles were not possible since the capture rate was fixed to 1. For H. garnotii, unadjusted mean juvenile abundances ranged from 0.03 to 0.73 juveniles/ building/ month with a mean of 0.22 (se = 0.05) for all buildings except 2811 Tamiami Trail which had a mean of 3.98 juveniles/ month. For H. mabouia, unadjusted mean juvenile abundances ranged from 1.48 to 2.27 juveniles/ building/ month with a mean of 1.88 (se = 0.23) 120

PAGE 121

Density For H. garnotii I estimated the mean subadult and adult densities to be 0.020 (se = .005, range = 0.005 to 0.068, n = 16) and 0.032 per linear meter of outer wall (se = .007, range = 0.005 to 0.100, n = 16), respectively. The overall mean was 0.052 (se = 0.012, range = 0.011 to 0.168, n = 16) H. garnotii per linear meter of outer wall (Table 4-26). For H. mabouia, I estimated the mean subadult and adult densities to be 0.200 (se = .021, range = 0.165 to 0.239, n = 3) and 0.234 per linear meter of outer wall (se = 0.030, range = 0.183 to 0.288, n = 3), respectively. The overall mean was 0.435 (se = 0.044, range = 0.348 to 0.485, n = 3) H. mabouia per linear meter (Table 4-27). The estimated mean density of H. mabouia was significantly greater than for H. garnotii (Subadults: t = -13.10, df = 17, p < 0.001; Adults: t = -9.83, df = 17, p < 0.001; Overall: t = -11.72, df = 17, p < 0.001). It should be noted that the actual difference may be somewhat smaller since the H. mabouia capture probabilities used to estimate abundance were less than 0.1 in nine of 30 periods and between 0.1 and 0.2 in five of 30 periods (Table 4-11). Such low capture probabilities did not allow precise estimation of abundance estimates for H. mabouia. Discussion Several traits common to both Hemidactylus garnotii and H. mabouia have promoted their successful colonization, establishment, and spread throughout large parts of Southwest Florida. Both species are human commensals, capable of rapid growth and maturity, and fecund. However, while similar, they are not the same. Small differences in their maturation rates, body size, reproductive output, or survival rates could lead to differential success when in sympatry. My theoretical maximum estimates of 13 clutches annually for H. garnotii and 16 for H. mabouia assumed a constant reproductive rate throughout the reproductive season. This 121

PAGE 122

assumption is unlikely to be true as it does not account for physiological constraints or environmental or other stresses which may limit clutch frequency (Du 2006). The field maxima I recorded of eight and four clutches annually are apt to be closer to realized maxima, although for H. mabouia may be underestimated due to low recapture rate. Previously published estimates based on follicle counts estimated at least three clutches annually for H. garnotii and a maximum of seven clutches annually for H. mabouia (Meshaka 1994, Meshaka et al. 1994a). This estimate for H. garnotii is much lower than the actual maximum count I recorded in this study. For H. mabouia this estimate does not account for the possibility of follicle reabsorption (Johnson 1960, Benabib 1994) which could lower actual counts. Regardless, the use of maximum values is not likely to be an accurate predictor of population wide values. Evaluation of reproductive output should instead focus on mean fecundity and fertility estimates which incorporate the proportion of breeding individuals in the population, the survival probability, and the sex ratio, which in this case offers a two-fold reproductive advantage to H. garnotii, due to parthenogenesis, over the sexually reproducing H. mabouia. At 0.904, the estimated mean monthly fecundity for H. garnotii is more than twice the estimate of 0.378 for H. mabouia. Both H. garnotii and H. mabouia are capable of rapid growth and maturity. H. garnotii can attain its 49 mm SVL maturation size in as little as 67 days (2.2 months) and on average in about 4.5 months while H. mabouia can attain its 51 mm SVL maturation size in as little as 100 days (3.3 months) and on average within 8 months. Rapid maturation could be advantageous as small juveniles are more vulnerable to predation. Further, rapid population growth could increase the availability of propagules for population spread. Compared to H. garnotii, fecundity may be depressed in H. mabouia due to a shortened activity period and reproductive season due to a greater intolerance to colder temperatures. 122

PAGE 123

Hemidactylus mabouia was more sensitive to low temperatures with capture rates dropping to just a few animals as temperatures fell to 21 o C and below while for H. garnotii such a reduction in activity was not seen until temperatures fell to around 12 o C. This resulted in overall H. mabouia activity being restricted to fewer months than H. garnotii. The approximate 6 to 8 week delay in H. mabouia peak juvenile and subadult capture rates as compared to H. garnotii is likely due to a greater and prolonged decrease in adult H. mabouia activity during the winter months. Though differences in hatching success could greatly influence overall fertility, I could not find any published data on the hatching success of these species. Selcer (1986) noted that hatching success of 100 eggs from the congeneric H. turcicus field collected in Southern Texas was 100%. Typically, however, reported rates are not that high. For instance rates of 83% to 91% have been reported for the lacertid lizard Podarcis bocagei (Galan 1997) and a natural rate of 59% and laboratory rate of 64% was reported for the non-indigenous gecko Lepidodactylus lugubris in Hawaii (Brown and Duffy 1992). While parthenogenesis could reduce hatching success (Lamb and Willey 1979, Corley and Moore 1999), no differences were found among parthenogenetic and sexual forms of the Australian gecko Heteronotia binoei (Kearney and Shine 2004). Further, egg predation by insects could be a significant factor as in Anolis limifrons from Central America (Andrews 1988). In most reproductive aspects, in this study, H. garnotii had a substantial reproductive advantage when compared to H. mabouia. But this was not the case for survival rates. At 0.283, I estimated juvenile monthly survival probability in H. garnotii to be about half the 0.570 monthly probability for H. mabouia. This difference considerably offsets the higher fecundity of H. garnotii as does a somewhat lower subadult survival probability (0.749 for H. garnotii and 123

PAGE 124

0.866 for H. mabouia). Estimated adult monthly survival rates for both species were fairly high and similar at 0.866 for H. mabouia and 0.857 for H. garnotii. Due to lower recapture probabilities, my estimates of abundance, and hence density, were not as accurate for H. mabouia as for H. garnotii. However, at ten times the average density of H. garnotii, the average density of H. mabouia far exceeded that of H. garnotii. Even the building with the highest H. garnotii density had less than half the density of any of the buildings with H. mabouia. Similarly, Meshaka (2000) noted that on several buildings that originally contained only H. garnotii, once colonized by H. mabouia had abundances of H. mabouia that far exceeded the precolonization abundance of H. garnotii. In sympatry, large differences in abundance could easily offset the reproductive advantages of H. garnotii. Numerical dominance has been implicated to play a part in the displacement of the non-indigenous Lepidodactylus lugubris by H. frenatus in Hawaii (Case et al. 1994). At just under 26 mm SVL, H. mabouia juveniles on average were significantly smaller than H. garnotii juveniles at 28 mm SVL. However, this pattern could be misleading if smaller H. garnotii juveniles were more wary and less likely to be observed or caught, or had lower survival rates than larger H. garnotii juveniles. Though smaller H. mabouia juvenile size, as compared to H. garnotii, could make them comparatively more vulnerable to predation or cannibalism, this does not appear to be important in that I determined H. mabouia juveniles to have more than twice the monthly survival rate of H. garnotii juveniles. Since the size difference of juvenile H. garnotii and H. mabouia is slight and growth rates are high, with a mean growth rate of 6 mm and maximum of 9 mm SVL per month, any size related disadvantages would likely be offset rapidly. Further, vulnerability due to small size could be buffered in the short run if ontogenetic changes exist in predator avoidance behaviors (Stamps 124

PAGE 125

1983, Keren-Rotem et al. 2006). However, such behavior could also reduce growth rate, furthering vulnerability, if accompanied by reduced food intake (Downes 2001). One behavior that may be employed to avoid predation without compromising access to food is autotomizing the tail. This would permit escape if the gecko was grasped by the tail or if the predator is distracted by the severed and wiggling tail (Congdon et al. 1974, Arnold 1984, 1994). In this study, I estimated that 13.5% of H. garnotii and 1.3% of H. mabouia juveniles were missing some part of their original tail. For H. garnotii this is slightly higher than the 10% previously reported (Cagle 1946). As suggested by Rand (1954) tail break frequencies can be used to compare the level of predation intensity among populations. More accurate estimations can be made when survival rates are incorporated into the estimation, under the assumption that tail breaks are due to failed predation attempts and not other causes such as intraspecific aggression (Schoener 1979). Using Schoeners method, I estimated predation intensities for juvenile H. garnotii and H. mabouia to be 1.459 and 0.570, respectively. This indicates 2.5 times the predation intensity on H. garnotii juveniles as compared to H. mabouia. This is consistent with my numerous observations of Cuban treefrogs, Osteopilus septentrionalis, a potential major predator of geckos (Meshaka 2001), on buildings in this study with H. garnotii while its presence on buildings with H. mabouia was less obvious. If smaller juveniles are more vulnerable to predation (Blomberg and Shine 2000, Head et al. 2002) then this could bias the mean juvenile H. garnotii SVL I presented. Further, the buildings with H. mabouia had an abundance of small hide spots that could not be accessed by frogs possibly reducing predation and allowing for a more accurate juvenile size estimation than for H. garnotii. Thus, the mean juvenile SVL I reported for H. garnotii may accurately reflect the population, however, it may be that a paucity of very small juvenile H. garnotii is an artifact of excessively high predation on H. 125

PAGE 126

garnotii thereby increasing the mean juvenile size I presented for this population. Though variation in predation levels among populations could bias the mean juvenile SVLs I reported, such predation differences are unlikely to be a factor in the displacement of H. garnotii. Competition could exclude H. garnotii from prime hidespots making them more vulnerable to predation, as was the case for L. lugubris by H. frenatus in Hawaii (Frogner 1967, as cited in Case et al. 1994). Nevertheless, presently there is no evidence to suggest that predation differentially affects these species in either allopatry or in sympatry. Though differences in juvenile size are slight, and may be of lesser consequence, the large differences seen between species in subadult and adult sizes could have important effects on survival probabilities and reproduction. Overall, H. mabouia was considerably larger than H. garnotii in both length and weight. When juveniles were excluded from the analysis, at 47.7 mm SVL and a weight of 2.00 grams the average H. mabouia is almost 13% longer and more than 90% heavier than the average H. garnotii at 42.3 mm SVL and 1.05 grams. Such differences could be substantial when competing for resources such as prime feeding locations (e.g. near lights) or prime escape or resting refugia. Such competition has been implicated in abundance reduction of the non-indigenous Hawaiian geckos H. garnotii and L. lugubris by H. frenatus (Bolger and Case 1992, Petren et al. 1993, Petren and Case 1996). Further, larger size could benefit H. mabouia by making it a more effective predator of smaller H. garnotii with even subadults potentially having an important effect. For example, I observed a subadult H. mabouia (SVL = 44 mm) prey on a juvenile H. garnotii (SVL = 26 mm) (8/8/04, 22:40, 25000 Harborside Boulevard, Punta Gorda, Charlotte County, 26.78698 o N / -82.03750 o W) indicating that predation by both subadults and adults could be a factor in H. garnotii displacement. While the population effect of predation on H. garnotii by H. mabouia is unknown, the importance of 126

PAGE 127

predation on juveniles has been implicated as a factor in the displacement of the native Anolis carolinensis by the non-indigenous A. sagrei in Florida (Campbell 2000, Gerber and Echternacht 2000). Predation of juveniles has also been documented on the non-indigenous Hawaiian gecko L. lugubris by H. frenatus. In laboratory experiments, Frogner (1967, as cited in Case et al. 1994) and Bolger and Case (1992) demonstrated that adult H. frenatus would eat juvenile L. lugubris but not the reverse. The authors hypothesized that the larger size of the juveniles compared to the adults may have influenced the inability of adults to prey on juveniles. Such asymmetrical intraspecific predation could also be present between H. garnotii and H. mabouia as H. mabouia are not only longer and heavier, but their mean head width to SVL ratio indicated that H. mabouia jaw width was nearly 20% larger than for H. garnotii (Klowden unpublished). It could be that H. mabouia can prey upon H. garnotii but not the reverse. Geckos of both species were almost always recaptured on the same building and often within meters of the original point of capture. Similar site fidelity was discussed for H. turcicus in Miami (Rose and Barbour 1968). Very few geckos appeared to move between my study buildings even when they were nearly touching. Analogous results were reported for H. turcicus in Texas and Tampa, Florida (Selcer 1986, Punzo 2001). This indicates that, similar to H. turcicus, dispersal by diffusion is likely not the dominant mode of dispersal (Trout and Schwaner 1994). However, I cannot account for movement out of my study area, including dispersal via vehicles. Jump dispersal may in fact be the most prevalent mode of dispersal as has been suggested for H. turcicus in Florida and Oklahoma (Meshaka 1995, Locey and Stone 2006). Although movements were recorded in all size classes, my data on this were too limited to allow for further analysis among size classes. 127

PAGE 128

128 This study indicated that dem ographic differences between H. garnotii and H. mabouia might have some bearing on their ability to coex ist. Differences in trai ts such as fecundity, juvenile survival rates, size, and abundance were substantial. Differences in other characters, such as subadult and adult surviv al rates and longevity were le ss pronounced but may still affect their survival in sympatry. Much work remain s to better understand if these differences in demographic and other life histor y characteristics are important in fluences on the distributions of H. garnotii and H. mabouia in Florida.

PAGE 129

Table 4-1. Description of buildings surveyed in Charlotte County. Address Latitude a Longitude a Year Built b Area c (m 2 ) Perimeter c (m) Accessible perimeter c (m) Notes 2841 Tamiami Tr 26.988424 W -82.108750 N 1984 178 59 59 2811 Tamiami Tr 26.988921 W -82.109515 N 1974 1,785 357 357 Linear dimension includes inner foyers 2785 Tamiami Tr 26.989408 W -82.110324 N 1983 241 71 71 2765 Tamiami Tr 26.989708 W -82.110538 N 1983 489 97 79 North side rear inaccessible until Hurricane Charley in August 2004 2751 Tamiami Tr ------------Excluded because fences made three sides inaccessible. 2745 Tamiami Tr 26.989963 W -82.110933 N 1979 149 50 38 Rear inaccessible until hurricane Charley in August 2004 2735 Tamiami Tr 26.990116 W -82.111175 N 1979 404 102 73 Rear inaccessible until hurricane Charley in August 2004 2733 Tamiami Tr 26.990250 W -82.111359 N 1982 502 93 93 2721 Tamiami Tr 26.990371 W -82.111540 N 1974 294 73 55 of northwest wall abuts southeast wall of 2715 2715 Tamiami Tr 26.990443 W -82.111646 N 1978 223 59 41 Southeast wall abuts 2721 2711 Tamiami Tr 26.990554 W -82.111774 N 1973 297 77 63 of northwest wall abuts southeast wall of 2705 2705 Tamiami Tr 26.990624 W -82.111878 N 1979 149 50 37 Southeast wall abuts 2711 2695 Tamiami Tr 26.990766 W -82.112066 N 1973 677 108 85 Most of northwest wall abuts southeast wall of 2691 2691 Tamiami Tr 26.990884 W -82.112232 N 1979 260 68 46 Southeast wall abuts 2695 2685 Tamiami Tr 26.990987 W -82.112367 N 1979 260 68 66 Small part of northwest wall abuts 2681 2681 Tamiami Tr 26.991047 W -82.112516 N 1974 328 80 46 Small part of southeast wall abuts 2685, northwest side abuts 2675, and of rear inaccessible 2675 Tamiami Tr 26.991110 W -82.112621 N 1975 297 73 49 Southeast side abuts 2681 129 (a) Latitudes and longitudes are in State Plane Florida West, NAD83 Meters, (b) Source: Charlotte County (2007a) (c) Estimated from Charlotte County website (Charlotte County 2007).

PAGE 130

Table 4-2. Description of buildings surveyed in Lee County. Address 130 Unit Latitude a Longitude a Year Built Area (m 2 ) Perimeter (m) 1342 Colonial Blvd D 26.595111 W -81.889833 N 1988 767 111 1342 Colonial Blvd E 26.594361 W -81.889806 N 1989 767 111 1342 Colonial Blvd F 26.594694 W -81.889556 N 1990 767 111 (a) Latitudes and longitudes are in State Plane Florida West, NAD83 Meters.

PAGE 131

Table 4-3. Summary of Charlotte County Hemidactylus garnotii surveys. Survey Date Survey Interval (days) New captures Recaptures Total captures 1 18-Aug-01 5 0 5 2 16-Sep-01 29 9 1 10 3 6-Nov-01 51 1 5 6 4 8-Mar-02 122 5 2 7 5 29-Apr-02 52 26 7 33 6 28-May-02 29 29 15 44 7 23-Jun-02 26 43 16 59 8 21-Jul-02 28 25 26 51 9 19-Aug-02 29 52 26 78 10 15-Sep-02 27 39 35 74 11 13-Oct-02 28 23 51 74 12 9-Nov-02 27 17 59 76 13 11-Dec-02 32 4 38 42 14 6-Jan-03 26 1 8 9 15 28-Jan-03 22 4 25 29 16 17-Feb-03 20 3 28 31 17 10-Mar-03 21 9 45 54 18 1-Apr-03 22 9 41 50 19 23-Apr-03 22 8 34 42 20 19-May-03 26 24 23 47 21 3-Jun-03 15 22 27 49 22 29-Jun-03 26 44 32 76 23 22-Jul-03 23 50 35 85 24 10-Aug-03 19 39 53 92 25 31-Aug-03 21 31 50 81 26 29-Sep-03 29 17 57 74 27 26-Oct-03 27 14 44 58 28 16-Nov-03 21 10 43 53 29 11-Dec-03 25 3 15 18 30 6-Jan-04 26 4 27 31 31 27-Jan-04 21 1 23 24 32 26-Feb-04 30 3 39 42 33 17-Mar-04 20 5 36 41 34 5-Apr-04 18 5 32 37 35 4-May-04 30 5 25 30 36 4-Jun-04 31 17 11 28 37 27-Jun-04 23 29 17 46 38 27-Jul-04 30 67 22 89 39 22-Sep-04 57 41 18 59 40 27-Oct-04 35 22 28 50 41 22-Nov-04 26 14 24 38 42 17-Feb-05 87 6 19 25 43 14-Mar-05 25 8 40 48 44 19-Apr-05 36 2 22 24 45 16-May-05 27 12 22 34 46 20-Jun-05 44 49 19 68 47 8-Aug-05 30 58 21 79 TOTAL 914 1286 2200 131

PAGE 132

Table 4-4. Summary of Lee County Hemidactylus mabouia surveys. Survey # Date Survey Interval (days) New captures Recaptures Total Captures 1 12-Mar-03 39 0 39 2 5-Apr-03 24 21 19 40 3 28-Apr-03 23 13 13 26 4 14-May-03 16 15 28 43 5 31-May-03 17 14 23 37 6 24-Jun-03 24 12 18 30 7 16-Jul-03 22 14 19 33 8 7-Aug-03 22 17 18 35 9 28-Aug-03 21 27 20 47 10 26-Sep-03 29 42 23 65 11 23-Oct-03 27 21 19 40 12 13-Nov-03 21 10 11 21 13 11-Dec-03 28 8 6 14 14 9-Jan-04 29 1 1 2 15 30-Jan-04 21 2 5 7 16 21-Feb-04 22 2 9 11 17 22-Mar-04 30 2 6 8 18 20-Apr-04 29 9 13 22 19 8-May-04 18 11 20 31 20 8-Jun-04 31 18 20 38 21 30-Jun-04 22 30 17 47 22 1-Aug-04 32 17 10 27 23 16-Sep-04 46 74 26 100 24 21-Oct-04 35 22 18 40 25 14-Nov-04 24 9 8 17 26 13-Feb-05 91 8 4 12 27 19-Mar-05 34 5 4 9 28 27-Apr-05 39 13 18 31 29 21-May-05 24 17 28 45 30 27-Jun-05 37 15 29 44 31 1-Aug-05 35 38 31 69 Total 546 484 1030 132

PAGE 133

Table 4-5. Proportion of total new captures for each stage. Absolute Proportion H. garnotii H. mabouia H. garnotii H. mabouia Juvenile 284 159 0.311 0.291 Subadult 454 225 0.497 0.412 Adult 176 162 0.193 0.297 Total 914 546 1.000 1.000 133

PAGE 134

Table 4-6. Number of each size class captured for each H. garnotii survey and the corresponding H. mabouia survey and the mean air temperatures for each survey. H. garnotii H. mabouia Date Juv a Sub b Adults ___oC c Date Juv a Sub b Adults ___oC c Combined ___oC d 29-Apr-02 5 3 25 24.2 24.2 28-May-02 10 13 21 25.4 25.4 23-Jun-02 17 21 21 26.8 26.8 21-Jul-02 8 24 18 27.3 27.3 19-Aug-02 11 40 27 27.7 27.7 15-Sep-02 4 56 14 28.1 28.1 13-Oct-02 1 47 26 26.1 26.1 9-Nov-02 1 39 35 23.5 23.5 11-Dec-02 1 17 24 19.7 19.7 6-Jan-03 1 4 4 11.4 11.4 28-Jan-03 0 12 17 14.5 14.5 17-Feb-03 0 8 23 16.9 16.9 10-Mar-03 1 18 35 22.6 12-Mar-03 0 11 28 24.2 23.4 1-Apr-03 3 9 38 16.0 5-Apr-03 0 9 31 25.6 20.8 23-Apr-03 6 1 34 23.9 28-Apr-03 0 5 21 24.4 24.2 19-May-03 18 5 24 26.3 14-May-03 2 9 32 27.2 26.8 3-Jun-03 14 11 24 26.9 31-May-03 6 3 28 27.5 27.2 29-Jun-03 23 22 31 28.6 24-Jun-03 6 2 22 26.8 27.7 22-Jul-03 20 38 27 27.5 16-Jul-03 6 6 21 28.1 27.8 10-Aug-03 18 51 23 26.2 7-Aug-03 9 10 16 27.9 27.0 31-Aug-03 5 51 25 27.9 28-Aug-03 11 21 15 28.5 28.2 29-Sep-03 1 48 25 24.6 26-Sep-03 14 35 16 26.8 25.7 26-Oct-03 3 26 29 25.0 23-Oct-03 9 20 11 25.2 25.1 16-Nov-03 4 18 31 22.2 13-Nov-03 6 9 6 21.9 22.0 11-Dec-03 2 5 11 13.7 14-Dec-03 3 8 3 17.4 15.6 6-Jan-04 0 9 22 16.2 9-Jan-04 0 1 1 14.5 15.4 27-Jan-04 0 11 13 13.5 30-Jan-04 1 2 4 18.5 16.0 26-Feb-04 0 18 24 15.5 21-Feb-04 1 4 6 20.8 18.1 17-Mar-04 1 8 32 20.1 22-Mar-04 0 4 4 19.3 19.7 5-Apr-04 1 4 32 19.4 20-Apr-04 0 12 10 21.9 20.6 4-May-04 2 3 25 21.9 8-May-04 1 12 18 24.1 23.0 4-Jun-04 11 5 12 26.2 8-Jun-04 3 10 25 27.3 26.8 27-Jun-04 19 11 16 28.1 30-Jun-04 10 9 28 28.2 28.1 27-Jul-04 32 42 15 27.6 1-Aug-04 8 6 13 27.1 27.4 22-Sep-04 2 38 19 26.2 16-Sep-04 23 44 33 29.0 27.6 27-Oct-04 4 21 25 24.0 21-Oct-04 7 16 17 26.2 25.1 22-Nov-04 2 15 21 21.2 14-Nov-04 5 6 6 21.4 21.3 17-Feb-05 0 3 22 12.4 13-Feb-05 4 6 2 18.0 15.2 14-Mar-05 1 6 41 21.2 19-Mar-05 1 6 2 18.9 20.0 19-Apr-05 1 1 22 20.3 27-Apr-05 1 9 21 23.5 21.9 16-May-05 8 4 22 25.5 21-May-05 0 25 20 26.1 25.8 20-Jun-05 16 23 29 27.5 27-Jun-05 8 7 29 27.9 27.7 8-Aug-05 11 43 25 28.1 1-Aug-05 18 12 39 29.2 28.7 (a) Juv=Juvenile, (b) Sub=subadult, (c) Mean air temperature in o C, (d) Average of mean temperatures in o C of corresponding H. garnotii and H. mabouia study sites. 134

PAGE 135

Table 4-7. Multistate goodness-of-fit test results and the estimated variance inflation factor () for survival probability estimation models. ^c 2a df b ^c c H. garnotii 229.306 280 0.819 H. mabouia 94.248 209 0.451 (a) Chi Squared goodness-of-fit test statistic for the Jolly-Move model calculated using program U-CARE version 2.2.5 (Choquet et al., 2005), (b) Model degrees freedom, (c) = ^c 2 /df Table 4-8. Survival probability model selection summary for H. garnotii. Model a AIC b w c k d S(g)p(g*t)(g) 0.00 1.00 88 S(g)p(t) (g) 21.40 0.00 46 S(g j,s=a )p(g*t) (g) 30.56 0.00 87 S(g j=s,a )p(g*t) (g) 86.30 0.00 87 S(g)p(g)(g) 97.41 0.00 6 S(g*t)p(g*t) (g) 100.69 0.00 209 S(.)p(g*t)(g) 181.86 0.00 86 (a) S=survival probability, p=recapture probability, = transition probability. Parentheses contain the factor for the associated parameter: g indicates probabilities for each stage were estimated separately or grouped as indicated by superscripts (j=juvenile, s=subadult, a=adult) and were time constant, g*t indicates probabilities were estimated separately for each stage and capture period (time-specific), indicates probabilities were estimated time constant and stage constant, (b) QAIC = difference in QAIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model. Table 4-9. Survival probability model selection summary for H. mabouia. Model a AIC b w c k d S(g j,s=f=m )p(t) (.) 0.00 0.52 33 S(g j,s=f=m )p(t) (g) 2.06 0.19 34 S(g j,s,f=m )p(t) (.) 2.10 0.18 34 S(g)p(t) (.) 3.27 0.10 35 S(g j=s,f=m )p(t) (.) 11.00 0.00 33 S(.)p(t)(.) 11.08 0.00 32 S(g j,s=f=m )p(g*t)(.) 42.27 0.00 93 S(g j,s=f=m *t)p(t)(.) 75.93 0.00 90 S(g j,s=f=m )p(.)(.) 196.33 0.00 4 (a) S=survival probability, p=recapture probability, = transition probability. Parentheses contain the factor for the associated parameter: g indicates probabilities for each stage were estimated separately or grouped as indicated by superscripts (j=juvenile, s=subadult, m=male adult, f=female adult) and were time constant, g*t indicates probabilities were estimated separately for each stage and capture period (time-specific), indicates probabilities were estimated time constant and stage constant, (b) QAIC = difference in QAIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model 135

PAGE 136

Table 4-10. H. garnotii survival, recapture, and transition probability estimates. 95% Confidence interval Parameter a Stage b Probability estimate for model S(g)p(g*t)(g) c Standard error Lower Upper S Juvenile 0.283 0.035 0.219 0.357 S Subadult 0.749 0.016 0.716 0.779 S Adult 0.857 0.008 0.840 0.873 p Juvenile 0 0 0 0 Fixed p Subadult 1.000 2.68E-06 1.000 1.000 p Subadult 0.356 0.156 0.127 0.678 p Subadult 0.439 0.124 0.226 0.678 p Subadult 0.380 0.115 0.190 0.616 p Subadult 0.690 0.090 0.493 0.836 p Subadult 0.684 0.084 0.502 0.822 p Subadult 0.748 0.097 0.521 0.890 p Subadult 0.429 0.089 0.269 0.606 p Subadult 0.133 0.063 0.050 0.308 p Subadult 0.355 0.098 0.192 0.561 p Subadult 0.313 0.101 0.154 0.533 p Subadult 0.911 0.083 0.580 0.987 p Subadult 0.568 0.191 0.222 0.858 p Subadult 1.22E-13 2.03E-07 -3.98E-07 3.98E-07 p Subadult 0.283 0.275 0.027 0.849 p Subadult 0.590 0.202 0.218 0.881 p Subadult 0.589 0.173 0.260 0.853 p Subadult 0.478 0.110 0.279 0.685 p Subadult 0.655 0.086 0.474 0.800 p Subadult 0.529 0.080 0.374 0.678 p Subadult 0.719 0.082 0.537 0.850 p Subadult 0.432 0.095 0.263 0.619 p Subadult 0.448 0.107 0.258 0.655 p Subadult 0.134 0.074 0.042 0.351 p Subadult 0.424 0.132 0.204 0.679 p Subadult 0.564 0.129 0.316 0.784 p Subadult 0.926 0.070 0.624 0.990 p Subadult 0.546 0.170 0.239 0.821 p Subadult 0.378 0.188 0.113 0.744 p Subadult 0.235 0.221 0.027 0.773 p Subadult 0.309 0.296 0.029 0.871 p Subadult 0.151 0.147 0.019 0.627 p Subadult 0.635 0.172 0.289 0.882 p Subadult 0.370 0.132 0.162 0.641 p Subadult 0.452 0.145 0.208 0.721 p Subadult 0.536 0.157 0.250 0.800 p Subadult 0.276 0.178 0.062 0.685 p Subadult 1.000 3.17E-05 1.000 1.000 136

PAGE 137

Table 4-10 Continued. 95% Confidence interval Parameter a Stage b Probability estimate for model S(g)p(g*t)(g) c Standard error Lower Upper P Subadult 0.391 0.341 0.037 0.914 p Subadult 5.72E-12 2.57E-06 -5.03E-06 5.03E-06 p Subadult 0.536 0.338 0.075 0.943 p Subadult 0.257 0.108 0.103 0.511 p Adult 0.494 0.109 0.293 0.697 p Adult 0.437 0.102 0.256 0.637 p Adult 0.423 0.095 0.255 0.610 p Adult 0.466 0.088 0.303 0.636 p Adult 0.201 0.059 0.109 0.340 p Adult 0.362 0.066 0.245 0.497 p Adult 0.451 0.063 0.332 0.575 p Adult 0.302 0.056 0.205 0.421 p Adult 0.055 0.027 0.021 0.139 p Adult 0.231 0.052 0.145 0.346 p Adult 0.306 0.057 0.207 0.427 p Adult 0.442 0.062 0.326 0.565 p Adult 0.507 0.064 0.385 0.629 p Adult 0.455 0.062 0.338 0.577 p Adult 0.339 0.061 0.232 0.466 p Adult 0.319 0.061 0.213 0.448 p Adult 0.415 0.066 0.294 0.548 p Adult 0.366 0.063 0.253 0.496 p Adult 0.334 0.059 0.229 0.457 p Adult 0.343 0.060 0.237 0.468 p Adult 0.315 0.058 0.213 0.438 p Adult 0.389 0.062 0.277 0.514 p Adult 0.364 0.060 0.257 0.487 p Adult 0.129 0.039 0.070 0.225 p Adult 0.255 0.052 0.167 0.369 p Adult 0.180 0.046 0.107 0.288 p Adult 0.358 0.061 0.248 0.485 p Adult 0.450 0.066 0.327 0.580 p Adult 0.449 0.068 0.322 0.583 p Adult 0.406 0.072 0.277 0.550 p Adult 0.207 0.064 0.108 0.359 p Adult 0.374 0.084 0.228 0.547 p Adult 0.338 0.087 0.193 0.523 p Adult 0.328 0.094 0.175 0.529 p Adult 0.443 0.085 0.288 0.611 p Adult 0.284 0.068 0.171 0.433 p Adult 0.399 0.078 0.260 0.558 p Adult 0.792 0.066 0.634 0.894 137

PAGE 138

Table 4-10 Continued. 95% Confidence interval Parameter a Stage b Probability estimate for model S(g)p(g*t)(g) c Standard error Lower Upper p Adult 0.484 0.081 0.332 0.639 p Adult 0.587 0.094 0.399 0.753 p Adult 0.542 0.106 0.339 0.733 p Adult 0.388 0.085 0.239 0.562 Juvenile to Subadult 1.000 0.000 1.000 1.000 Fixed Subadult to Adult 0.218 0.017 0.187 0.252 (a) S=survival, p=recapture, =transition (b) Parentheses contain the factor for the associated parameter: S(g) indicates survival probabilities for each stage were estimated separately; p(g*t) indicates recapture probabilities were estimated separately for each stage and capture period; (g) indicates transition probabilities were estimated separately for each stage but time constant. 138

PAGE 139

139 Table 4-11. H. mabouia survival, recapture, and tran sition probability estimates. 95% Confidence interval Parameter a Stage Probability estimate for model S(g j, s=m=f )p(t) (.) b Standard error Lower Upper S Juvenile 0.570 0.074 0.422 0.706 S Subadult/adult 0.866 0.009 0.849 0.882 p Juvenile 0.000 0.000 0.000 0.000 fixed p Subadult/adult 0.547 0.088 0.376 0.707 p Subadult/adult 0.271 0.066 0.163 0.416 p Subadult/adult 0.511 0.072 0.372 0.648 p Subadult/adult 0.363 0.063 0.251 0.492 p Subadult/adult 0.262 0.054 0.169 0.381 p Subadult/adult 0.264 0.054 0.173 0.381 p Subadult/adult 0.222 0.049 0.141 0.331 p Subadult/adult 0.231 0.048 0.150 0.338 p Subadult/adult 0.244 0.046 0.165 0.344 p Subadult/adult 0.154 0.034 0.098 0.233 p Subadult/adult 0.081 0.025 0.044 0.146 p Subadult/adult 0.052 0.021 0.023 0.112 p Subadult/adult 0.009 0.009 0.001 0.062 p Subadult/adult 0.050 0.022 0.021 0.115 p Subadult/adult 0.087 0.030 0.044 0.166 p Subadult/adult 0.073 0.029 0.033 0.154 p Subadult/adult 0.162 0.044 0.093 0.267 p Subadult/adult 0.262 0.053 0.172 0.377 p Subadult/adult 0.261 0.055 0.169 0.381 p Subadult/adult 0.186 0.045 0.113 0.291 p Subadult/adult 0.111 0.034 0.060 0.197 p Subadult/adult 0.317 0.058 0.216 0.439 p Subadult/adult 0.149 0.033 0.094 0.227 p Subadult/adult 0.064 0.022 0.032 0.123 p Subadult/adult 0.047 0.023 0.018 0.120 p Subadult/adult 0.052 0.026 0.019 0.132 p Subadult/adult 0.271 0.058 0.172 0.399 p Subadult/adult 0.400 0.066 0.281 0.533 p Subadult/adult 0.393 0.066 0.274 0.527 p Subadult/adult 0.415 0.070 0.287 0.555 Juvenile to adult 1.000 0.000 1.000 1.000 fixed Juvenile to adult 0.000 0.000 0.000 0.000 fixed Subadult to juvenile 0.00 0 0.000 0.000 0.000 fixed Subadult to adult 0.130 0.012 0.108 0.155 Adult to juvenile 0.000 0.000 0.000 0.000 fixed Adult to subadult 0.000 0.000 0.000 0.000 fixed (a) S=survival, p=recapture, =transition (b) Parentheses contain the factor for the associated parameter: S(g j,s=m=f ) indicates the juvenile (j ) survival probability was estimated separately from the subadult (s) and male (m) and female (f) adults which were estimated together; p(t) indicates recapture probabilities were estimated time specific and stage constant; (.) indicates transition probabilities were estimated time constant and stage constant.

PAGE 140

140 Table 4-12. Mean temperature ( o C) and monthly growth rate (mm/day) for non-adult, adult, and all H. garnotii and H. mabouia. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Mean Mean temperature ( o C) 63.0 66.0 70.5 73.0 79.0 82.3 82.8 83.2 82.3 78.5 71.8 64.3 Growth Rates (mm/d a 11 768 1 11265 0 111 2059 y) H. garnotii all Mean 0.026 0.040 0.049 0.083 0.206 0.236 0.197 0.217 0.151 0.083 0.042 0.038 0.147 Standard error 0.009 0.019 0.012 0.018 0.028 0.014 0.013 0.013 0.009 0.009 0.007 0.000 0.006 Sample size 7 11 14 22 13 42 36 28 55 25 19 1 273 H. garnotii nonadult Mean 0.033 0.134 0.091 0.159 0.242 0.267 0.215 0.224 0.163 0.090 0.050 0.038 0.183 Standard error 0.012 0.018 0.045 0.044 0.026 0.012 0.011 0.012 0.010 0.010 0.009 0.000 0.007 Sample size 4 3 2 7 10 34 32 27 43 21 12 1 196 H. garnotii adult Mean 0.016 0.005 0.036 0.048 0.089 0.080 0.023 0.037 0.071 0.046 0.028 0.043 Standard error ize 0.016 0.005 3 0.010 8 0.009 1 0.022 5 0.021 3 0.011 6 0.000 3 0.014 1 0.018 7 0.008 4 0.005 0 Sample s H. mabouiaall Mean 0.045 0.033 0.096 0.106 0.188 0.123 0.149 0.209 0.132 0.090 0.038 0.135 Standard error 0.000 0.000 0.015 0.015 0.025 0.022 0.030 0.026 0.017 0.025 0.038 0.008 Sample size 1 1 17 30 20 17 8 18 18 6 2 0 138 H. mabouianonadult Mean 0.045 0.033 0.147 0.160 0.302 0.243 0.210 0.260 0.151 0.125 0.038 0.191 Standard error ize 0.000 0.000 1 0.031 1 0.025 5 0.041 2 0.019 7 0.031 4 0.017 4 0.017 3 0.018 2 0.038 4 0.012 0 Sample s H. mabouiaadult Mean 0.063 0.071 0.110 0.066 0.071 0.034 0.029 0.021 0.068 Standard error ize 0.013 0 0.014 0 0.010 8 0.014 9 0.033 1 0.011 3 0.029 4 0.021 3 0.007 0 Sample s

PAGE 141

141 ^c Table 4-13. Multistate goodness-of-fit test results and the estimated variance inflation factor () for breeding proportion estimation models. 2a df b ^c c H. garnotii 161.300 247 0.653 H. mabouia 23.465 47 0.499 (a) Chi Squared goodness-of-fit test statistic for the Jolly-Move model calculated using program U-CARE version 2.2.5 (Choquet et al., 2005), (b) Model degrees freedom, (c) c= ^2 /df Table 4-14. Breeding proportion model selection summary for H. garnotii. Model a AIC AIC b w c Model Likelihood k d Deviance S(.)p(g)(.) 2568.94 0 0.47 1.00 4 1909.26 S(g)p(g)(.) 2569.22 0.28 0.41 0.87 5 1907.51 S(.)p(g)(g) 2571.79 2.86 0.11 0.24 4 1912.12 S(.)p(gt)(.) 2587.12 18.18 0.00 0.00 58 1808.41 S(.)p(.)(.) 2588.23 19.29 0.00 0.00 3 1930.58 S(.)p(g)(t) 2606.92 37.98 0.00 0.00 31 1890.22 S(t)p(g)(.) 6899.60 4330.66 0.00 0.00 30 6185.10 (a) S = survival probability, p = recapture probability, = transition probability. Parentheses contain the factor for the associated parameter: g indicates probabilities for each stage were estimated separately and were time constant, t indicates probabilities were estimated separately for each capture period, gt indicates probabilities were estimated separately for each stage and capture period (time-specific), indicates probabilities were estimated time constant and stage constant, (b) AIC = difference in AIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model. Table 4-15. Breeding proportion model selection summary for H. mabouia. Model a AIC AIC b w c Model Likelihood k d Deviance S(.)p(g)(.) 543.35 0.00 0.35 1.00 4 436.25 S(.)p(.)(.) 544.26 0.92 0.22 0.63 3 439.30 S(.)p(g)(g) 545.15 1.80 0.14 0.41 5 435.88 S(g)p(g)(.) 545.51 2.16 0.12 0.34 5 436.25 S(g)p(g)(g) 547.35 4.00 0.05 0.14 6 435.88 S(t)p(g)(.) 1682.80 1139.45 0.00 0.00 31 1501.15 (a) S = survival probability, p = recapture probability, = transition probability. Parentheses contain the factor for the associated parameter: g indicates probabilities for each stage were estimated separately and were time constant, t indicates probabilities were estimated separately for each capture period, gt indicates probabilities were estimated separately for each stage and capture period (time-specific), indicates probabilities were estimated time constant and stage constant, (b) AIC = difference in AIC value from the highest ranking model, (c) Akaike model weight, (d) number of parameters in each model.

PAGE 142

Table 416. Parameter estimates for breeding proportion models for H. garnotii. Model a S r b S n b p r c p n c r n d nr d s(.)p(g)(.) 0.833 0.833 0.344 0.539 0.468 0.468 se (0.012) (0.012) (0.025) (0.031) (0.028) (0.028) 95% CI (0.809 0.856) (0.809 0.856) (0.297 0.394) (0.478 0.598) (0.413 0.524) (0.413 0.524) s(g)p(g)(.) 0.802 0.864 0.342 0.534 0.474 0.474 se (0.028) (0.026) (0.025) (0.030) (0.029) (0.029) 95% CI (0.742 0.851) (0.805 0.907) (0.295 0.393) (0.474 0.593) (0.418 0.531) (0.418 0.531) s(.)p(g)(g) 0.835 0.835 0.36 0.488 0.524 0.524 se (0.012) (0.012) (0.107) (0.114) (0.177) (0.177) 95% CI (0.811 0.857) (0.811 0.857) (0.185 0.584) (0.280 0.700) (0.215 0.815) (0.215 0.815) Model Averaged Estimates 0.821 0.846 0.345 0.531 0.477 0.473 se (0.025) (0.024) (0.043) (0.051) (0.068) (0.047) 95% CI (0.765 0.865) (0.793 0.888) (0.265 0.434) (0.432 0.628) (0.349 0.608) (0.382 0.565) 142 (a) S = survival probability, p = recapture probability, = transition probability. Parentheses contain the factor for the associated parameter: g indicates probabilities for each stage were estimated separately and were time constant, t indicates probabilities were estimated separately for each capture period, gt indicates probabilities were estimated separately for each stage and capture period (time-specific), indicates probabilities were estimated time constant and stage constant, (b) S r = survival probability for gravid females, S n = survival probability for non-gravid adult females (c) p r = recapture probability for gravid adult females, p n = recapture probability for non-gravid adult females, (d) r n = transition probability from gravid state to non-gravid state, nr = transition probability from non-gravid state to gravid state.

PAGE 143

143 Table 417. Parameter estimates for breeding proportion models for H. mabouia. Model a S r b S n b p r c p n c r n d nr d S(.)p(g)(.) 0.857 0.857 0.259 0.172 0.582 0.582 se (0.022) (0.022) (0.041) (0.033) (0.090) (0.090) 95% CI (0.807 0.895) (0.807 0.895) (0.187 0.347) (0.117 0.246) (0.402 0.742) (0.402 0.742) S(.)p(.)(.) 0.857 0.857 0.215 0.215 0.565 0.565 se (0.022) (0.022) (0.028) (0.028) (0.097) (0.097) 95% CI (0.807 0.896) (0.807 0.896) (0.166 0.274) (0.166 0.274) (0.374 -0.739) (0.374 0.739) S(.)p(g)(g) 0.856 0.856 0.340 0.140 0.719 0.448 se (0.022) (0.022) (0.181) (0.050) (0.193) (0.263) 95% CI (0.807 0.895) (0.807 0.895) (0.095 0.715) (0.067 0.270) (0.282 0.943) (0.092 0.867) S(g)p(g)(.) 0.851 0.863 0.259 0.173 0.582 0.582 se (0.086) (0.084) (0.041) (0.033) (0.090) (0.090) 95% CI (0.603 0.956) (0.609 0.962) (0.186 0.347) (0.117 0.247) (0.402 0.742) (0.402 0.742) S(g)p(g)(g) 0.856 0.856 0.340 0.140 0.719 0.448 se (0.105) (0.066) (0.182) (0.051) (0.195) (0.263) 95% CI (0.528 0.969) (0.676 0.945) (0.095 0.715) (0.067 0.272) (0.279 0.944) (0.092 0.867) Model Averaged Estimates 0.856 0.858 0.265 0.176 0.607 0.549 se (0.045) (0.040) (0.100) (0.045) (0.135) (0.156) 95% CI (0.745 0.924) (0.760 0.920) (0.116 0.497) (0.105 0.281) (0.337 0.824) (0.261 0.807) (a) S = survival probability, p = recapture probability, = transition probability. Parentheses contain the factor for the associated parameter: g indicates probabilities for each stage were estimated separately and were time constant, t indicates probabilities were estimated separately for each capture period, gt indicates probabilities were estimated separately for each stage and capture period (time-specific), indicates probabilities were estimated time constant and stage constant, (b) S r = survival probability for gravid females, S n = survival probability for non-gravid adult females (c) p r = recapture probability for gravid adult females, p n = recapture probability for non-gravid adult females, (d) r n = transition probability from gravid state to non-gravid state, nr = transition probability from non-gravid state to gravid state.

PAGE 144

Table 4-18. Number of gravid and nongravid adult female H. garnotii captured each survey and the nave and adjusted breeding proportion. Survey number Gravid Nongravid Total Nave estimate Adjusted estimate 19 11 23 34 0.324 0.424 20 8 16 24 0.333 0.435 21 7 17 24 0.292 0.388 22 12 19 31 0.387 0.493 23 5 22 27 0.185 0.259 24 10 13 23 0.435 0.542 25 11 14 25 0.440 0.547 26 7 18 25 0.280 0.374 27 8 21 29 0.276 0.370 28 12 19 31 0.387 0.493 29 2 9 11 0.182 0.255 30 9 13 22 0.409 0.516 31 5 8 13 0.385 0.490 32 7 17 24 0.292 0.388 33 19 13 32 0.594 0.692 34 16 16 32 0.500 0.606 35 8 17 25 0.320 0.420 36 3 9 12 0.250 0.339 37 6 10 16 0.375 0.480 38 4 11 15 0.267 0.359 39 5 14 19 0.263 0.355 40 7 18 25 0.280 0.374 41 8 13 21 0.381 0.486 42 12 10 22 0.545 0.649 43 18 23 41 0.439 0.546 44 6 16 22 0.273 0.366 45 13 9 22 0.591 0.690 46 14 15 29 0.483 0.590 47 3 22 25 0.120 0.173 Mean 0.355 0.452 144

PAGE 145

145 Table 4-19. Number of gravid and nongravid adult female H. mabouia captured each survey and the nave and adjusted breeding proportion. Survey number Gravid Nongravid Total Nave estimate Adju sted estimate 3 8 4 12 0.667 0.571 4 11 4 15 0.733 0.646 5 8 4 12 0.667 0.571 6 7 1 8 0.875 0.823 7 4 3 7 0.571 0.470 8 2 2 4 0.500 0.399 9 4 4 8 0.500 0.399 10 6 1 7 0.857 0.799 11 1 3 4 0.250 0.181 12 0 3 3 0.000 0.000 13 0 0 0 0.000 0.000 14 0 0 0 0.000 0.000 15 0 0 0 0.000 0.000 16 2 0 2 1.000 1.000 17 2 0 2 1.000 1.000 18 0 0 0 0.000 0.000 19 2 1 3 0.667 0.571 20 2 2 4 0.500 0.399 21 6 2 8 0.750 0.666 22 0 1 1 0.000 0.000 23 0 6 6 0.000 0.000 24 1 3 4 0.250 0.181 25 1 0 1 1.000 1.000 26 0 1 1 0.000 0.000 27 0 0 0 0.000 0.000 28 3 2 5 0.600 0.499 29 1 2 3 0.333 0.249 30 3 4 7 0.429 0.332 31 3 8 11 0.273 0.199 Mean 0.428 0.378

PAGE 146

146 Table 4-20. Amount of original tail missing at the first capture of each gecko for H. garnotii and H. mabouia. Absolute amount Proportion of stage Proportion of total Amount of tail missing Juvenile Subadult Adult All Juvenile Subadult Adult Juvenile Subadult Adult All H. garnotii >2/3 5 32 38 75 0.018 0.072 0.224 0.006 0.036 0.043 0.084 2/3 4 21 25 50 0.014 0.048 0.147 0.004 0.024 0.028 0.056 1/2 6 27 23 56 0.021 0.061 0.135 0.007 0.030 0.026 0.063 1/3 13 42 16 71 0.046 0.095 0.094 0.015 0.047 0.018 0.080 Tip 10 38 9 57 0.036 0.086 0.053 0.011 0.043 0.010 0.064 Any 38 160 111 309 0.135 0.362 0.653 0.043 0.179 0.124 0.346 None 243 282 59 584 0.865 0.638 0.347 0.272 0.316 0.066 0.654 Total 281 442 170 893 1.000 1.000 1.000 0.315 0.495 0.190 1.000 H. mabouia >2/3 0 9 22 31 0.000 0.042 0.138 0.000 0.017 0.041 0.058 2/3 1 14 6 21 0.006 0.065 0.038 0.002 0.026 0.011 0.039 1/2 0 5 10 15 0.000 0.023 0.063 0.000 0.009 0.019 0.028 1/3 0 6 11 17 0.000 0.028 0.069 0.000 0.011 0.021 0.032 Tip 1 10 7 18 0.006 0.046 0.044 0.002 0.019 0.013 0.034 Any 2 44 56 102 0.013 0.204 0.352 0.004 0.083 0.105 0.192 None 155 172 103 430 0.987 0.796 0.648 0.291 0.323 0.194 0.808 Total 157 216 159 532 1.000 1.000 1.000 0.295 0.406 0.299 1.000

PAGE 147

Table 4-21. Amount of tail missing for male and female adult H. mabouia at the first capture for each gecko. Absolute amount Proportion of sex Proportion of total adults Amount of tail missing Males Females All Males Females Males Females All >2/3 15 7 22 0.185 0.092 0.096 0.045 0.140 2/3 2 4 6 0.025 0.053 0.013 0.025 0.038 1/2 4 6 10 0.049 0.079 0.025 0.038 0.064 1/3 8 3 11 0.099 0.039 0.051 0.019 0.070 Tip 7 0 7 0.086 0.000 0.045 0.000 0.045 Any 36 20 56 0.444 0.263 0.229 0.127 0.357 None 45 56 101 0.556 0.737 0.287 0.357 0.643 Total 81 76 157 1.000 1.000 0.516 0.484 1.000 Table 4-22. Number of H. garnotii captured on each survey building. Building address a 2675 2681 2685 2691 2695 2705 2711 2715 Total # of captures 43 59 30 26 72 33 93 27 Total # of individuals 28 26 17 16 44 19 40 14 Mean # of captures per individual 1.54 2.27 1.76 1.63 1.64 1.74 2.33 1.93 Minimum # of individuals captured per survey 0 0 0 0 0 0 0 0 Maximum # of individuals captured per survey 6 5 2 2 5 5 7 3 # of survey periods 43 43 43 43 43 43 43 43 Mean # of captures per survey 1.00 1.37 0.70 0.60 1.67 0.77 2.16 0.63 Standard Error (0.19) (0.20) (0.11) (0.09) (0.22) (0.18) (0.24) (0.12) Table 4-22 Continued. Building address a 2721 2733 2735 2745 2765 2785 2811 2841 Total # of captures 23 45 21 22 189 170 1172 147 Total # of individuals 15 20 18 14 100 77 428 57 Mean # of captures per individual 1.53 2.25 1.17 1.57 1.89 2.21 2.74 2.58 Minimum # of individuals captured per survey 0 0 0 0 0 0 6 0 Maximum # of individuals captured per survey 2 4 3 2 14 9 55 8 # of survey periods 43 43 43 43 43 43 43 43 Mean # of captures per survey 0.53 1.05 0.49 0.51 4.40 3.95 27.26 3.42 Standard Error (0.11) (0.16) (0.11) (0.11) (0.40) (0.34) (1.76) (0.31) (a) Address on Tamiami Trail in Port Charlotte, Florida. 147

PAGE 148

Table 4-23. Number of H. mabouia captured on each survey building. Building Unit a D E F Total # of captures 347 385 298 Total # of individuals 176 217 160 Mean # of captures per individual 1.97 1.77 1.86 Minimum # of individuals captured per survey 1 1 1 Maximum # of individuals captured per survey 31 28 41 # of survey periods 31 31 31 Mean # of captures per survey 11.57 12.42 9.93 Standard Error (1.16) (1.29) (1.52) Table 4-24. Estimated mean number of subadult and adult H. garnotii on each survey building during each survey period (SE). Building address a 2675 2681 2685 2691 2695 2705 2711 2715 Mean # of subadults 0.9 0.6 0.7 0.4 1.2 0.7 1.6 0.6 (0.31) (0.17) (0.40) (0.08) (0.29) (0.16) (0.41) (0.12) Mean # of adults 1.3 2.8 0.7 0.7 1.8 0.9 3.3 0.8 (0.23) (0.53) (0.14) (0.13) (0.34) (0.18) (0.66) (0.18) 2.2 3.4 1.4 1.1 3.1 1.6 4.9 1.4 Mean # of subadults and adults (0.27) (0.39) (0.30) (0.11) (0.32) (0.17) (0.55) (0.15) Table 4-24 Continued. Building address a 2721 2733 2735 2745 2765 2785 2811 2841 Mean # of subadults 0.4 1.1 0.6 0.4 4.2 3.0 24.2 2.7 (0.12) (0.38) (0.24) (0.10) (1.56) (0.88) (7.83) (0.98) Mean # of adults 0.6 1.5 0.5 0.7 5.2 4.9 35.7 4.7 (0.10) (0.30) (0.08) (0.13) (0.96) (1.02) (7.38) (0.88) 1.0 2.6 1.1 1.1 9.4 8.0 59.9 7.4 Mean # of subadults and adults (0.11) (0.34) (0.18) (0.12) (1.30) (0.95) (7.61) (0.93) (a) Address on Tamiami Trail in Port Charlotte, Florida. Table 4-25. Estimated mean number of subadult and adult H. mabouia on each survey building during each survey period (SE). Building Unit D E F Mean # of subadults 26.5 21.8 18.3 (10.4) (6.4) (5.1) Mean # of adults 25.8 32.0 20.3 (6.4) (10.8) (5.1) Total # of subadult and adults 52.3 53.8 38.6 (8.6) (8.9) (5.1) 148

PAGE 149

Table 4-26. Estimated density of H. garnotii per linear meter of outer wall surface. Building address a 2675 2681 2685 2691 2695 2705 2711 2715 Outer wall linear meters 73 80 68 68 108 50 77 59 Density (geckos/linear meter) Subadults 0.012 0.008 0.010 0.006 0.011 0.014 0.021 0.010 Adults 0.018 0.035 0.010 0.010 0.017 0.018 0.043 0.014 Subadults and adults 0.030 0.043 0.021 0.016 0.029 0.032 0.064 0.024 Table 4-26 Continued. Building address a 2721 2733 2735 2745 2765 2785 2811 2841 Outer wall linear meters 73 93 102 50 97 71 357 59 Density (geckos/linear meter) Subadults 0.005 0.012 0.006 0.008 0.043 0.042 0.068 0.046 Adults 0.008 0.016 0.005 0.014 0.054 0.069 0.100 0.080 Subadults and adults 0.014 0.028 0.011 0.022 0.100 0.070 0.168 0.125 (a) Address on Tamiami Trail in Port Charlotte, Florida. Table 4-27. Estimated density of H. mabouia per linear meter of outer wall surface. Building Unit D E F Outer wall linear feet 111 111 111 Density (geckos/linear meter) Subadults 0.239 0.196 0.165 Adults 0.232 0.288 0.183 Subadults and adults 0.471 0.485 0.348 149

PAGE 150

150 Figure 4-1. Aerial view of Charlotte County study area.

PAGE 151

Figure 4-2. Aerial views of Lee County study area. Figure 4-3. Ground level views of Lee County study area. 151

PAGE 152

152 Figure 4-4. Peering behind gutters with helmet mounted light and holding extendable pole with squeegee mounted at the end. Figure 4-5. Fold out rulers on end of extendable painters pole.

PAGE 153

Figure 4-6. Number of H. garnotii captures for each snout-vent length (mm) for new captures and all captures (new plus recaptures). 153 Figure 4-7. Number of H. mabouia captures for each snout-vent length (mm) for new captures and all captures (new plus recaptures).

PAGE 154

0102030405060708090100AprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulMonth/ YearNumber Captured1012141618202224262830Mean Temperature (C) H. garnotii captures H. mabouia captures Mean Temperature 2004 20 0 3 2005 2 002 154 Figure 4-8. Number of H. garnotii and H. mabouia captured throughout the study and the mean temperature ( o C).

PAGE 155

05101520253035AprMa y JunJulAu g Se p OctNovDecJanFebMarAprMa y JunJulAu g Se p OctNovDecJanFebMarAprMa y JunJulAu g Se p OctNovDecJanFebMarAprMa y JunJulMonth/ yearNumber Captured10.012.014.016.018.020.022.024.026.028.030.0 H. garnotii Juveniles H. mabouia Juveniles Mean Temperature 2 002 20 0 3 2004 2005 155 Figure 4-9. Number of juvenile H. garnotii and H. mabouia captured and the mean temperature throughout the study.

PAGE 156

0102030405060AprMa y JunJulAu g Se p OctNo v De c JanFebMarAprMa y JunJulAu g Se p OctNo v De c JanFebMarAprMa y JunJulAu g Se p OctNo v De c JanFebMarAprMa y JunJulMonth/ yearNumber Captured10.012.014.016.018.020.022.024.026.028.030.0 H. garnotii Subadults H. mabouia Subadults Mean Temperature 2 002 20 0 3 2004 2005 156 Figure 4-10. Number of subadult H. garnotii and H. mabouia captured and the mean temperature throughout the study.

PAGE 157

157 051015202530354045AprMa y JunJulAu g Se p OctNo v DecJanFebMarAprMa y JunJulAu g Se p OctNo v DecJanFebMarAprMa y JunJulAu g Se p OctNo v DecJanFebMarAprMa y JunJulMonth/ yearNumber Captured10.012.014.016.018.020.022.024.026.028.030.0 H. garnotii Adults H. mabouia Adults Mean Temperature 2005 Figure 4-11. Number of adult H. garnotii and H. mabouia captured and the mean temperature throughout the study. 2004 20 0 3 2 002

PAGE 158

00.050.10.150.20.25JanFebMarAprMayJunJulAugSepOctNovDecMonthMean daily growth rate606570758085Mean Temperature (C) H. garnotii -all H. mabouia-all Temp Figure 4-12. Mean daily growth rate (mm) and temperature ( o C) for each month for all size classes combined.. 0.000.050.100.150.200.250.30JanFebMarAprMayJunJulAugSepOctNovDecMonthMean daily growth rate15.020.025.030.0Mean Temperature (C) H. garnotii -non-adult H. mabouia-non-adult Mean Temperature Figure 4-13. Mean non-adult daily growth rate (mm) and temperature ( o C) for each month. 0.000.050.10JanFebMarAprMayJunJulAugSepOctNovDecMonthMean daily growth rate15.020.025.030.0Mean Temperature (C) H. garnotii -adult H. mabouia-adult Mean Temperature Figure 4-14. Mean adult daily growth rate (mm) and temperature ( o C) for each month. 158

PAGE 159

Figure 4-16. Mean number of gravid H. garnotii and H. mabouia captured each month. 159 7814914521397251240264264118137050100150200250300AprMayJunJulAugSepOctNovMonthNumber of days to maturity H. garnotii H. mabouia Figure 4-15. Mean H. garnotii and H. mabouia maturation times, in days, for juveniles hatched 0.02.04.06.08.010.012.014.016.018.020.0 Month H. garnotii7.09.518.511.09.7 8.64.56.57.77.510.02.0 H. mabouia0.51.01.55.36.3JanFebMarAprMay Mean number of gravid geckos captured in each month. 8.07.05.04.03.01.00.0JunJulAugSepOctNovDec 8.5 3.6 Mean

PAGE 160

Total 701 256 160 Figure 4-17. Number of gravid H. garnotii compared to the total number of adult females captured for each snout-vent length for surveys 19 and after.

PAGE 161

.000.050.100.150.200.250.300.350.400.450.500Snout-vent length (mm)Proportio n Proportion gravid of each SVL.033.111.244.333.406.385.468.490.260.462.449.473.333.500.000 Proportion gravid of total gravid.004.020.043.055.102.078.113.094.078.164.121.102.020.0080 Proportion gravid of total population.001.007.016.020.037.029.041.034.029.060.044.037.007.003.000 Proportion of each SVL of total population.043.064.064.060.091.074.088.070.110.130.098.078.021.006.001495051525354555657585960616263 161 Figure 4-18. Proportion of adult H. garnotii that were gravid for each snout-vent length (SVL)(mm) compared to the total number captured at each SVL (gravid at each SVL / (gravid + nongravid at each SVL)), the proportion that were gravid at each SVL as compared to the total number of gravid at all SVLs (gravid at each SVL / gravid at all SVLs), the proportion gravid for each SVL as compared to the total number of adults captured (gravid at each SVL / (gravid + nongravid adults at all SVLs), and the proportion of each SVL as compared to the total number of adults captured ((gravid + nongravid adults at each SVL / (gravid + nongravid adults at all SVLs).

PAGE 162

T ot a l 162 228 123 Figure 4-19. Number of gravid H. mabouia compared to the total number of adult females captured for each snout-vent length for surveys 3 and after.

PAGE 163

.000.100.200.300.400.500.600.700.800.9001.000Snout-vent length (mm)Proportion Proportion gravid of each SVL.0830.333.400.586.455.737.750.579.444.625.909.833.4001.500.66700 Proportion gravid of total gravid.0080.033.033.138.081.114.146.089.098.081.081.041.016.016.008.01600 Proportion gravid of total population.0040.018.018.075.044.061.079.048.053.044.044.022.009.009.004.00900 Proportion of each SVL of total population.053.035.053.044.127.096.083.105.083.118.070.048.026.022.009.009.0130.00451525354555657585960616263646566676869 163 Figure 4-20. Proportion of adult H. mabouia that were gravid for each snout-vent length (SVL)(mm) compared to the total number captured at each SVL (gravid at each SVL / (gravid + nongravid at each SVL)), the proportion that were gravid at each SVL as compared to the total number of gravid at all SVLs (gravid at each SVL / gravid at all SVLs), the proportion gravid for each SVL as compared to the total number of adults captured (gravid at each SVL / (gravid + nongravid adults at all SVLs), and the proportion of each SVL as compared to the total number of adults captured ((gravid + nongravid adults at each SVL / (gravid + nongravid adults at all SVL

PAGE 164

CHAPTER 5 ADDITIONAL DISTRIBUTIONAL OBSERVATIONS In the course of surveying Charlotte and Lee Counties for Hemidactylus garnotii and H. mabouia (Chapter 3) as well as during other opportunistic surveys from 2001 to 2006, I encountered other non-indigenous gecko species on some buildings. Supplementary surveys of southern Sarasota County also yielded additional distributional records. Following, I present these observations. Hemidactylus turcicus Hemidactylus turcicus, native to the Mediteranean region (Loveridge 1947), was first recorded from Florida in Key West in 1910 (Fowler, 1915). With a maximum reported size of 59 mm SVL (Schwartz and Henderson 1991) it is somewhat smaller than H. garnotii. In Fort Myers, Lee County I encountered two H. turcicus on one building (Table 5-1 and Figure 5-1). As this was a restaurant, Red Lobster, these may have been transported on a delivery truck. Though this was a new county record, I do not consider this species to be established in the county since out of hundreds of buildings surveyed this was the only observation. Hemidactylus frenatus Hemidactylus frenatus is native to southern and southeast Asia and Indonesia (Bauer and Henle 1994). It was first reported in Florida in 1993 in Monroe County (Meshaka et al. 1994c) and is now known from Broward, Lee, and Miami-Dade Counties (Krysko and Daniels 2005). Its maximum size is 62 mm SVL (Klowden unpublished). No data have been published on its distribution in Lee County. I encountered H. frenatus on 26 buildings in Fort Myers, Lee County (Table 5-1 and Figure 5-1). Though I observed this species at a low abundance in several locations, they were concentrated in three main areas. The highest concentration was on buildings near the intersection of McGregor Boulevard and College Parkway. On these 164

PAGE 165

buildings from 5 May 2005 to 12 May 2005 I collected 157 individuals of this species and observed many more. The second area of concentration was approximately in the 3700 to 3800 block of Cleveland Avenue (US-41) and numerous buildings west of this road. A third area in which I observed numerous, but comparitively fewer, H. frenatus was in downtown Fort Myers near USB-41, SR-82, and CR-80. Though this area primarily contained H. mabouia and H. garnotii, I observed H. frenatus on several buildings. Gekko gecko The tokay gecko, Gekko gecko, is native to southern and southeast Asia and Indonesia and north to southern China (Brown and Alcala 1978, Bauer and Henle 1994). In Florida, it has been reported from five counties (Krysko and Daniels 2005). At approximately 185 mm SVL this large species is popular in the pet trade and is often released inside houses in the belief that it will control insects. I observed G. gecko on four scattered buildings in Lee County and one in Charlotte County (Table 5-1 and Figure 5-1). At 1031 NE Pine Island Rd (CR-78) I observed 3 large adults and several very small juveniles. This building is a pet store and there were no adjacent buildings. I observed one large adult on a veterinary office at 9580 Cypress Lake Drive (CR-876). I also observed a medium sized adult on 12720 McGregor Boulevard (SR-867) and I captured two male adults on 12700 McGregor that measured 178 and 145 mm snout-vent length, respectively. In Charlotte County, I observed one large adult on 24123 Peachland Blvd. Port Charlotte. This building was a large stripmall that contained a pet store. It appears that this species generally remains at or near its point of introduction. Tarentola anularis Tarentola annularis is native to the desert and savanah regions of northern Africa and has a maximum SVL of about 100 mm (Schleich et al. 1996). In Florida, it has been reported from 165

PAGE 166

Lee and Miami-Dade Counties (Krysko and Daniels 2005). In Lee County, on 11 August 2004, I captured 5 and observed 6 additional T. annularis (Table 5-1 and Figure 5-1) of all size classes on 954 NE Pine Island Rd (CR-78), Cape Coral. In October 2004 I observed 10 individuals on this same building. This building was on the opposite side of the road from the pet store on which I observed G. gecko. In May 2005, I captured and observed approximately 20 T. annularis of all size classes on 12710, 12720, and 12730 McGregor Boulevard (SR-867). On two of these buildings I also observed H. frenatus and one G. gecko, and it was next to another building that contained G. gecko. It is likely that a pet store was previously in one of these buildings. Hemidactylus mabouia To ascertain if H. mabouia had dispersed into Sarasota County, I surveyed 130 buildings in southern Sarasota County (Table C-1 and Figure 5-2). On 15 May 2005, I surveyed 44 buildings on south US-41 near the Charlotte County border. Twenty-eight were unoccupied, H. garnotii was present on 14, H. mabouia was present on one, and both species were present on one additional building. On 21 November 2005, I surveyed 37 buildings on US-41 in South Venice. Twenty were unoccupied, H. garnotii was present on 15, and H. mabouia was present on two. On 2 November 2005 I surveyed 49 buildings on SR-776 in Englewood. Fourteen were empty, H. garnotii was present on 29, H. mabouia was present on three, and both species were present on three additional buildings. The farthest north H. mabouia was observed was on US-41 at 1868 South Tamiami Trail, 27.06346 o W latitude, -82.41791 o N longitude. 166

PAGE 167

Table 5-1. Gekko gecko, Hemidactylus frenatus, H. turcicus, and Tarentola annularis observed or captured from 2001 to 2006. Species a Latitude Longitude Street # Street Name SR/ CR b Location details City G. gecko 26.66210 -81.95407 1031 NE Pine Island Rd CR-78 Pet store / River of Life Worship Center Cape Coral G. gecko 26.54592 -81.91314 9580 Cypress Lake Dr CR-876 Ft Myers G. gecko 26.55825 -81.91008 12700 McGregor Blvd SR-867 Ft Myers G. gecko 26.55769 -81.91033 12720 McGregor Blvd SR-867 Ft Myers G. gecko 27.01529 -82.05354 24123 Peachland Blvd (Kings Hwy) CR-769 Address on Peachland but faces CR-769 Port Charlotte H. frenatus 26.64325 -81.86355 1773 Fowler Ave USB-41 Banyan Tree Gifts Ft Myers H. frenatus 26.60902 -81.87206 3706 Cleveland Ave US-41 Pizza Hut Ft Myers H. frenatus 26.66512 -81.83230 4008 Palm Beach Blvd/ Van Buren CR-80 Jack MacDonald Auto/ Coqui Restaurant Ft Myers H. frenatus 26.54556 -81.91197 6610 Warwick At corner of Warwick and CR-876 Ft Myers H. frenatus 26.55697 -81.90869 9280 College Pkwy Near SR-867 Ft Myers H. frenatus 26.55706 -81.90919 9320 College Pkwy Near SR-867 Ft Myers H. frenatus 26.57150 -81.87147 11601 S. Cleveland US-41 Vet/Middle East Deli/Dan's Fans Ft Myers H. frenatus 26.57016 -81.87168 11705 S Cleveland Ave US-41 SCUBA/ Ocean Enironments Ft Myers H. frenatus 26.56011 -81.90942 12600 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55950 -81.90969 12640 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55881 -81.90825 12651 (Bldgs 1-5) McGregor Blvd SR-867 Not sure which bldg but put Lat/long as bldg 5 Ft Myers H. frenatus 26.55917 -81.90978 12670 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55886 -81.90992 12680 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55856 -81.91006 12690 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55842 -81.90964 12691 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55839 -81.90900 12693 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55825 -81.91008 12700 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55764 -81.90897 12713 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55764 -81.90897 12715 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55769 -81.91033 12720 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55764 -81.90897 12721 McGregor Blvd SR-867 Ft Myers H. frenatus 26.55753 -81.91028 12730 McGregor Blvd SR-867 Ft Myers H. frenatus 26.57056 -81.87158 116xx or 117xx S. Cleveland US-41 Not certain of address Ft Myers 167

PAGE 168

Table 5-1 Continued. Species Latitude Longitude Street # Street Name SR/ CR a Location details City H. frenatus 26.63648 -81.86212 2203-73 Fowler Ave USB-41 Stripmall Ft Myers H. frenatus 26.64073 -81.85653 2769-73 Dr MLK Blvd/ Palm Ave SR-82 Blanding Bailbonds/ True Holiness Church Ft Myers H. turcicus 26.60700 -81.87158 3801 Cleveland Ave US-41 Red Lobster Ft Myers T. annularis 26.66042 -81.95554 954 NE Pine Island Rd CR-78 Home Dcor/ Keller Pools/ El Arepazo Cape Coral T. annularis 26.55753 -81.91042 12710 McGregor Blvd SR-867 Ft Myers T. annularis 26.55765 -81.91031 12720 McGregor Blvd SR-867 McGregor Antique Mall Ft Myers T. annularis 26.55753 -81.91028 12730 McGregor Blvd SR-867 Ft Myers (a) State Road/ County Road 168

PAGE 169

Legend 169 Figure 5-1. Other geckos observed in Lee County

PAGE 170

170 Figure 5-2. Geckos observed in southern Sarasota County. Legend Sarasota Count y Charlotte Count y

PAGE 171

CHAPTER 6 SUMMARY My research was directed toward discerning if a causal link existed between changes in Hemidactylus garnotii distribution and abundance and the co-occurrence of H. mabouia and if so, what role differences in life history might play. I presented a significant amount of data on their distribution, co-occurrence, colonization and extinction rates, demographics, and various other life history traits. In my literature review and survey of museum specimens (Chapter 2), I discovered a paucity of distributional and life history data and the use of very limited surveys to conclude replacement of one species by another (Meshaka 1995, 2000, Meshaka et al. 1994b, 2005). Despite these assertions and those echoed by other authors (e.g. Powell et al. 1998, Punzo 2001, 2005) quantitative analyses of this pattern have been minimal and based on very small sample sizes. For example, one study (Meshaka et al. 2005) inferred replacement by comparing the relative proportion of geckos on 10 buildings from 17 surveys over a 70 day period. However, data collection occurred during the cooler winter and early spring months when activity levels are typically reduced (Table 4-3 and 4-4), sample size was very small, and no prior surveys occurred, making conclusions of the dynamic replacement process from this static pattern unfounded. Observed patterns may be due to expected fluctuations or random extinction of small populations. Two studies have provided limited evidence of localized abundance shifts over time. Enge et al. (2004) noted that in five urban Miami parks, from 1997 to 2002, H. mabouia colonized several parks and became abundant while H. garnotii was extirpated from two parks and its abundance decreased in the remaining three parks to only a few individuals. Meshaka (2000) discussed occupancy and abundance shifts in Everglades National Park on 12 buildings occupied by H. garnotii and/ or H. mabouia. This studys geographic scope was limited to just a 171

PAGE 172

few buildings and did not provide sufficient quantitative data to support his conclusion that invasion by H. mabouia caused the reduction of H. garnotii populations. The relative dearth of data on a reportedly extensive phenomena, illustrated a need for systematic widespread surveys to evaluate how H. garnotii and H. mabouia distributions were related. I conducted such surveys in Charlotte County from 2003 to 2006 and in Lee County in 2004 and 2006 (Chapter 3). During my survey period I found that, in both counties, the proportion of buildings on which H. garnotii occurred decreased while for H. mabouia increased. This increase was especially significant in Charlotte County where H. mabouia was limited to a few places in the county in 2003 but by 2006 was found throughout the county. Further corroborating this pattern, I also showed for both counties that for H. garnotii the extinction rate from surveyed buildings was high while the colonization rate was low. In contrast, H. mabouia had a low extinction and high colonization rate. Moreover, in Lee County in 2004, H. garnotii and H. mabouia appeared to be distributed randomly as they co-occurred as often as would be expected by chance. However, by 2006 they were no longer randomly distributed and had a decreased likelihood of co-occurring than would be predicted by chance. Whereas it appears that colonization of a building by H. mabouia is not linked with the immediate extinction of H. garnotii, longer term associations may have different consequences. Though colonization by H. mabouia did not appear to be the primary cause of extinction of H. garnotii from a building, on buildings on which they co-occured, the extinction rate of H. garnotii was significantly higher than for H. mabouia. However, these species seem to be able to co-exist for at least some period of time as indicated by their continued co-occurrence in 2006 on 41% of the buildings on which they co-occurred in 2004. Longer term sampling is required to determine if they can persist together and under what conditions. 172

PAGE 173

The comparatively low colonization rate of H. garnotii appears to be due in part to the presence of H. mabouia. The ability of H. garnotii to colonize buildings that were occupied by H. mabouia the previous year was significantly lower than for buildings that were unoccupied. In contrast, there was not a significant difference in the ability of H. mabouia to colonize buildings previously occupied by H. garnotii versus unoccupied buildings. Further, it appears that once H. garnotii becomes extinct from a building and it is colonized by H. mabouia, there is a decreased probability for re-colonization by H. garnotii. Though this pattern is intriguing, by itself, it is not enough to conclude causation (Case 1983) and the mechanism underlying this pattern warrants further investigation. In Chapter 4, I investigated life history traits of H. garnotii and H. mabouia. I conducted a mark-recapture study on a population of H. garnotii in Charlotte County from August 2001 to August 2005. Simultaneously I conducted a similar study on a population of H. mabouia in Lee County from March 2003 to August 2005. I estimated stage specific survival, fecundity, mean snout-vent length, weight, maturation size, growth rate, time to maturation, inter-clutch interval, maximum longevity, and other parameters. I showed that both species mature rapidly and are similar in having high growth rates. Mean fecundity for H. garnotii more than double that of H. mabouia in part because it is an all female parthenogenetic species and all adult H. garnotii are capable of reproduction. As H. garnotii had a fairly high extinction rate, a high fecundity rate may be advantageous when colonizing or re-colonizing buildings. Adult survival in both species was similar, however, subadult survival was somewhat higher for H. mabouia and juvenile survival was double that of H. garnotii. Low H. garnotii juvenile survival would be disadvantageous when colonizing new buildings since a low population growth rate may increase the likelihood of random extinction. If Cuban treefrogs, Osteopilus septentrionalis (Meshaka 173

PAGE 174

2001) or, as I documented, H. mabouia, prey upon juvenile H. garnotii, even if only occasionally, this could further reduce the colonization ability of H. garnotii. Additionally, I documented that the average H. mabouia is nearly 13% longer, more than 90% heavier, and has a markedly wider head than the average H. garnotii. This could be important by also allowing H. mabouia to prey upon larger subadult H. garnotii and could prevent H. garnotii from obtaining the most secure hiding spots upon colonization of a building. Such differences could account for the reduced ability of H. garnotii to colonize buildings that contain H. mabouia as opposed to those which are unoccupied. In Chapter 5, I presented data on additional non-indigenous gecko species I encountered in Charlotte and Lee Counties. Though I did not find Gekko gecko and Tarantola annularis on more than a few buildings, H. frenatus occurred in several areas of Fort Myers, Lee County. Though this species has not spread as rapidly as H. mabouia, it warrants close attention as it seems to be expanding its range (Meshaka et al. 1994c, Krysko et al. 2005), can occur at extremely high densities (Case et al. 1994, Klowden see Chapter 5), and has been an extremely prolific invader in many places around the world (Bauer and Henle 1994). I also presented data on the continued northward dispersal of H. mabouia into Sarasota County. It is my hope that other researchers will continue to survey for these species and not simply rely on opportunistic collecting to document their spread and persistence. The techniques I have established should act as a guide for future investigations. 174

PAGE 175

APPENDIX A DETAILED METHODOLOGY FOR GENERATION OF RANDOM POINTS I determined random survey points using ArcView GIS v.3.1, Florida Geographic Digital Library (FGDL) v. 3.0 2000 road layers, ArcView GIS Random Point Generator v. 1.1 extension by Jeff Jenness, Garmin etrex GPS, Waypoint + v. 1.8.03 by B. Hildebrand, and Microsoft Excel 2000. Following is the detailed methodology necessitated by the quirks and limitations of the computer programs available. 1. Copy the desired FGDL files from the appropriate county compact disc to the hard drive. The roads theme is called majrds_XX where XX is a number that differs for each county. Also transfer other desired themes such as county boundaries. 2. Start new project and add the theme majrds_XX. 3. Make the majrds_XX theme active and under the View menu select Properties. 4. In the properties dialog box change the Map Units to meters and select OK. 5. Reproject the majrds_XX theme to lat/long: A. Start the projector module B. In the Projection Properties dialog box: 1. Select Custom and enter the following: a. Projection: Albers Equal-Area Conic b. Spheroid: GRS 80 c. Central Meridian: -84 d. Reference Latitude: 24 e. Standard Parallel 1: 24 f. Standard Parallel 2: 31.5 g. False Easting: 400000 h. False Northing: 0 i. select o.k. j. output units: decimal degree k. Recalculate area, perimeter, and length fields using decimal degrees: Yes l. Add projected shapefile as theme to view: Yes m. Retain theme names and legend files for the new view: Yes n. Add theme to: Either yes or no o. Save the file under the desired name and location. p. Make the new theme active and select zoom to active theme. 6. Modify the new theme so it only has state & county roads: A. View the data table B. Select the state and county roads C. Switch the selection D. View the map E. Start editing F. Delete the selected G. Save edits as new name H. Stop editing 175

PAGE 176

7. Make a copy of the modified roads theme and merge all of the roads so that they appear as a single road. This works around a problem with the point generator extension allowing you to obtain desired # of sample points on all roads instead of a particular # per road. A. View map of state & county roads B. Start editing (Theme menu) C. Select all (use select feature tool or select all in table) D. Edit menu Union features E. Save edits as name(merged) F. Stop editing 8. Obtain Random Points using random point generator A. Open random point generator (May need to activate the extension under File/Extension) B. Dialog box 1: 1. Select With Respect To Lines In A Theme 2. Select ok C. Dialog box 2: 1. Under select Line Theme, select the merged roads theme 2. Under Select Line Theme ID Field choose any option as this is not important. 3. Under Number of Random Points Per Line enter desired number of points (I used 2X the number of points needed) 4. Under Locate Random Points in This Area choose Random Points Located On line 5. Under Output Data For Random Points select X coordinate and Y coordinate 6. Under Output Data Format select New Shapefile D. Select OK E. Save file with desired name and location 9. Join random points and unmerged road attribute tables to i.d. the points on a specific road. G. Open road attribute table and select shape column H. Open random point attribute table and select shape column I. From the Table menu select Join 10. Export random point table to excel. A. While viewing the newly joined random points table, under the File menu select Export B. Choose Delimited Text and save the file. C. Open the saved comma delimited file in Excel. D. Erase any undesired columns. Retain the Unique ID, Rnd_X_Crd, Rnd_Y_Crd, and columns with road identification data. i. Unique ID: Will be used in random point determination later ii. Rnd_X_Crd: Longitude iii. Rnd_Y_Crd: Latitude 11. Reorder the points by road name. 12. Insert a row between different roads to separate into groups. 13. Randomly (coin flip) determine the direction the points will be searched for along each road (e.g heads you will search north to south and tails south to north). 14. Sort each group to make it easier to locate points. A. To sort predominately north-south roads in a north to south order, sort in Z-A order by Latitude. To sort in a south to north order sort A-Z by latitude. 176

PAGE 177

B. To sort predominately east-west roads in an east to west order, sort in a Z-A order by longitude. To sort in a west to east order sort A-Z by longitude. C. If a road has a portion that is distinctly NS and another portion that is EW, separate these portions into 2 groups for easier sorting. 15. Uniquely number the points in the sorted order to make point location with a GPS easier. 16. Too retain the order of points when uploaded into a GPS, all points must be in a 3-digit format (e.g. 001, ., 020,., 100). A. Can be done quickly by using the concatenate function. 1. add column before and after unique id. 2. highlight all 3 columns and change format to text (Format/cells/text) 3. in column before put for single digit unique ids and for double digit unique ids. 4. place cursor in first row (below headings) of column after unique id. 5. type =concatenate(a2,b2) (w/o the quotes) where a2 and b2 refer to the preceding 2 columns. 6. copy this formula to all cells in this column that correspond to single and double digit unique ids. 7. copy the triple digit ids from the unique ids to this new column. 8. Highlight and copy this entire column. 9. Select edit/paste special/unique values. 10. Erase columns 1 and 2. 11. Add Search ID title to column 3. B. If you plan to have multiple county random point data sets on your GPS at one time, add an identifier to the unique point IDs (e.g. LEE for Lee county or SAR for Sarasota county): 1. Add a column before and after the new Search ID column. 2. In column 1 type the unique identifier (e.g. LEE for Lee County) to all rows. 3. In the first row of column 3 (below headings) type =concatenate(a2,b2) (w/o the quotes) 4. copy this formula to all rows in column 3. 5. highlight and copy column 3. 6. Select Edit/Paste special/unique values 7. Erase columns 1 and 2 8. Retitle column 3 to Search ID 17. If your GPS will not handle decimal degrees format, in Excel, convert lat/long from DD.dddd to DD/MM.mmm if needed (Not needed for use in Garmin Etrex GPS but needed for use with Magellan Tracker): A. add six new blank columns: three after the lat column and three after the long column B. Select the entire X coordinate column C. From the data menu select text to columns D. Choose fixed width and select next E. Place a dividing line so that the whole number is to the left and the decimal and following digits are to the right then choose next F. Verify that that the line is accurately placed. G. Under Destination select the two blank columns to the right of the data. 177

PAGE 178

H. In the remaining blank column next to the newly created columns write a formula that multiplies the decimal column by 60 (e.g. =c3*60). This will result in decimal minutes. I. Select the new decimal minutes column, select copy, and then paste special and select values. This results in a column with numbers only and no formula. The decimal column can now be deleted. J. Repeat for longitude 18. Transfer Random Point Lat/long to GPS. A. For Garmin Etrex use Waypoint + 1. In Excel file, put latitudes and longitudes for each road in the order they will be visited. i. Sort points by road. ii. Separate different roads with a blank row. iii. Sort each road. 1. N/S roads sort from latitude column a. N to S sort Z to A b. S to N sort A to Z 2. E/W roads sort from longitude column. iv. Remove blank rows between roads. 2. Save sorted latitudes and longitudes in new Excel file 3. To import from computer to Etrex: i. Save random point file as new file. ii. Delete blank rows. iii. Delete all columns except search ID, latitude, and longitude (and road description if your GPS has this capability). iv. data must be in the following format: Waypoint,data format, waypoint name, latitude, longitude,,, e.g. waypoint,D,CC001,26.12345,-82.12345,,, 1. Add 3 columns before lat and long columns 2. In column A put waypoint (no quotes) 3. In column B put D (no quotes) 4. In column C put waypoint name. If using numbers make sure to use 3 numbers (e.g. SAR001 not SAR1 or 001 not 1). Since the lats and longs were previously sorted, naming waypoints in numeric order will make it easier to find locations. 5. Latitude in column D 6. Longitude in column E 7. put a in columns F, G, and H 8. Delete the header row. 9. Save as a comma delimited csv file and close Excel. 10. In Windows Explorer change file extension from csv to txt. 11. Open txt file in notepad. 12. From edit menu select replace and put in the find and leave replace box empty. 4. In Waypoint+, under the File menu open the txt file. 5. Under the Waypoint menu verify that the waypoints were opened properly by selecting list waypoints 178

PAGE 179

6. Under the Waypoint menu select upload. 19. Save a copy with points reordered according to road and lat/long for easier use in GPS (To sort roads easier I added a direction column (NS or EW) so I know whether to sort them by latitude or longitude). 179

PAGE 180

APPENDIX B SURVEY BUILDING STATISTICS AND RAW DATA Table B-1. Charlotte County survey building locations and number of Hemidactylus garnotii and H. mabouia seen. 2003 d 2004 d 2006 d Latitude Longitude Street# Street name Road # a City b Business c g m u g m u g m u 26.94767 -81.99898 27690 Bermont Rd CR74/ US17 PG Winn Dixie 1 0 0 0 0 0 0 1 0 26.94785 -81.98793 28330 Bermont Rd CR74 PG New Life Family Worship 00 0 30 0 22 000 031 100 206 1 0 0 0 0 0 0 26.94723 -81.98245 5881 Shalimar (CR74) PG Spirit and LifeChristian Ministries 0 0 0 0 0 0 26.89087 -82.02432 10001 Burnt Store Rd CR765/ US41 PC Wachovia Bank 0 0 0 1 0 0 0 0 0 26.89025 -82.02403 10021 Burnt Store Rd CR765 PC Juicy Lucy's 0 0 0 0 0 0 0 2 0 26.88985 -82.02278 10001 Burnt Store Rd CR765/ US41 PC Citgo 0 0 0 0 0 0 0 0 0 26.88855 -82.02310 10050 Burnt Store Rd CR765 PC Burnt Store Travel 2 0 0 1 0 0 0 0 0 26.88477 -82.02352 10101 Burnt Store Rd CR765 PC Parkhill RV Park office 1 3 2 2 10 2 26.88257 -82.02357 10205 Burnt Store Rd CR765 PC Encore RV Park office 0 0 0 0 1 0 26.86403 -82.02375 11330 Burnt Store Rd CR765 PC Burnt Store Presbyterian Church 2 1 5 1 12 6 26.86083 -82.02377 26001 Notre Dame (CR765) PC Unknown building 0 0 0 0 0 0 2 0 0 26.84993 -82.02183 12301 Burnt Store Rd CR765 PC Charlotte Harbor Buffer Preserve 0 0 0 1 4 1 26.80028 -82.03695 15500 Burnt Store Rd CR765 PC Star Center 3 0 2 4 0 0 4 0 0 26.78698 -82.03750 25000 Burnt Store Rd CR765 PC Coldwell Banker 1 1 0 1 4 0 1 2 0 26.78655 -82.03773 25001 Burnt Store Rd CR765 PC Burnt Store Properties 1 0 9 3 10 0 180

PAGE 181

181 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m0 u 1 1 g2 m0 u g m u 26.91793 -82.03373 2300 Taylor Rd CR765A PG Church of Christ (Kingdom Hall of Jehovah's Witnesses) 0 0 0 0 26.91670 -82.03215 25349 Airport Rd (CR765A) PG Punta Gorda Lodge 20 100 0 40 1 39 ----10 040 030 000 0 10 0 0 0 0 0 0 0 26.91292 -82.02920 2851 Taylor Rd CR765A PG Red Barn Auction Warehouse (Used Furniture Outlet) 0 0 0 0 0 0 26.88077 -82.00615 6305 Taylor Rd CR765A PG Sprint Substation 1 0 0 6 0 0 5 0 0 26.87922 -82.00655 26530 Mallard (CR765A) PG Luxury Services 0 0 0 0 0 0 2 0 0 26.87847 -82.00630 6601 Taylor Rd CR765A PG KPM FAA Repair Center 0 1 0 1 11 4 26.89348 -81.99802 26901(aprox) Jones Loop Rd CR768/ I-75 PG I-75 Rest stop mile 160 1 26.89580 -81.98673 27200 Jones Loop Rd CR768 PG Royal Palm Memorial Gardens 0 4 0 0 0 0 26.89492 -81.98457 27265 Jones Loop Rd CR768 PG Kingdom Hall of Jehovah's Witnesses 1 2 0 0 0 0 26.88485 -81.99488 29201 Jones Loop Rd CR768 PG Blue Heron Pines Proshop 0 0 0 0 0 0 26.98967 -82.05837 23456 Olean (CR769) PC Christ the King Lutheran Church 0 1 0 0 2 0 26.99358 -82.05837 23970 Suncoast (CR769) PC CharlotteSurgery Center 0 1 0 0 0 0

PAGE 182

182 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m u0 g m0 u 7 10 g m u 26.99645 -82.05762 2200Bldg 3 Kings Hwy CR769 PC Dental/ Cap'n and the Cowboy 0 3 0 0 26.99707 0 003----0 103 100 04 00 -82.05748 2200Bldg 2 Kings Hwy CR769 PC Barber/ Allstate 2 0 0 2 0 0 1 0 0 27.00400 -82.05520 2000 Kings Hwy CR769 PC King's Crossing 2 0 0 1 0 0 0 3 5 27.01410 -82.05355 1021 Kings Hwy CR769 PC Hess 1 0 0 1 0 0 1 2 0 27.01480 -82.05398 24150 Tiseo (CR769) PC Barber/ pizza 1 0 0 0 0 0 0 1 0 27.02402 -82.04448 330 Kings Hwy CR769 PC 7-Eleven 1 3 1 0 0 0 0 4 0 27.02310 -82.04393 24410 Sandhill Blvd (CR769) PC Carwash 0 0 0 0 0 0 0 0 1 27.02232 -82.04470 24451 Sandhill Blvd (CR769) PC Medical center 0 0 0 0 0 0 0 1 0 26.93378 -82.22163 1595 Mccall Rd SR776 PC Citgo 1 0 1 0 0 0 1 0 0 26.93302 -82.22357 13435 Mccall Rd SR776 PC Gulf Cove Publix Shopping Center 0 0 3 0 0 0 26.92523 -82.36057 2075 Beach Rd CR776 EW Utilities on Manasota Key 0 0 26.90943 -82.22830 8252 Wiltshire Dr (CR771) PC Duffy's Realty Station 2 0 2 0 0 0 26.90918 -82.22843 8264 Wiltshire Dr (CR771) PC South Gulf Cove Realty 0 0 1 0 0 0 26.90890 -82.22868 8282 Wiltshire Dr (CR771) PC Hoot Gibson Construction 0 0 0 0 0 0 0 26.83632 -82.26180 15001 Gasparilla Rd CR771 PL Ingman Marine 4 0 0 3 0 0 1 2 1 26.91007 -82.32515 3031 Placida Rd CR775 EW Willow Plaza 0 0 0 0 0 0 2 0 0 26.90993 -82.32555 3030 Placida Rd CR775 EW Bizzy Buzzy's Coin Laundry 0 0 6 0 0 0 26.90913 -82.32517 3060 Placida Rd CR775 EW Dockside Marine 6 0 0 26.85913 -82.29927 8501 Placida Rd CR775 EW Cape Haze Plaza 12 0 0 11 0 0 5 0 1 26.85372 -82.29115 8675 Placida Rd CR775 EW Mobil 0 0 0 1 0 0 0 0 0

PAGE 183

183 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m0 u 1 00 g m2 u g m u 26.85295 -82.29013 8725 Placida Rd CR775 EW Publix at Cape Haze 1 11 1 0 26.93970 -82.34977 1261 Beach Rd CR776 EW Surfside Styling Salon 10 0001 -23 ---01 ---01 --03 --12 12 10 10 10 10 0 0 0 1 3 26.93953 -82.35000 1271 Beach Rd CR776 EW Surfside Realty 1 0 0 0 0 0 0 1 1 26.93900 -82.35012 1350 Beach Rd CR776 EW The Red Pelican 0 0 0 0 0 0 02 026.92913 -82.36280 2395 Beach Rd CR776 EW Surfside Realty (Englewood Realty) 1 0 0 0 26.92627 -82.36098 2095 Beach Rd CR776 EW Gulf View Grille 1 0 1 2 0 0 26.92422 -82.35903 1950-54 Beach Rd CR776 EW Calico Jack's/ Geraldi's Pizza 0 0 2 0 26.92407 -82.35822 1927 Beach Rd CR776 EW Manasota Key Realty 4 0 0 0 26.92400 -82.35868 1939 Beach Rd CR776 EW Island Court Seafood 0 0 0 0 26.96540 -82.06535 23264 Harborview Rd CR776 PC GingerbreadTrim 0 0 1 0 26.96548 -82.06462 23296 Harborview Rd CR776 PC Circle K 0 1 0 0 3 0 0 1 0 26.96573 -82.06383 23330 Harborview Rd CR776 PC Self Storage 0 0 0 3 0 0 1 0 0 26.96600 -82.06112 23415 Janice (CR776) PC Clune's Auto Body 2 0 0 0 8 26.96632 -82.05647 24038 Harborview Rd CR776 PC HarborviewChristian Church 0 3 0 0 1 26.97180 -82.03532 25180 Harborview Rd CR776 PC Peace RiverWesleyan Church 1 4 0 0 0 27.01207 -82.14668 1331 Tamiami Tr SR776 PC Red Lobster 3 0 1 0 0 1 0 12 3

PAGE 184

184 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m2 u 0 00 g2 m5 u g m u 27.01217 -82.14757 1020 El Jobean Rd SR776 PC Auto Express (Embarq) 0 1 3 27.01172 -82.14780 1050 El Jobean Rd SR776 PC Bob Evans 1 0 0 0 0 0 1 1 0 27.01293 13 00 03 00 02 00 00 00 10 00 10 1007 1300 1000 00 -82.14917 1225 Tamiami Tr SR776/US41 PC Charlotte TradeCenterEmerald City 0 0 2 0 2 27.01238 -82.15010 1242 Market (SR776) PC Embroidery byDesign/ American Cabinets 1 0 1 0 2 27.01240 -82.15027 1232 Market (SR776) PC SignatureHomes/ Ace Door 0 0 2 0 2 27.01122 -82.15237 1312 Market (SR776) PC Model HomeFurniture 0 0 0 0 1 27.01033 -82.15270 17701 Murdock Cir (SR776) PC Bacon'sFurniture 1 1 0 0 3 27.01105 -82.15297 1300 Enterprise (SR776) PC Pinch-a-PennyPools/ Red Cross 1 3 0 0 0 27.00527 -82.16748 1720 El Jobean Rd SR776 PC Oak Park Professional Center 0 0 5 0 6 27.00482 -82.16903 1770 El Jobean Rd SR776 PC Country Peddler 1 0 0 0 0 0 0 0 0 27.00455 -82.16972 1800 El Jobean Rd SR776 PC Grants Pools 2 0 0 2 0 1 2 0 0 27.00195 -82.17760 2128 El Jobean Rd SR776 PC Continental Tile and Marble 0 1 1 0 1 26.93380 -82.22408 5665 S McCall Rd SR776 PC Banditos 4 0 0 2 0 0 0 0 0 26.93300 -82.22528 13423 S McCall Rd SR776 PC Mobil (Amoco) (Bp) 0 0 0 0 0 26.93237 -82.27607 3993 S. Access Rd SR776 EW Robert's Homes 1 0 1 2 0 0 0 0 0

PAGE 185

185 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m u g m u g m u 26.93242 -82.27697 3973 S. Access Rd SR776 EW Value Carpet 1 0 1 1 0 0 0 0 0 26.93242 -82.27718 3969 S. Access Rd SR776 EW Sybil's 1 0 0 1 00 0 0 0 26.93252 -82.28558 3691 S. Access Rd SR776 EW Lots-a-Tot Day Care 00 0000 -0000 0020 1020 00 00 0010 --00 30 0 0 26.93268 -82.28645 3651 S. Access Rd SR776 EW Access Road Hair Salon 0 0 26.93282 -82.28720 3631 S. Access Rd SR776 EW State Farm 0 0 0 0 0 0 26.93423 -82.31997 2411 S McCall Rd SR776 EW Eason Plaza 1 0 1 2 0 1 1 3 0 26.93432 -82.32070 2391 S McCall Rd SR776 EW Florida Furnishings 0 0 0 0 0 26.93437 -82.32110 2381 S McCall Rd SR776 EW Basil Appliance 1 0 0 1 0 0 0 0 0 26.93863 -82.33967 1720 S McCall Rd SR776 EW Cornerstone Shopping Plaza 0 6 0 0 0 26.93875 -82.33997 1680 S McCall Rd SR776 EW Speedway 0 0 0 0 0 0 26.93873 -82.34043 1630 S McCall Rd SR776 EW Lemon Bay Motel 1 4 2 0 9 26.94298 -82.03342 25197 Marion Ave US17 PG StorybookAcademy 0 1 0 0 0 26.94298 -82.03293 25176 Marion Ave US17 PG Emerald Point Sales and Rentals 0 0 0 0 26.94140 -82.03092 6220 Marlympia Way (US17) PG Circle K 5 0 0 0 1 0 0 1 0 26.94318 -82.01193 6161 Duncan Rd US17 PG Baby Barrier 1 0 0 3 1 0 1 0 0 26.94372 -82.01127 6117 Duncan Rd US17 PG Shell 0 0 0 0 0 0 26.94403 -82.00982 6101 Duncan Rd US17 PG Country Club Plaza (Towle's Plaza) 0 0 0 0 0 26.95828 -81.99033 5179 Duncan Rd US17 PG Video Country 0 0 0 0 1 0 0 3 0 26.95845 -81.99010 5169 Duncan Rd US17 PG Circle K 0 2 1 0 0 0 0 2 0

PAGE 186

186 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m0 u 2 00 g1 m0 u g m u 26.95860 81.99020 28097 Cleveland Ave (US17) PG Fry Brothers Aluminum 0 1 0 26.97003 -81.96153 4214 Duncan Rd US17 PG Royal Thai 8 0 0 6 0 0 3 0 0 26.97345 -81.95958 4043 Duncan Rd US17 PG Tower Bar 3 0 0 26.98467 -81.95927 29501 Taralane Dr US17 PG Empty Bldg 1 0 1 5 0 0 0 0 0 26.98723 -81.95922 3149 Duncan Rd US17 PG Deb's 17 Lounge 1 0 0 4 0 0 1 0 0 26.99480 -81.95867 2390 Duncan Rd US17 PG Mini Mart (Farm Store) 10 0100 0050 0030 0001 00 50 1220 00 1 4 0 0 1 27.03275 -82.20503 41 Tamiami Tr US41 PC Circle K/ 76 Station 0 0 0 0 0 27.03118 -82.20342 59 Tamiami Tr US41 PC North Port Tires 1 0 0 0 0 0 0 0 0 27.03103 -82.20293 71 Tamiami Tr US41 PC Barry's Buildings 2 0 0 0 0 0 0 0 0 27.02243 -82.18902 481 Tamiami Tr US41 PC Tree Store 0 0 0 0 0 0 0 0 0 27.02177 -82.18862 514 Tamiami Tr US41 PC Class Auto Rentals 0 1 0 0 0 27.02092 -82.18858 531 Tamiami Tr US41 PC Port Charlotte Shops 0 0 0 0 0 26.99988 -82.12387 2195 Tamiami Tr US41 PC Barnes and Phillips 0 2 0 0 1 26.99955 -82.12352 2205 Tamiami Tr US41 PC Irongate North 2 0 0 5 0 0 0 0 0 26.99938 -82.12308 2241 Tamiami Tr US41 PC Eyeglass World 0 0 0 1 0 0 0 0 0 26.98563 -82.10480 3095 Tamiami Tr US41 PC Blueprint 0 0 0 2 0 0 0 0 0 26.98525 -82.01045 3109 Tamiami Tr US41 PC Sorentino Plaza 2 0 0 0 0 0 0 0 0 26.98502 -82.10307 3129 Tamiami Tr US41 PC Soto Building 3 0 0 1 0 0 0 0 0 26.98435 -82.10213 3195 Tamiami Tr US41 PC Suntrust Bank (Wachovia) 0 3 5 0 7 26.98342 -82.10187 3221 Tamiami Tr US41 PC Randal Realty 1 0 0 0 0 0 0 1 0 26.98330 -82.10167 3231 Tamiami Tr US41 PC Mini mall/ dry cleaners 1 3 1 0 1

PAGE 187

187 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m0 u 0 00 g0 m0 u g m u 26.97782 -82.09288 3575 Tamiami Tr US41 PC Chef's Garden/ Harry's Steakhouse 0 0 2 26.97760 -82.09368 3565 Tamiami Tr US41 PC Tire Kingdom 0 0 0 0 1 1 0 2 0 26.97695 -82.09300 3591 Tamiami Tr US41 PC La Romana 2 2 0 0 2 0 0 2 0 26.97070 -82.08403 4069 Tamiami Tr US41 PC The School Box 1 0 0 0 0 0 0 0 0 26.97062 -82.08392 4075 Tamiami Tr US41 PC Case Realty 0 0 0 0 1 0 1 0 0 26.97052 -82.08378 4079 Tamiami Tr US41 PC A-1 Mower 0 0 0 0 0 0 0 0 0 26.96805 -82.07957 4255 Tamiami Tr US41 PC Old Citgo By Town and Country Plaza 00 ---00 01 00 0001 0001 00 03 00 0 0 0 0 26.96805 -82.07957 4265 Tamiami Tr US41 PC Town and Country Plaza 0 1 0 0 2 26.96622 -82.07812 22627 Bayshore Rd (US41) PC Southtrust Bank 0 3 0 0 0 0 0 2 1 26.96025 -82.06860 4678 Tamiami Tr US41 PC Tamiami Motors 0 0 0 0 0 0 26.96013 -82.06852 4694 Tamiami Tr US41 PC Goodwill(Cronin's Patio Furniture) 0 0 0 0 0 26.95987 -82.06793 23188 Freedom Ave (US41) PC Coast Financial Advisors 1 0 0 0 4 26.95955 -82.06717 23212 Freedom Ave (US41) PC Jeff's Auto 0 0 0 0 0 0 0 0 0 26.95908 -82.06742 4732 Tamiami Tr US41 PC Touch of Class Carwash 0 0 0 0 1 26.95888 -82.06712 4760 Tamiami Tr US41 PC SkyslamFireworks 0 0 2 0 1 26.95307 -82.06193 5000 Tamiami Tr US41 PC Harbour Inn 3 1 0 2 2 0 26.95252 -82.06153 5041 Tamiami Tr US41 PC Sea Tow 1 0 0 4 0 0

PAGE 188

188 Table B-1 Continued. 2003 d 2004 d 2006 d Latitude Longitude Street # Street name Road # a City b Business c g m0 u 0 10 g2 m0 u g m u 26.95147 -82.06138 5054 Tamiami Tr US41 PC Charlotte Harbor Rawbar 0 0 3 26.90745 10 10 30 8100 2000 0000 0020 00 00 0000 00 -82.03965 3435 Tamiami Tr US41 PG SmokehouseRestaurant (barber) + back Bldg 0 3 0 0 0 26.90492 -82.03687 3655 Tamiami Tr US41 PG Pinch-a-PennyPool and Spa 0 8 0 3 6 26.90323 -82.03548 3691 Tamiami Tr US41 PG State Farm 0 0 0 1 0 0 0 0 0 26.88257 -82.01567 10175 Tamiami Tr US41 PG Kash N' Karry (Sweet Bay) 0 1 0 0 0 26.88120 -82.01418 10251 Tamiami Tr US41 PG Tae Kwan Do 1 0 0 1 0 0 26.87855 -82.01157 10381 Tamiami Tr US41 PG The Mecca (Goodwill) 0 0 0 0 1 26.84942 -81.98190 12350 Tamiami Tr US41 PG Davis Cypress 1 0 1 26.84713 -81.98057 12443 Tamiami Tr US41 PG House next to Tropical Gulf Acres Realty 0 0 0 0 0 26.84682 -81.97940 12459 Tamiami Tr US41 PG Tropical Gulf Acres Realty 0 1 0 0 0 26.84340 -81.97780 12594 Tamiami Tr US41 PG Brockman AutoSales 1 0 0 1 0 26.83973 -81.97655 12705 Tamiami Tr US41 PG Skip Epper's RVs 2 0 0 3 0 0 0 0 1 26.81250 -81.95790 15040 Tamiami Tr US41 PG Airborne Express (DHL) 0 0 0 0 0 (a) Road # in parenthesis indicates building was parallel to this road but the address was not on this road, (b) EW= Englewood, PC= Port Charlotte, PG= Punta Gorda, PL= Placida, (c) Business name in parenthesis indicates a change to a business name, (d) g= Hemidactylus garnotii, m= H. mabouia, u= unidentified individuals, -= missing observation.

PAGE 189

Table B-2. Lee County survey building locations and number of Hemidactylus garnotii and H. mabouia seen each survey. 2004 c 2006 c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.68237 -81.89818 532 Pine Island Rd CR 78 NFM Napa Auto Parts 0 0 0 0 3 0 0 0 0 0 1 1 0 0 0 0 1 0 26.68224 -81.88634 90 Pine Island Rd CR 78 NFM State Farm 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 26.68321 -81.87430 5080 Bayshore Rd CR 78 NFM Auto Zone 0 3 0 0 3 0 0 1 0 0 0 0 0 1 0 0 2 0 26.69701 -81.85232 6350 Bayshore Rd CR 78 NFM Abandoned bldg 0 2 0 0 1 0 0 1 0 0 0 0 0 1 0 0 0 0 26.70527 -81.84311 6950 Bayshore Rd CR 78 NFM Faith Assembly of God 1 4 0 1 3 0 0 3 0 0 0 0 1 4 1 0 6 0 26.71488 -81.80206 17181 Tarpon Way (CR78) NFM New HopeChristian Church 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 26.71495 -81.78875 10291 Bayshore Rd CR 78 NFM All Saints Byzantine Catholic Church 0 3 0 0 3 0 0 4 0 0 2 0 0 3 0 0 2 0 26.71473 -81.78664 10440 Bayshore Rd CR 78 NFM Hogbody's 0 3 1 0 3 3 0 4 1 0 1 0 0 1 6 0 1 2 26.71712 -81.60761 21361 N River Road CR 78 Alva Alva Church of God 0 4 0 0 6 0 0 3 0 0 3 2 0 4 2 0 5 0 26.70856 -81.61127 21200 Palm BeachBlvd CR 80 Alva Alva Country Diner 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 26.70960 -81.63373 19850 Palm BeachBlvd CR 80 Alva Intracoastal Real Estate 1 6 0 0 7 0 1 5 0 0 3 0 0 3 0 0 3 0 26.71289 -81.65623 18500 Palm BeachBlvd CR 80 Alva Alva Village Market 0 4 1 0 4 3 0 2 1 2 4 1 0 1 5 2 1 3 26.71443 -81.66822 17750 Palm BeachBlvd CR 80 Alva Countryside Wesleyan Church 1 2 0 0 1 0 0 2 0 0 1 0 0 1 0 0 0 0 26.71106 -81.71537 14830/14870 Palm Beach Blvd CR 80 FM Into All the World 7th Day Adventists 0 4 1 0 2 2 0 3 1 0 1 1 1 3 1 0 6 1 26.71033 -81.71837 14600 Palm BeachBlvd CR 80 FM Winn Dixie 0 0 0 0 1 0 1 1 0 0 1 0 0 0 0 1 3 0 189

PAGE 190

Table B-2 Continued. 2004 c 2006 c Survey Survey 1 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.70754 -81.73282 13806 Palm BeachBlvd CR 80 FM Oscar's Handy Pantry 0 0 8 0 0 7 0 0 5 0 0 4 0 0 10 0 0 7 26.70462 -81.74602 12925 Palm BeachBlvd CR 80 FM Sonshine Worship Center 1 2 2 0 0 4 1 0 2 0 0 10 2 2 6 1 1 9 26.70187 -81.75970 12002/12010 Palm Beach Blvd CR 80 FM Abandoned Bldg 0 1 3 0 1 4 0 0 2 2 3 1 0 1 3 1 2 2 26.69861 -81.76901 11431 Palm BeachBlvd CR 80 FM Victory Christian Center 0 1 0 0 2 0 0 3 0 0 1 0 0 1 0 0 1 0 26.68579 -81.79494 5450 Palm BeachBlvd CR 80 FM BP/ Amoco 0 1 0 0 0 1 0 0 1 0 0 0 0 1 0 0 1 0 26.67871 -81.80863 4841 Palm BeachBlvd CR 80 FM Spear and Assoc 0 1 0 0 0 1 0 3 0 0 0 1 0 0 0 0 0 0 26.67634 -81.81188 4786/4788 Palm Beach Blvd CR 80 FM Flexbon Paint 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 26.67546 -81.81469 4701 Palm BeachBlvd CR 80 FM Dairy Queen 1 1 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 1 26.66989 -81.82429 4401 Palm BeachBlvd CR 80 FM Come Back Inn 0 0 0 0 0 0 0 0 0 1 0 4 0 0 1 0 0 2 26.66512 -81.83230 4008 Palm BeachBlvd CR 80 FM Jack MacDonald Auto/ Coqui Restaurant 0 0 0 0 0 1 0 0 0 0 0 1 0 0 2 0 0 1 26.65333 -81.85402 2923 Palm BeachBlvd CR 80 FM Palm Beach Plaza 0 0 21 1 0 23 0 0 25 26.64720 -81.86455 2524 1st St CR 80 FM Fried & Fried Law Firm 1 0 2 1 0 1 1 0 1 1 0 2 0 0 1 0 0 2 26.52083 -81.87996 7152 Coca SabalLn (SR 865) FM Digestive Health 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 0 26.52055 -81.88656 7981 GladiolusDr SR 865 FM Nephrology 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 26.52092 -81.90005 8712-50 Gladiolus Dr SR 865 FM Shoppes at Paddle Creek 0 0 13 0 0 13 0 0 6 2 0 11 2 0 15 1 0 15 190

PAGE 191

Table B-2 Continued. vey 2004 c 2006 c Sur 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.52069 -81.90277 8950 Gladiolus Dr SR865 FM PrincetonAcademy 1 3 0 0 2 0 0 1 0 1 0 0 0 1 0 0 1 0 26.52036 -81.91305 9650 Gladiolus Dr SR865 FM Iona HopeEpiscopal Church 0 0 1 1 0 3 0 0 1 1 0 1 1 0 5 1 0 5 26.51848 -81.92535 10320 Gladiolus Dr SR865 FM BrightestHorizons Mission 0 1 1 0 2 0 0 1 0 0 0 0 0 0 0 0 1 0 26.51815 -81.92800 10511 Gladiolus Dr SR865 FM Joe'sOutreach and Thrift Store 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 26.51800 -81.93378 10891 Gladiolus Dr SR865 FM Luz en el Desierto Iglesia Cristiana 0 0 9 3 0 10 0 0 6 2 0 3 2 0 4 0 0 12 26.49744 -81.97105 16842 McGregorBlvd SR 867 FM ACEPerformer 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 26.51344 -81.95042 15690 McGregorBlvd SR 867 FM Faith UnitedMethodist Church 1 0 2 1 1 2 1 0 4 0 0 2 1 0 8 0 0 9 26.53906 -81.91924 13970 McGregorBlvd SR 867 FM DeBono'sMarket 0 1 0 2 3 2 0 2 0 0 0 2 1 0 3 1 0 4 26.54115 -81.91601 13761 McGregorBlvd SR 867 FM Foot &Ankle Care 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 26.54675 -81.91365 13391 McGregorBlvd SR 867 FM CypressOffice Bldg 3 1 18 2 1 18 0 0 16 5 0 11 0 0 22 2 0 24 26.57493 -81.89907 11481 McGregorBlvd SR 867 FM Clancey'sRestaurant 0 0 2 0 0 2 0 1 3 0 0 2 0 0 1 0 1 3 26.59208 -81.89226 10291 McGregorBlvd SR 867 FM SalvationArmy Worship Center 0 0 0 0 0 0 0 0 0 1 1 4 0 0 3 0 0 3 26.59969 -81.88947 4600 McGregorBlvd SR 867 FM Giovanni'sItalian Market/ Flower Spot 0 1 0 1 0 2 0 1 0 0 0 0 0 0 3 0 0 2 26.60162 -81.88846 4305 McGregorBlvd SR 867 FM McGregorCaf 2 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 191

PAGE 192

Table B-2 Continued. vey 2004 c 2006 c Sur 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.62043 -81.88395 3049 McGregorBlvd SR 867 FM Saint Johnthe Apostle Metro. Com. Church 0 1 0 0 4 0 0 1 0 0 0 0 0 1 0 0 3 0 26.62660 -81.88424 2756 McGregorBlvd SR 867 FM TrinityCommunity Church 0 1 0 0 0 0 0 0 1 0 0 1 0 0 0 0 1 0 26.63635 -81.87781 2259 McGregorBlvd SR 867 FM JerryKennedy Realty 0 0 0 0 0 1 0 0 3 26.74942 -81.91992 19591 N Tamiami Tr US 41 NFM Sprint Substation 0 1 0 0 3 0 0 0 0 0 1 0 0 2 0 0 0 0 26.68248 -81.90124 15050 N Cleveland Ave US 41 NFM Eckerds 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 1 0 0 26.68109 -81.90101 14940 N ClevelandAve US 41 NFM Bedding Outlets 0 0 2 0 0 1 0 0 0 0 0 1 0 0 1 0 0 1 26.66024 -81.88455 13181 N ClevelandAve US 41 NFM Abandoned Fast Food 0 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 2 26.64148 -81.87309 1542 Carson St (US41) FM RivaConsulting 0 2 0 0 0 0 0 1 0 0 1 3 0 1 1 0 2 2 26.63501 -81.87246 2275 Cleveland Ave US 41 FM Abandoned Orange bldg 0 0 3 0 0 5 0 0 3 0 0 0 0 0 1 0 0 1 26.62867 -81.87223 2635 Cleveland Ave US 41 FM St Luke's Episcopal Church 0 1 0 0 4 0 0 2 1 2 0 6 0 0 3 0 1 5 26.58748 -81.87218 5030 S ClevelandAve US 41 FM Abandoned bldg near Blockers 0 0 1 0 0 1 0 0 0 26.55366 -81.87116 12879 S ClevelandAve US 41 FM A Party World 1 0 2 0 0 1 0 0 2 0 0 0 0 0 1 0 0 0 26.53146 -81.87162 14500 S Tamiami Tr US 41 FM 8 ball Lounge 0 0 1 0 0 2 0 0 0 1 0 1 0 0 2 0 0 2 26.52280 -81.87072 15120 S Tamiami Tr US 41 FM Modern Auto/ Beverly Hills Tinting 0 0 0 0 0 0 0 0 1 1 0 5 0 0 9 0 0 7 192

PAGE 193

193 Table B-2 Continued. vey 2004 c 2006 c Sur 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.51754 -81.86710 15460 S Tamiami Tr US 41 FM Tires Plus 0 0 1 0 4 0 0 2 2 1 0 4 0 0 5 0 0 5 26.50218 -81.85648 16520 S Tamiami Tr US 41 FM Southtrust Bank 0 0 0 e 0 -82.00894 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 26.49528 -81.84986 16850 S Tamiami Tr US 41 FM Classics Forever Furniture 0 0 5 0 0 2 0 0 2 0 0 2 0 0 3 0 0 2 26.47629 -81.83793 18080 S Tamiami Tr US 41 FM Citgo 0 0 2 1 1 3 0 1 1 0 0 2 0 0 4 0 0 4 26.46666 -81.83256 18767 S Tamiami Tr US 41 FM Andre's Steak House 0 6 0 0 5 0 0 3 0 0 3 0 0 0 0 1 1 0 26.73765 -81.76129 18841 SR 31 CR 78 FM Temple Baptist 1 1 0 0 2 0 0 0 1 0 0 0 0 1 0 0 1 0 26.36132 -81.78844 4800 Palm BeachBlvd CR 80 FM American Legion 0 1 5 1 1 3 0 0 2 0 1 0 0 2 1 0 1 2 26.68098 -81.90100 14930 N ClevelandAve US 41 FM Video 0 0 0 0 0 0 0 0 2 0 0 1 0 0 1 0 0 1 26.49785 -81.97093 16836 McGregorBlvd SR 867 FM Benchmark 0 0 0 0 0 2 0 1 2 0 0 2 0 0 1 0 0 4 26.50233 -81.96414 16581 McGregorBlvd SR 867 FM Bread ofLife Ministries 0 1 0 0 2 0 0 2 0 0 2 0 0 1 1 0 1 0 26.51431 -81.94948 15630 McGregorBlvd SR 867 FM NexttoSantiva Plaza 0 0 8 0 0 10 1 0 9 0 0 8 0 0 8 0 0 12 26.53091 -81.87173 14540 S Tamiami Tr US 41 FM Bike Route 0 0 7 0 0 9 0 0 5 0 0 1 0 0 7 0 0 5 26.66461 -81.81247 11751 S Tamiami Tr US 41 FM Kinkos 0 0 0 0 0 1 0 2 0 0 0 0 0 0 1 0 0 0 26.62721 -82.07301 4700 NW PineIsland Rd CR 78 Mat Real Estate 0 1 26.63782 -82.04857 12016 Matlacha Blvd (CR78) CC BackbayHomes 0 1 26.63733 -82.04782 3640 SW PineIsland Rd CR 78 CC Greater Pine Island Chamber of Commerc 0 1 26.63787 1630 SW PineIsland Rd CR 78 CC Coral Ridge Cemetery 1 2 26.67950 -81.91672 2691 NE Pine Island Rd CR 78 CC Messiah Lutheran Church 0 2 0

PAGE 194

Table B-2 Continued. vey 2004 c 2006 c Sur 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.68079 -81.91345 2805 NE Pine Island Rd CR 78 NFM New Testament Baptist Church 0 2 0 26.68085 -82.14944 7573 Raymary St (CR767) Bok Hair Net 0 0 0 26.66361 -82.13976 13960 StringfellowRd CR 767 Bok Pine IslandExecutive Suites 0 0 1 26.63751 -82.12522 12175 StringfellowRd CR 767 Bok Our Lady ofthe Miraculous Medal Catholic Church 2nd bldg (S of Main Bldg) 0 5 1 26.62033 -82.12006 10954 StringfellowRd CR 767 Bok Burnt StoreProperties/ Pine Island Printing 0 2 0 26.61593 -82.11849 10700 StringfellowRd CR 767 Bok IslandExchange 0 1 0 26.63614 -82.06495 10530 StringfellowRd CR 78 Bok Sunbelt Realty 0 1 1 26.60534 -82.11492 9830-9872 Stringfellow Rd CR 767 SJC Winn Dixieshopping Center 0 0 2 26.59047 -82.11064 8903 StringfellowRd CR 767 SJC MooseLodge 0 0 9 26.52556 -82.08622 4480 StringfellowRd CR 767 SJC Post Office 0 3 0 26.51997 -82.08499 4106 StringfellowRd CR 767 SJC AmericanLegion 0 1 13 26.51224 -82.08320 3366 StringfellowRd CR 767 SJC SellstateIsland Realty 0 2 1 194

PAGE 195

Table B-2 Continued. vey 2004 c 2006 c Sur 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road # a City b Business U G M U G M U G M U G M U G M U G M 26.61763 -81.81148 3451 Ortiz Ave CR865 FM 7th DayAdventist Church 0 0 0 26.64072 26.59332 -81.81283 10511 Ben C Pratt Six Mile Cypress Rd 865 FM Banks Engineering 0 10 26.59375 -81.81206 10491 Ben C Pratt Six Mile Cypress Rd CR 865 FM Six Mile Cypress Office Park 0 30 26.60955 -81.81061 8081 Dani Dr (CR 865) FM Bank of America 0 00 26.52762 -81.85660 14921 Technology Court (CR 865) FM Wallpaper World 0 00 26.50284 -81.94321 16385 San Carlos Blvd SR 865 FM DMS Custom Choppers 0 02 26.50192 -81.94365 16450 San Carlos Blvd SR 865 FM One Stop Outlet 1 09 26.48919 -81.94297 17305 San Carlos Blvd SR 865 FMB Faith Baptist Church 1 81 26.48264 -81.94638 17811 San Carlos Blvd SR 865 FMB Enterprise 0 03 26.48222 -81.94702 17853 San Carlos Blvd SR 865 FMB Mermaid Grille 0 02 26.47926 -81.95144 17980 San Carlos Blvd SR 865 FMB Empty Office 0 04 -81.81160 2041 Ortiz Ave CR865 FM Outpost Barand Grill 0 0 0 26.65200 -81.81186 1441-71 Ortiz Ave CR 865 FM LaMexicana 0 0 1 26.66461 -81.81247 724 Ortiz Ave CR865 FM Penny SaverSupermarket 1 0 1 26.52864 -81.85325 14690 Metro Pkwy (SR865) CR FM Hess 1 1 0 195

PAGE 196

196Table B-2 Continued. 2004c2006c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road #aCitybBusiness UGM U GM UGM UGM UGM UGM 26.46378 -81.95161 19051 San Carlos Blvd SR 865 FMB Beach Plaza/ Island Interiors 1 133 26.45046 -81.94736 1901 Estero Blvd CR 865 FMB Tattoo/ May's 0 06 26.44130 -81.92841 3751 Estero Blvd CR 865 FMB St Peter Lutheran Church 2 09 26.42490 -81.90614 6051 Estero Blvd CR 865 FMB Dentist/ Florist 0 012 26.41431 -81.90029 7205 Estero Blvd CR 865 FMB Skye's/ Italian Fisherman/ Lions Den 0 111 26.35825 -81.85803 26105 Hickory Ave CR 865 BS Hickory Island Center 0 03 26.33132 -81.78471 27911 Vermont St (CR 865) BS Living Waters Community Church 0 12 26.32938 -81.80954 3421 Bonita Beach Rd SW CR 865 BS Beach Road Plaza 0 02 26.33071 -81.83265 4836 Bonita Beach Rd SW CR 865 BS Ditto Kiddo/ House of Brass 0 02 26.32329 -81.78477 28480 Old 41 CR 887 BS Brand New Strip Mall 0 01 26.36132 -81.78844 25999 Old 41 CR 887 BS Hope Lutheran Church 1 15 26.56523 -81.94370 4635 Del Prado Blvd S CR 867A CC Law Offices 0 00

PAGE 197

197Table B-2 Continued. 2004c2006c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road #aCitybBusiness UGM U GM UGM UGM UGM UGM 26.58054 -81.94268 3818 Del Prado Blvd S CR 867A CC Eagle Realty/ Allied Construction 0 00 26.58065 -81.94265 3810 Del Prado Blvd S CR 867A CC Books New and Used 0 00 26.59573 -81.94184 3030-8 Del Prado Blvd S CR 867A CC Subway/ Sakura/ Pools & Spas 0 00 26.59653 -81.94190 3002 Del Prado Blvd S CR 867A CC State Farm Insurance 0 00 26.60937 -81.94135 2302 Del Prado Blvd S CR 867A CC Stacie Lee Nails/ Palm & Tarot Readings 0 00 26.63522 -81.94101 910 Del Prado Blvd S CR 867A CC Car Quest Auto Parts 0 00 26.63568 -81.94144 902 Del Prado Blvd S CR 867A CC Abandoned Gas Station 0 00 26.64395 -81.94079 455 Del Prado Blvd S CR 867A CC General Dentistry 0 00 26.65252 -81.94086 9 Del Prado Blvd N CR 867A CC Gulf Coast Realty 0 10 26.66687 -81.94127 798 Del Prado Blvd N CR 867A CC 7 Eleven 0 20 26.54680 -81.80242 9230 Daniels Pkwy CR 876 FM Tuffy/ Eagle Cleaners/ Real Estate 1 00 26.54557 -81.90744 6715 Winkler (CR 876) FM Wachovia 0 00 26.54545 -81.90805 9231 Cypress Lake Blvd CR 876 FM Phil Kiser Funeral Home 0 30 26.54538 -81.91316 9571 Cypress Lake Blvd CR 876 FM Kurokawa Martial Arts 0 01

PAGE 198

198Table B-2 Continued. 2004c2006c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road #aCitybBusiness UGM U GM UGM UGM UGM UGM 26.64725 -81.85835 2663 2nd st SR 80 FM Peniel 7th Day Adventist Church 1 00 26.64104 -81.86905 1822 Broadway St (SR 82) FM Perkin's Bail Bonds 0 00 26.63703 -81.80815 2240 Park Dr (SR 82) FM Citgo 0 01 26.60218 -81.74778 12777 SR 82 SR 82 FM Lee Memorial Park Cemetery 0 00 26.58754 -81.88327 28 Barkley Circle (SR 869) FM Risk Management Insurance 0 00 26.58532 -81.88313 5172 Mason Corbin (SR 869) FM N of Foot & Ankle Clinic 0 00 26.57823 -81.88138 5285 Summerlin SR 869 FM Summerlin Medical Clinic 0 63 26.56215 -81.88360 1537C Brantley Rd (SR 869) FM Atkins & Assoc Insurance 1 50 26.51528 -81.89677 15620 New Hampshire Ct (SR 869) FM Eyelid Surgery 0 00 26.49325 -81.93019 17080 Safety St (SR 869) FM No Worry's Racing/ Quality Cheese 0 03 26.49306 -81.96828 20021 Summerlin Rd SR 869 FM Circle K/ 76 0 00 26.60294 -81.62678 200 E Leeland Heights Blvd SR 884 LA 1st Community Congressional Church 0 30 26.60273 -81.63856 30 Colorado Rd (SR 884) LA Realty World 0 00

PAGE 199

199Table B-2 Continued. 2004c2006c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road #aCitybBusiness UGM U GM UGM UGM UGM UGM 26.61473 -81.65080 1154 Lee Blvd SR 884 LA Pinewood Medical Plaza 0 16 26.61465 -81.65627 1400-1470 Lee Blvd SR 884 LA Lee Blvd Plaza 1 20 26.61690 -81.67592 403 Joan Ave N (SR 884) LA Allstate 0 10 26.61698 -81.68453 3305 Lee Blvd SR 884 LA Professional Mechanic 0 11 26.61660 -81.71951 4519 Lee Blvd SR 884 LA Century 21 0 00 26.62057 -81.75786 5846 8th St W (SR 884) LA Round Yellow Bldg 0 10 26.59711 -81.84096 3550 Colonial Blvd SR 884 FM Abandoned Hospital 0 00 26.59690 -81.88817 1409 Colonial Blvd SR 884 FM Abandoned Gas Station 0 01 26.63860 -81.59800 1050 Joel Blvd SR 884 LA Elks Lodge 2 013 26.68102 -81.59856 2205-21 Joel Blvd SR 884 Alva Lippincott Square 2 20 26.70064 -81.60078 2790 Joel Blvd SR 884 Alva 1st Baptist Church of Alva 0 20 26.70870 -81.60114 21870 Palm Beach Blvd SR 80 Alva Lee's Trees 0 00 26.44286 -81.81187 20400 Trailside Dr (US 41) Est Salon & Day Spa 0 00 26.44081 -81.81085 20583 S Tamiami Tr US 41 Est Estero Real Estate 0 00 26.43808 -81.81070 20771 S Tamiami Tr US 41 Est Ro-Lin Rentals/ Salon 0 22 26.41895 -81.81213 22050 S Tamiami Tr US 41 Est West Bay Club Sales Center 0 00

PAGE 200

200Table B-2 Continued. 2004c2006c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road #aCitybBusiness UGM U GM UGM UGM UGM UGM 26.38893 -81.80989 3440 Renaissance Blvd (US 41) BS JB's/ Pizza Hut/ Blind Expert 0 00 26.38793 -81.80861 24241 S Tamiami Tr US 41 BS Allstate/ Suntamers/ Resale Furniture 0 03 26.34353 -81.80577 27241 Bay Landing Dr (US 41) BS Baylanding/Arla Bonten Realty 1 60 26.33377 -81.80579 27821 S Tamiami Tr US 41 BS Ameri Gas/ Pro Audio Lighting 0 10 26.32755 -81.80600 28194 S Tamiami Tr US 41 BS Best of Everything Jewelry Outlet 0 00 26.68787 -81.88605 2001 N Tamiami Trail NE USB 41 FM North Ft. Myers Library 0 30 26.67036 -81.88259 1357 N Tamiami Trail NE USB 41 FM Wonder Bread 0 13 26.66283 -81.87743 1027 N Tamiami Trail NE USB 41 FM Printmaking 1 37 26.64486 -81.86467 1651 Fowler St USB 41 FM Bill Smith Electronics 2 00 26.63485 -81.86236 2280-90 Fowler St USB 41 FM AAIM Engine Machine/ Tandem 1 02 26.63159 -81.86207 2509 Fowler St USB 41 FM Enterprise 0 00 26.62744 -81.86236 2730 Fowler St USB 41 FM Used Appliance Center 0 00 26.59913 -81.86199 4350 Fowler St USB 41 FM Edison Square/ Tubby's 0 02

PAGE 201

201Table B-2 Continued. 2004c2006c Survey 1 Survey 2 Survey 3 Survey 1 Survey 2 Survey 3 Latitude Longitude Street # Street name Road #aCitybBusiness UGM U GM UGM UGM UGM UGM 26.60549 -81.85076 3940 Metro Pkwy USB 41 FM Parkway Center/Nextel/ Quiznos 0 015 26.59395 -81.84984 10231 Metro Pkwy USB 41 FM Parkway Plaza 0 01 26.58563 -81.85355 10800 Metro Pkwy USB 41 FM Circle K 0 01 26.58243 -81.85358 10994 Metro Pkwy USB 41 FM Broedell 0 02 26.56890 -81.85273 11931 Metro Pkwy USB 41 FM Duron 0 03 (a) Road # in parenthesis indicates building was parallel to this road but the address was not on this road, (b) Bok= Bokeelia, BS= Bonita Springs, CC= Cape Coral, Est= Estero, FMB= Fort Myers Beach, FM= Fort Myers, LA= Lehigh Acres, Mat= Matlacha, NFM= North Fort Myers, SJC= St. James City (c) g= Hemidactylus garnotii, m= H. mabouia, u= unidentified individuals, -= missing observation.

PAGE 202

APPENDIX C GECKOS CAPTURED IN SARASOTA COUNTY Table C-1. Geckos captured in Sarasota Date Lat Long Street # Street Name Road # Miscellaneous information City Species 5/15/2005 27.03866 -82.21812 15121-81 Tamiami Trail US41 North Port Health Park North Port none 5/15/2005 27.04244 -82.23047 14601 Tamiami Trail US41 Prudential North Port none 5/15/2005 27.04276 -82.23088 14595 Tamiami Trail US41 Shark's Restaurant North Port H. garnotii 5/15/2005 27.04279 -82.23139 14575 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04303 -82.23193 14525 Tamiami Trail US41 North Port none 5/15/2005 27.04318 -82.23230 14505 Tamiami Trail US41 North Port none 5/15/2005 27.04378 -82.23421 14375-85 Tamiami Trail US41 North Port none 5/15/2005 27.04400 -82.23472 14355 Tamiami Trail US41 Babe's Hardware North Port H. garnotii 5/15/2005 27.04263 -82.23562 14380 Tamiami Trail US41 San Pedro Church Main Church bldg North Port none 5/15/2005 27.04217 -82.23473 14380 Tamiami Trail US41 San Pedro Church Parish Life Center North Port H. garnotii 5/15/2005 27.04168 -82.23504 14380 Tamiami Trail US41 San Pedro Church Activity Center NorthPort H.garnotii 5/15/2005 27.32062 -82.97009 14380 Tamiami Trail US41 San Pedro Church Office North Port none 5/15/2005 27.04210 -82.23379 14400 Tamiami Trail US41 Medical Arts Plaza North Port none 5/15/2005 27.04166 -82.23222 14538 Tamiami Trail US41 James McKee Funeral Home North Port none 5/15/2005 27.04170 -82.23263 14512-30 Tamiami Trail US41 Alvaro's Restaurant North Port none 5/15/2005 27.04183 -82.23298 14506 Tamiami Trail US41 Holiday Plaza North Port none 5/15/2005 27.04191 -82.23317 14500 Tamiami Trail US41 North Port Veterinarian North Port none 5/15/2005 27.04131 -82.23132 14580 Tamiami Trail US41 Bicycle Shop etc North Port H. garnotii 5/15/2005 27.04640 -82.24587 13807-31 Tamiami Trail US41 Medical Plaza North Port H. garnotii 5/15/2005 27.04648 -82.24648 13801 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04643 -82.24746 5900 Biscayne Drive / Tamiami US41 Funeral Home North Port none 5/15/2005 27.14660 -82.24993 13601-65 Tamiami Trail US41 North Port none 5/15/2005 27.04685 -82.25221 13355 Tamiami Trail US41 North Port none 5/15/2005 27.04692 -82.25283 13221 Tamiami Trail US41 North Port none 5/15/2005 27.04697 -82.25315 13201 Tamiami Trail US41 North Port none 5/15/2005 27.04695 -82.25353 13125 Tamiami Trail US41 North Port none 202

PAGE 203

203 Table C-1 Continued. Date Lat Long Street # Street Name Road # Miscellaneous information City Species 5/15/2005 27.04700 -82.25423 13001-35 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04729 -82.25796 12705-11 Tamiami Trail US41 North Port Both 5/15/2005 27.04688 -82.25588 12781 Tamiami Trail US41 North Port none 5/15/2005 27.04716 -82.25679 12731-67 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04496 -82.25319 13000 Tamiami Trail US41 Church North Port none 5/15/2005 27.04478 -82.24869 13600-88 Tamiami Trail US41 Biscayne Plaza North Port H. garnotii 5/15/2005 27.04397 -82.24043 14132 Tamiami Trail US41 Family Table North Port none 5/15/2005 27.04390 -82.24021 14140 Tamiami Trail US41 State Farm North Port none 5/15/2005 27.04397 -82.23988 14156 Tamiami Trail US41 Mosse Lodge North Port none 5/15/2005 27.04377 -82.23955 14164 Tamiami Trail US41 Car Wash North Port none 5/15/2005 27.04358 -82.23872 14212 Tamiami Trail US41 Quality TAV North Port none 5/15/2005 27.04349 -82.23822 14236 Tamiami Trail US41 Cedar Profesional Center North Port none 5/15/2005 27.04339 -82.23797 14242 Tamiami Trail US41 Audrey's North Port none 5/15/2005 27.04339 -82.23743 14260 Tamiami Trail US41 North Port Century 21 North Port none 5/15/2005 27.04208 -82.23472 14380 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04397 -82.23467 14355 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04279 -82.23138 14575 Tamiami Trail US41 North Port H. garnotii 5/15/2005 27.04729 -82.25796 12705-11 Tamiami Trail US41 Northport H. mabouia 11/21/2005 27.09030 -82.43165 851-59 US 41 Bypass S US41 S. Venice none 11/21/2005 27.08986 -82.43124 901 US 41 Bypass S US41 S. Venice H. garnotii 11/21/2005 27.08942 -82.43096 955 US 41 Bypass S US41 S. Venice none 11/21/2005 27.08334 -82.42596 1267 US 41 Bypass S US41 S. Venice none 11/21/2005 27.06650 -82.41911 1807 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06621 -82.41892 1829-39 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06568 -82.41866 1841 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06548 -82.41854 1843 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06535 -82.41837 1847-55 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06396 -82.41732 1861 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06374 -82.41723 1865 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06353 -82.41706 1867 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06346 -82.41791 1868 S Tamiami Trail US41 S. Venice H. mabouia 11/21/2005 27.06331 -82.41683 1869 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06305 -82.41766 1872 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06301 -82.41674 1875 S Tamiami Trail US41 S. Venice none

PAGE 204

204 Table C-1 Continued. Date Lat Long Street # Street Name Road # Miscellaneous information City Species 11/21/2005 27.06289 -82.41650 1879 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06283 -82.41726 1880 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06255 -82.41714 1884 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.06231 -82.41709 1890 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06134 -82.41545 1915-25 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06095 -82.41520 1935 S Tamiami Trail US41 S. Venice H. mabouia 11/21/2005 27.06064 -82.41509 1939 S Tamiami Trail US41 S. Venice none 11/21/2005 27.06042 -82.41478 1945 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.05492 -82.41078 2131-43 S Tamiami Trail US41 S. Venice none 11/21/2005 27.05443 -82.41042 2155-75 S Tamiami Trail US41 S. Venice none 11/21/2005 27.05306 -82.40963 2225 S Tamiami Trail US41 S. Venice none 11/21/2005 27.05334 -82.40981 2227-33 S Tamiami Trail US41 S. Venice none 11/21/2005 27.05069 -82.40659 2357 S Tamiami Trail US41 S. Venice none 11/21/2005 27.05011 -82.40313 2383 S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.04945 -82.40154 3944-86 S Tamiami Trail US41 S. Venice none 11/21/2005 27.04909 -82.40131 3990 unit A S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.04881 -82.40135 3990 Unit C S Tamiami Trail US41 S. Venice H. garnotii 11/21/2005 27.05115 -82.40215 2401 Seaboard S. Venice H. garnotii 11/21/2005 27.05198 -82.40306 2375 Seaboard S. Venice H. garnotii 11/21/2005 27.05289 -82.40381 2235 Seaboard S. Venice none 11/21/2005 27.05153 -82.40394 351 Orange S. Venice H. garnotii 11/2/2005 27.01123 -82.38346 2222 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 27.00924 -82.38131 2201 Englewood Rd SR776 Englewood none 11/2/2005 27.00712 -82.38150 2100 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 27.00611 -82.37976 2051 Englewood SR776 Englewood H. garnotii 11/2/2005 27.00623 -82.37989 2061 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 27.00542 -82.37925 2021 Englewood Rd SR776 Englewood Both 11/2/2005 27.00506 -82.37905 2011 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 27.00382 -82.37847 1975 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 27.00331 -82.37815 1959 Englewood Rd SR776 Englewood none 11/2/2005 27.00309 -82.37802 1949 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 27.00072 -82.37631 1885 Englewood Rd SR776 Englewood H. garnotii

PAGE 205

205 Table C-1 Continued. Date Lat Long Street # Street Name Road # Miscellaneous information City Species 11/2/2005 26.99513 -82.37243 1600 Englewood Rd SR776 Englewood H. garnotii 11/2/2005 26.97587 -82.35577 726 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97565 -82.35561 720 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 27.97546 -82.35556 712 Indiana Ave N SR776 Englewood none 11/2/2005 26.97520 -82.35424 701 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97449 -82.35386 685 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97406 -82.35362 655 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97399 -82.35446 650 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97360 -82.35469 634 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97436 -82.35465 686 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.97138 -82.35336 500-514 Indiana Ave N SR776 Englewood none 11/2/2005 26.97101 -82.35339 450 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.96969 -82.35399 406 Indiana Ave N SR776 rear bldg Englewood Both 11/2/2005 26.96970 -82.35350 406 Indiana Ave N SR776 front bldg Englewood none 11/2/2005 26.96916 -82.35320 390 Indiana Ave N SR776 Englewood H. mabouia 11/2/2005 26.96821 -82.35322 340 Indiana Ave N SR776 Englewood H. mabouia 11/2/2005 26.96734 -82.35320 264 Indiana Ave N SR776 Englewood none 11/2/2005 26.96703 -82.35318 242 Indiana Ave N SR776 Englewood none 11/2/2005 26.96549 -82.35332 180 Indiana Ave N SR776 Englewood none 11/2/2005 26.96476 -82.35321 140 Indiana Ave N SR776 Englewood none 11/2/2005 26.96431 -82.35321 120 Indiana Ave N SR776 Englewood Both 11/2/2005 26.96359 -82.35323 70 Indiana Ave N Englewood H. mabouia 11/2/2005 26.96091 -82.35363 31 Green St Englewood H. garnotii 11/2/2005 26.96060 -82.35323 60 Indiana Ave N SR776 Englewood H. garnotii 11/2/2005 26.96090 -82.35321 50 Green St/ Indiana Englewood none 11/2/2005 26.95453 -82.35334 400 Indiana AveS SR776 Englewood none 11/2/2005 26.95425 -82.35329 450 Indiana Ave S SR776 Englewood none 11/2/2005 26.95401 -82.35333 460 Indiana Ave S SR776 Englewood none 11/2/2005 26.95419 -82.35281 473 Indiana Ave S SR776 Englewood none 11/2/2005 26.95365 -82.35275 501 Indiana Ave S SR776 Englewood H. garnotii 11/2/2005 26.95268 -82.35293 579 Indiana Ave S SR776 Englewood H. garnotii 11/2/2005 26.95119 -82.35293 649 Indiana Ave S SR776 Englewood H. garnotii 11/2/2005 26.95107 -82.35299 651 Indiana Ave S SR776 Englewood H. garnotii 11/2/2005 26.95078 -82.35294 659 Indiana Ave S SR776 Englewood H. garnotii

PAGE 206

206 Table C-1 Continued. Date Lat Long Street # Street Name Road # Miscellaneous information City Species 11/2/2005 26.95052 -82.35295 699 Indiana AveS SR776 Englewood H. garnotii 11/2/2005 26.94908 -82.35295 737 Indiana Ave S SR776 Englewood H. garnotii 11/2/2005 26.94845 -82.35296 777 Indiana Ave S SR776 Englewood H. garnotii 11/2/2005 26.94875 -82.35289 773 Indiana Ave S SR776 Englewood H. garnotii

PAGE 207

LITERATURE CITED AGRESTI, A., AND B. A. COULL. 1998. Approximate is better than 'exact' for interval estimation of binomial proportions. The American Statistician 52(2):119-126. AKAIKE, H. 1973. Information theory and an extension of the maximum likelihood principle. Pages 267-281 in B. N. Petran, and F. Csaaki, editors. Second International Symposium on Information Theory. Akadeemiai Kiadi, Budapest, Hungary. ANDERSON, D. R., K. P. BURNHAM, AND G. C. WHITE. 1994. AIC model selection in overdispersed capture-recapture data. Ecology 75(6):1780-1793. ANDREWS, R. M. 1988. Demographic correlates of variable egg survival for a tropical lizard. Oecologia 76(3):376 -382. ARNASON, A. N. 1972. Parameter estimates from mark-recapture experiments on two populations subject to migration and death. Researches on Population Ecology 13:97-113. -----. 1973. The estimation of population size, migration rates, and survival in a stratified population. Researches on Population Ecology 15:1-8. ARNOLD, E. N. 1984. Evolutionary aspects of tail shedding in lizards and their relatives. Journal of Natural History 18(1):127-169. -----. 1994. Caudal autotomy as a defense, p. 253-273, In: Gans, C. and Huey, R. B. (eds.), Biology of the reptilian, Volume 16, Ecology B, Defense and life history. Branta Books, Ann Arbor, MI, USA. ARAUJO, A. F. B. 1991. Structure of a white sand-dune lizard community of coastal Brazil. Revista Brasileira De Biologia 51(4):857-865. BARBOUR, T. 1936. Two introduced lizards in Miami, Florida. Copeia 1936(2):113. BARTLETT, R.D., AND P. P. BARTLETT. 1999. A field guide to Florida reptiles and amphibians. Gulf Publishing Company, Houston, TX, USA. BAUER, A. M., AND K. HENLE. 1994. Familia Gekkonidae (Reptilia, Sauria): part I Australia and Oceania. in: H. Wermuth, and M. Fischer (editors), Das tierreich: the animal kingdom, a compilation and characterization of the recent animal groups, part 109. Walter de Gruyter., Berlin, Germany. BEEBE, W. 1944. Field notes on the lizards of Kartabo, British Guiana, and Caripito, Venezuela. Part 1. Gekkonidae. Zoologica 29:145-160. BEHLER J. L., AND F. W. KING. 1979. The Audubon Society field guide to North American reptiles and amphibians. Alfred A. Knopf, New York, USA. 207

PAGE 208

BENABIB, M. 1994. Reproduction and lipid utilization of tropical populations of Sceloporus variabilis. Herpetological Monographs 8:160-180. BLOMBERG, S. P., AND R. SHINE. 2000. Size-based predation by kookaburras (Dacelo novaeguineae) on lizards (Eulamprus tympanum: Scincidae): what determines prey vulnerability? Behavioral Ecology and Sociobiology 48(6):484-489. BOCK, B. C. 1996. Interclutch interval and egg aggregations in the tropical house gecko, Hemidactylus mabouia. Herpetological Review 27(4):181-183. BOLGER, D. T., AND T. J. CASE. 1992. Intraand interspecific interference behaviour among sexual and asexual geckos. Animal Behaviour 44:21-30. BOURQUIN, O. 1987. The recent geographical range extension of Hemidactylus mabouia mabouia. Lammergeyer 38:12-14. BRANCH, W. R. 1987. Introduced reptiles in the Addo Elephant National Park. Koedoe 30:165. BROADLEY, D. G. 1977. A review of the Hemidactylus mabouia complex in southeastern Africa (Sauria: Gekkonidae). Arnoldia Rhodesia 8(19):1-15. BROWN, J. H., D. A. KELT, AND B. J. FOX. 2002. Assembly rules and competition in desert rodents. American Naturalist 160(6) :815 BROWN, S. G., AND P. K. DUFFY. 1992. The effects of egg-laying site, temperature, and salt water on incubation time and hatching success in the gecko Lepidodactylus lugubris. Journal of Herpetology 26(4):510-513. BROWN, W. C., AND A. C. ALCALA. 1978. Philippine lizards of the family Gekkonidae. Silliman University Press, Dumaguete City, Philippines. BROWNIE, C., J. E. HINES, J. D. NICHOLS, K. H. POLLOCK, AND J. B. HESTBECK. 1993. Capture-recapture studies for multiple strata including non-Markovian transitions. Biometrics 49:1173-1187. BUCKLAND, S. T., AND P. H. GARTHWAITE. 1991. Quantifying precision of markrecapture estimates using the bootstrap and related methods. Biometrics 47:255-268. BURNHAM, K. P., AND D. R. ANDERSON. 1992. Data-based selection of an appropriate biological model: the key to modern data analysis. Pages 16-30 in D. R. McCullough, and R. H. Barrett, editors. Wildlife 2001: Populations. Elsevier Appl. Sci., New York. -----, and -----. 1998. Model selection and inference: a practical information-theoretic approach. Springer-Verlag, New York. 208

PAGE 209

-----, and -----. 2002. Model selection and multimodel inference: a practical information-theoretic approach, 2 nd edition. Springer-Verlag, New York. -----, -----, G. C. WHITE, C. BROWNIE, AND K. POLLOCK. 1987. Design and analysis methods for fish survival experiments based on release-recapture. American Fisheries Society Monograph 5:1-437. BUTTERFIELD, B. P. 1996. Patterns and processes of invasions by amphibians and reptiles into the West Indies and south Florida. PhD Dissertation, Auburn University. -----, B. HAUGE, AND W. E. JR. MESHAKA. 1993. The occurrence of Hemidactylus mabouia on the United States mainland. Herpetological Review 24(3):111-112. -----, W. E. MESHAKA, JR., AND C. GUYER. 1997. Nonindigenous amphibians and reptiles, In: D. Simberloff, D. C. Schmitz, and T. C.Brown, editors, Strangers in paradise: impact and management of nonindigenous species in Florida. Island Press, Washington, D.C., USA. CAGLE, F. R. 1946. Tail loss and regeneration in a Pacific island gecko. Copeia 1946(1):45. CAMPBELL, T. S. 2000. Analysis of the effects of an exotic lizard (Anolis sagrei) on a native lizard (Anolis carolinensis) in Florida, using islands as experimental units. Ph. D. Dissertation, University of Tennessee, Knoxville, TN. CARRANZA, S., AND E. N. ARNOLD. 2006. Systematics, biogeography, and evolution of Hemidactylus geckos (Reptilia : Gekkonidae) elucidated using mitochondrial DNA sequences. Molecular Phylogenetics and Evolution 38(2):531-545. CASE, T. J. 1983. Sympatry and size similarity in Cnemidophorus. Pages 297-325, In: R. B. Huey, E. R. Pianka, and T. W. Schoener, editors, Lizard ecology: studies of a model organism, Harvard University Press, Cambridge, Massachusetts, USA. -----, D. T. BOLGER, AND K. PETREN. 1994. Invasions and competitive displacement among house geckos in the tropical Pacific. Ecology 75(2):464-477. CHARLOTTE COUNTY. 2007a. Charlotte County Geographical Information Systems (GIS), http://www.ccgis.com. -----. 2007b. Charlotte County Government Community Development Department, http://charlottecountyfl.com/CommunityDevelopment/PlanningZoning. CHOQUET, R., A. M. REBOULET, J. D. LEBRETON, O. GIMENEZ, AND R. PRADEL. 2005. U-CARE 2.2 Users Manual. CEFE, Montpellier, France. (http://ftp.cefe.cnrs.fr/biom/Soft-CR/). CONANT, R., AND J. T. COLLINS. 1991. A field guide to reptiles and amphibians, eastern and central North America, third edition. Houghton Mifflin Company, Boston, MA, USA. 209

PAGE 210

-----, AND -----. 1998. A field guide to reptiles and amphibians, eastern and central North America, third edition, expanded. Houghton Mifflin Company, Boston, USA. CONGDON, J. D., L. J. VITT, AND W. W. KING. 1974. Geckos: adaptive significance and energetics of tail autotomy. Science 184(4144):1379-1380. COOCH, E., AND G. WHITE, editors. 2006. Program Mark: A gentle Introduction, 5 th edition. On line at: http://www.phidot.org/software/mark/docs/book/ CORLEY, L. S., AND A. J. MOORE. 1999. Fitness of alternative modes of reproduction: developmental constraints and the evolutionary maintenance of sex. Proceedings of the Royal Society of London. B 266:471. CORMACK, R. M. 1964. Estimates of survival from the sighting of marked animals. Biometrika 51:429-438. D'AGOSTINO, R. B., W. CHASE, AND A. BELANGER. 1988. The appropriateness of some common procedures for testing the equality of two independent binomial populations. The American Statistician 42:(3):198-202. DIAMOND, J. M. 1975. Assembly of species communities. Pages 342. in M. L. Cody, J. M. Diamond, editors, Ecology and evolution of communities Belknap, Cambridge, Massachusetts, USA. DINSMORE, J. J. 1969. Reptiles of Little Tobago Island, West Indies. Quarterly Journal of the Florida Academy of Sciences 32(4):307-309. DIXON, J. R., AND SOINI, P. 1986. The reptiles of the upper Amazon basin, Iquitos region, Peru. Milwaukee Public Museum., Milwaukee WI, USA. DODD, C. K. JR. AND M. L. GRIFFEY. 2002. Remarks on the current status of the non-marine herpetofauna of Egmont Key, Florida. Florida Scientist 65(1):62-66. DOWNES, S. 2001. Trading heat and food for safety: costs of predator avoidance in a lizard. Ecology 82 (10):2870-2881. DU, W. 2006. Phenotypic plasticity in reproductive traits induced by food availability in a lacertid lizard, Takydromus septentrionalis. Oikos 112(2):363. EGAN, V. 2005. Hemidactylus mabouia (De Jonnes, 1818) tropical house gecko: mortality. African Herp News 38:19. ENGE, K. M., M. S. ROBSON, AND K. L. KRYSKO. 2004. Reptile surveys of pine rockland habitat in six Miami-Dade County parks. Florida Scientist 67(3):194-204. 210

PAGE 211

FGDL. 2000. Florida Geographic Data Library version 3.0. Http://www.fgdl.org/. FITCH, H. S. 1981. Sexual size differences in reptiles. University of Kansas Museum of Natural History Miscellaneous Publication 70:1-72. FLMNH. 2007. University of Florida, Florida Museum of Natural History, Division of Herpetology, online database: http://www.flmnh.ufl.edu. FOWLER, H. W. 1915. Cold-blooded vertebrates from Florida, the West Indies, Costa Rica, and eastern Brazil. Proceedings of The Academy of Natural Sciences of Philadelphia 67:244-269. FRANKENBERG, E. 1982. Social-Behavior of the parthenogenetic Indo-Pacific gecko, Hemidactylus garnotii. Zeitschrift fur Tierpsychologie 59(1):19-28. FROGNER, K. J. 1967. Some aspects of the interaction between the gecko species Hemidactylus frenatus and Lepidodactylus lugubris in Hawaii. Masters Thesis, University of Hawaii at Manoa, Manoa, Hawaii, USA. FUENMAYOR, G. R., G. UGETO, A. M. BAUER, T. BARROS, AND J. MANZANILLA. 2005. Expansion and natural history of a successful colonizing gecko in Venezuela (Reptilia: Gekkonidae: Hemidactylus mabouia) and the discovery of H. frenatus in Venezuela. Herpetological Review 36(2):121-125. GALDINO, C. A. B., AND M. VAN SLUYS. 2004. Tropidurus torquatus (collared lizard, calango) saurophagy. Herpetological Review 35(2):173. GALAN, P. 1997. Reproductive ecology of the lacertid lizard Podarcis bocagei. Ecography 20(2):197-209. GERBER, G. P., AND A. C. ECHTERNACHT. 2000. Evidence for asymmetrical intraguild predation between native and introduced Anolis lizards. Oecologia 124:599-607. GIBBONS, J. R. H. AND G. R. ZUG. 1987. Gehyra, Hemidactylus, and Nactus (Pacific geckos). eggs and hatchlings. Herpetological Review 18(2):35-36. GILL, B. J., D. BEJAKOVICH, AND A. H. WHITAKER. 2001. Records of foreign reptiles and amphibians accidentally imported to New Zealand. New Zealand Journal of Zoology 28(3):351-359. GOMEZ-ZLATAR, P., AND M. P. MOULTON. 2005. Habitat use by the nonindigenous Mediterranean gecko (Hemidactylus turcicus) in North Central Florida. Florida Scientist 68(3):206-214. -----, -----, AND R. FRANZ. 2006. Microhabitat use by introduced Hemidactylus turcicus (Mediterranean Geckos) in north-central Florida. Southeastern Naturalist 5(3):425-434. 211

PAGE 212

GRAMENTZ, D. 2000. Zur Biologie von Hemidactylus brookii angulatus in Ghana sowie H. mabouia und H. muriceus in Gabun. Salamandra 36(3):185-198. GRANT, C. 1932a. The Hemidactyls of the Porto Rico region. The Journal of the Department of Agriculture of Porto Rico 16:51-57. -----. 1932b. Herpetology of Tortola; notes on Anegada and Virgin Gorda, British Virgin Islands. Journal of the Department of Agriculture of Porto Rico 16:339-346. HAAGNER, G. V. 1997. Hemidactylus mabouia, tropical house gecko: predation. African Herp News (26):25. -----, AND W. R. BRANCH. 1996. Hemidactylus mabouia. African Herp News (25):43-44. HAUGE, J. B., AND B. P. BUTTERFIELD. 2000. Geographic distribution: Cosymbotus platyurus (Asian house gecko). Herpetological Review 31:52. HEAD, M. L., J. S. KEOGH, AND P. DOUGHTY. 2002. Experimental evidence of an agespecific shift in chemical detection of predators in a lizard. Journal of Chemical Ecology 28(3):541-554. HENDERSON, R. W., AND R. A. SAJDAK. 1996. Diets of West Indian racers (Colubridae: Alsophis): composition and biogeographic implications, pages 327-328 In: R. Powell and R. W. Henderson, editors, Contributions to West Indian herpetology: a tribute to Albert Schwartz. Society for the Study of Amphibians and Reptiles, Contributions to Herpetology Volume 12, Ithaca, NY, USA. HENKEL, F. W., AND W. SCHMIDT. 1995. Geckoes: biology, husbandry, and reproduction. Krieger Publishing Company, Malabar, FL, USA. HESTBECK, J. B., J. D. NICHOLS, AND R. A. MALECKI. 1991. Estimates of movement and site fidelity using mark-resight data of wintering Canada geese. Ecology 72:523-533. HILDEBRAND, B. 2001. Waypoint + v. 1.8.03. http://www.tapr.org/~kh2z/Waypoint/. HINES, J. E., AND D. I. MACKENZIE. 2006. Program Presence, version 2.05. http://www.mbr-pwrc.usgs.gov/software/doc/presence/presence.html. HOGE, A. R. 1950. Notas erpetologica. 7. Fauna erpetologica da Ihla da Queimada Grande. Memorias do Instituto Butantan 22:151-171. HOOGMOED, M. S. 1973. Notes on the herpetofauna of Surinam. IV. the lizards and amphisbaenians of Surinam. Biogeographica 4:1-419. The Hague: W. Junk. 212

PAGE 213

HOWARD, K. G., J. S. JR. PARMERLEE, AND R. POWELL. 2001. Natural history of the edificarian geckos Hemidactylus mabouia, Thecadactylus rapicauda, and Sphaerodactylus sputator on Anguilla. Caribbean Journal of Science 37(3-4):285-288. JENNESS, J. 2001. Random point generator (randpts.avx) extension for ArcView 3.x, v. 1.1. Jenness Enterprises. http://www.jennessent.com/arcview/random_points.htm. JOHNSON, C. 1960. Reproductive cycle in females of the greater earless lizard, Holbrookia texana. Copeia 1960(4):297-300. JOE, M., AND K. H. POLLOCK. 2002. Separation of survival and movement rates in multistate tag-return and capture-recapture models. Journal of Applied Statistics 29:373-384. JOLLY, G. M. 1965. Explicit estimates from capture-recapture data with both death and immigration stochastic model. Biometrika 52:225-247. KEARNEY, M., AND R. SHINE. 2004. Developmental success, stability, and plasticity in closely related parthenogenetic and sexual lizards (Heteronotia, Gekkonidae). Evolution 58(7):1560. KEREN-ROTEM, T., A. BOUSKILA, AND E. GEFFEN. 2006. Ontogenetic habitat shift and risk of cannibalism in the common chameleon (Chamaeleo chamaeleon). Behavioral Ecology and Sociobiology 59(6):723-731. KING, W. AND T. KRAKAUER. 1966. The exotic herpetofauna of southeast Florida. Quarterly Journal of the Florida Academy of Sciences 29(2):144-154. KLOWDEN, G. S. 2002. Hemidactylus mabouia (Amerafrican House Gecko). Herpetological Review 33(3):224. KLUGE, A. G. 1969. The evolution and geographical origin of the New World Hemidactylus mabouia-brookii complex (Gekkonidae, Sauria). Miscellaneous Publications Museum of Zoology, University of Michigan 138:1-78. -----, AND M. J. ECKARDT. 1969. Hemidactylus garnotii (Dumeril and Bibron), a triploid all-female species of Gekkonid lizard. Copeia (4):651-664. KHLER, G. 2003. Reptiles of Central America. Herpeton Verlag, Offenbach, Germany. KRYSKO, K. L., AND K. J. DANIELS. 2005. A key to the geckos (Sauria : Gekkonidae) of Florida. Caribbean Journal of Science 41(1):28-36. -----, K. M. ENGE, J. H. TOWNSEND, E. M. LANGAN, S. A. JOHNSON, AND T. S. CAMPBELL. 2005. New county records of amphibians and reptiles from Florida. Herpetological Review 36(1):85-87. 213

PAGE 214

-----, A. N. HOOPER, AND C. M. SHEEHY III. 2003a. The Madagascar giant day gecko, Phelsuma madagascariensis grandis Gray 1870 (Sauria: Gekkonidae): A new established species in Florida. Fla. Sci. 66:222-225. -----, C. M. I. SHEEHY, AND A. N. HOOPER. 2003b. Interspecific communal oviposition and reproduction of four species of lizards (Sauria: Gekkonidae) in the lower Florida Keys. Amphibia Reptilia 24(3):390-396. LAMB R. Y., AND R. B. WILLEY. 1979. Are parthenogenetic and related bisexual insects equal in fertility? Evolution. 33:774. LANDO, R. V., AND E. E. WILLIAMS. 1969. Notes on the herpetology of the U.S. naval base at Guantanamo Bay, Cuba. Studies on the Fauna of Curacao and Other Caribbean Islands 31(116):159-201. LAWSON, R., P.G. FRANK, AND D. L. MARTIN. 1991. A gecko new to the United States herpetofauna, with notes on geckoes of the Florida keys. Herpetological Review 22(1):11-12. LEBRETON, J. D., K. P. BURNHAM, J. CLOBERT, AND D. R. ANDERSON. 1992. Modeling survival and testing biological hypotheses using marked animals a unified approach with case-studies. Ecological Monographs 62(1):67-118. LEE COUNTY PROPERTY APPRAISER. 2007. Web page: http://www.leepa.org. LIN, J. Y., AND H. Y. CHENG. 1984. Ovarian cycle in the house gecko, Hemidactylus frenatus, in Taiwan with reference to food stress in winter. Bulletin of the Institute of Zoology Academia Sinica 23(1):21-28. LOCEY, K. J., AND P. A. STONE. 2006. Factors affecting range expansion in the introduced Mediterranean Gecko, Hemidactylus turcicus. Journal of Herpetology 40(4):526-530. LOVERIDGE, A. 1941. Some geckos of Tanganyika Territory. Tanganyika Notes and Records 12:32-37. -----. 1947. Revision of the African lizards of the family Gekkonidae. Bulletin of the Museum of Comparative Zoology at Harvard College 98(1):1-474. MACKENZIE, D. I., AND L. L. BAILEY. 2004. Assessing the fit of site occupancy models. Journal of Agricultural, Biological, and Environmental Statistics 9(3):300-318. -----, -----, AND J. D. NICHOLS. 2004. Investigating species co-occurrence patterns when species are detected imperfectly. Journal of Animal Ecology 73:546-555. -----, J. D. NICHOLS, J. E. HINES, M. G. KNUTSON, AND A. B. FRANKLIN. 2003. Estimating site occupancy, colonization, and local extinction when a species is detected imperfectly. Ecology 84(8):2200-2207. 214

PAGE 215

-----, -----, G. B. LACHMAN, S. DROEGE, A. ROYLE, AND C. LANGTIMM. 2002. Estimating site occupancy rates when detection probabilities are less than one. Ecology 83(8):2248-2255. -----, -----, J. A. ROYLE, K. H. POLLOCK, L. L. BAILEY, AND J. E. HINES. 2006. Occupancy estimation and modeling: inferring patterns and dynamics of species occurrence. Academic Press, Amsterdam. 324 pp. MASCHIO, G. F., M. DI-BERNARDO, AND J. MELCHIORS. 2004. Oxyrhopus rhombifer rhombifer (NCN). diet. Herpetological Review 35(1):71. MCCOY, C. J. 1972. Hemidactylus garnotii. U.S.A., Florida, Lee Co., Sanibel Island. Herpetological Review 4(1):23. MESHAKA, W. E., JR. 1994. Reproductive cycle of the Indo-pacific gecko, Hemidactylus garnotii, in south Florida. Florida Scientist 57(1-2):6-9. -----. 1995. Reproductive cycle and colonization ability of the Mediterranean gecko (Hemidactylus turcicus) in south-central Florida. Florida Scientist 58(1):10-15. -----. 1996. Vagility and the Florida Distribution of the Cuban treefrog (Osteopilus septentrionalis). Herpetological Review 27(1):37-40. -----. 1999. The herpetofauna of the Doc Thomas House in south Miami, Florida. Florida Field Naturalist 27(3):121-123. -----. 2000. Colonization dynamics of two exotic geckos (Hemidactylus garnotii and H. mabouia) in Everglades National Park. Journal of Herpetology 34(1):163-168. -----. 2001. The Cuban treefrog in Florida. life history of a successful colonizing species. University Press of Florida, Gainesville, FL, USA. -----, B. P. BUTTERFIELD, AND J. B. HAUGE. 1994a. Reproductive notes on the introduced gecko Hemidactylus mabouia in southern Florida. Herpetological Natural History 2(1):109-110. -----, -----, AND -----. 1994b. Hemidactylus mabouia as an established member of the Florida herpetofauna. Herpetological Review 25(2):80-81. -----, -----, AND -----. 1994c. Hemidactylus frenatus established on the lower Florida Keys. Herpetological Review 25(3):127-128. -----, -----, AND -----. 2004. The exotic amphibians and reptiles of Florida. Krieger Publishing Company, Malabar, FL, USA. -----, AND J. LEWIS. 1994. Cosymbotus platyurus in Florida: ten years of stasis. Herpetological Review 25(3):127. 215

PAGE 216

-----, AND B. A. MOODY. 1996. The Old World tropical house gecko (Hemidactylus mabouia) on the Dry Tortugas. Florida-Scientist 59(2):115-117. -----, H. T. SMITH, R. SEVERSON, AND M. A. SEVERSON. 2005. Spatial picture of a gecko assemblage in flux. Florida Scientist 68(1):53-55. MSN 2007. Microsoft Network, http://weather.msn.com. MURPHY, J. C. 1997. Amphibians and reptiles of Trinidad and Tobago. Krieger Publishing Company, Malabar, Florida, USA. MYERS, G. S. 1945. A natural habitat of the house gecko (Hemidactylus mabouia) in Brazil. Copeia (2):120. MYERS, S. 1978. Hemidactylus garnotii (Indo-Pacific gecko). USA. Florida: Martin Co: Hobe Sound. Herpetological Review 9(3):107. NICHOLS, J. D., J. E. HINES, K. H. POLLOCK, R. L. HINZ, AND W. A. LINK. 1994. Estimating breeding proportions and testing hypotheses about costs of reproduction with capture-recapture data. Ecology 75(7):2052-2065. -----, J. R. SAUER, K. H. POLLOCK, AND J. B. HESTBECK. 1992. Estimating transition probabilities for stage-based population projection matricies using capture-recapture data. Ecology 73(1):306-312. OLIVER, J. A. AND C. E. SHAW. 1953. The amphibians and reptiles of the Hawaiian Islands. Zoologica 38(5):65-95. OWEN, J., AND G. PERRY. 2005. Anolis cristatellus wileyae (Virgin Island crested anole). saurophagy. Herpetological Review 36(4):444. PAULISSEN, M. A., AND H. A. MEYER. 2000. The effect of toe-clipping on the gecko Hemidactylus turcicus. Journal of Herpetology 34(2):282-285. PENDLEBURY, G. B. 1972. Nesting sites, eggs and young of Hemidactylus mabouia from Carriacou, West Indies. Herpetological Review 4(6):203. PETREN, K., D. T. BOLGER, AND T. J. CASE. 1993. Mechanisms in the competitive success of an invading sexual gecko over an asexual native. Science 259(5093):354-358. -----, AND T. J. CASE. 1996. An experimental demonstration of exploitation competition in an ongoing invasion. Ecology 77(1):118-132. POLLOCK, K. H., J. D. NICHOLS, C. BROWNIE, AND J. E. HINES. 1990. Statistical inference for capture-recapture experiments. Wildlife Monographs 107:1-97. 216

PAGE 217

POWELL, R., R. I. CROMBIE, AND H. E. A. BOOS. 1998. Hemidactylus mabouia. Catalogue of American amphibians and reptiles 674:1-11. PRADEL R., C. M. A. WINTREBERT, AND O. GIMENEZ. 2003. A proposal for a goodnessof-fit test to the Arnason-Schwarz multisite capture-recapture model. Biometrics 59(1):43-53 -----, O. GIMENEZ, AND J. D. LEBRETON. 2005. Principles and interest of GOF tests for multistate capture-recapture models. Animal Biodiversity and Conservation 28(2):189204. PUNZO, F. 2001. The mediterranean gecko, Hemidactylus turcicus: life in an urban landscape. Florida Scientist 64(1):56-66. -----. 2005. The introduction of Hemidactyline geckos in Florida: colonization and impact on native fauna, In: W. E. Meshaka, and K. Babbitt, editors, Amphibians and Reptiles: Status and Conservation in Florida, Krieger Publishing Company Melbourne, FL, USA RAND, A. S. 1954. Variation and predator pressure in an island and a mainland population of lizards. Copeia 1954(4):260-262. RICE, W. R. 1988. A new probability model for determining exact p-values for 2 x 2 contingency tables when comparing binomial proportions. Biometrics 44(1):1-22. RIVERO, J. A. 1978. Los anfibios y reptiles de Puerto Rico. the amphibians and reptiles of Puerto Rico. Editorial Universitaria. Universidad de Puerto Rico. Mayaguez, Puerto Rico. RODRIGUES, M. T. 1996. Lizards, snakes, and amphisbaenians from the Quaternary sand dunes of the Middle Rio So Francisco, Bahia, Brazil. Journal of Herpetology 30(4):513-523. ROSE, F. L., AND C. D. BARBOUR. 1968. Ecology and reproductive cycles of the introduced gecko, Hemidactylus turcicus, in the southwestern United States. American Midland Naturalist 79(1):159-168. SAVAGE, J. M. 2002. The amphibians and reptiles of Costa Rica: A herpetofauna between two continents, between two seas. Illinois: University of Chicago Press. SAZIMA, I., AND A. J. S. ARGOLO. 1994. Siphlophis pulcher (NCN). prey. Herpetological Review 25(4):126. SCHLEICH, H. H., W. KSTLE, AND K. KABISCH. 1996. Amphibians and Reptiles of North Africa: biology, systematics, field guide. Koeltz Sceintific Books, Koenigstein, Germany. SCHOENER, T. W. 1979. Inferring the properties of predation and other injury-producing agents from injury frequencies. Ecology 60(6):1110-1115. 217

PAGE 218

SCHWARTZ, A., AND HENDERSON, R. W. 1991. Amphibians and reptiles of the West Indies: descriptions, distributions, and natural history. University of Florida Press., Gainesville, FL. SCHWARZ, C. J., J. F. SCHWEIGERT, AND A. N. ARNASON. 1993. Estimating migration rates using tag recovery data. Biometrics 49:177-193. SEBER, G. A. F. 1965. A note on the multiple recapture census. Biometrika 52:249-259. -----. 1982. The estimation of animal abundance and related parameters. Second edition. Griffin, London, England SELCER, K. W. 1986. Life history of a successful colonizer the Mediterranean gecko, Hemidactylus turcicus, in Southern Texas. Copeia 1986(4):956-962. SHANBHAG, B. A., L. SUBRAYA, S. K. SAIDAPUR. 1998. Pattern of recruitment, growth of developing follicles, and germinal bed activity in the tropical gecko, Hemidactylus brooki. Journal of Herpetology 32(4):566-572. SIMO, M., C. PRIGIONI, J. LANGONE, AND F. ACHAVAL. 1988. Introduccion accidental de fauna al Uruguay en cargamentos de banana, Musa acumintaI (Musoidae). Boletin de la Sociedad Zoologica del Uraguay (2o poca) 4:15-18. STAMPS, J. A. 1983. The relationship between ontogenetic habitat shifts, competition and predator avoidance in a juvenile lizard (Anolis aeneus). Behavioral Ecology and Sociobiology 12(1):19-33. STEINER, T. M. 1983. Hemidactylus garnotii (indo-pacific gecko). Predation. Herpetological Review 14(3):74. STEJNEGER, L. 1922. Two geckos new to the fauna of the United States. Copeia 108(8):56. TOWNSEND, J. H., AND K. L. KRYSKO. 2003. The distribution of Hemidactylus (Sauria: Gekkonidae) in northern peninsular Florida. Florida Scientist 66(3):204-208. -----, -----, A. T. REPPAS, AND C. M. I. SHEEHY. 2002. Noteworthy records for introduced reptiles and amphibians from Florida, USA. Herpetological Review 33(1):75. TROUT, L., AND T. D. SCHWANER. 1994. Allozyme evidence for insularity in exotic populations of the Mediterranean Gecko (Hemidactylus turcicus). Journal of Herpetology 28:391. UNITED STATES CENSUS BUREAU. 2000. http://www.census.gov. UNITED STATES CENSUS BUREAU. 2006. http://www.census.gov. 218

PAGE 219

VITT, L. J. 1986. Reproductive tactics of sympatric Gekkonid lizards with a comment on the evolutionary and ecological consequences of invariant clutch size. Copeia 1986(3):773786. -----. 1992. Diversity of reproductive strategies among Brazilian lizards and snakes: the significance of lineage and adaption. In: W. C. Hamlett, editor, Reproductive Biology of South American Vertebrates. pp. 135-149. -----. 1995. The ecology of tropical lizards in the Caatinga of northeast Brazil. Occasional Papers of the Oklahoma Museum of Natural History 1:1-29. -----, AND P. A. ZANI. 1997. Ecology of the nocturnal lizard Thecodactylus rapicauda (Sauria: Gekkonidae) in the lower Amazon. Herpetologica 53(2):165-179. VOSS, R. 1975. Notes on the introduced gecko, Hemidactylus garnotii, in south Florida. Florida Scientist 38(3):174. WEATHER UNDERGROUND. 2007. http://www.wunderground.com. WHITE, G.C., AND K. P. BURNHAM. 1999. Program MARK: Survival estimation from populations of marked animals. Bird Study 46 Supplement:120-138. -----, -----, AND D. R. ANDERSON. 2001. Advanced features of Program Mark, p. 368-377, In: R. Field, R. J. Warren, H. Okarma, and P. R. Sievert, editors, Wildlife, land, and people: priorities for the 21st century. The Wildlife Society, Bethesda, MD. WILLIAMS, B. K., J. D. NICHOLS, AND M. J. CONROY. 2002. Analysis and management of animal populations: modeling, estimation, and decision making. Academic Press, San Diego. 817 pp. WILSON, E. B. 1927. Probable inference, the law of succession, and statistical inference. Journal of the American Statistical Association 22(158):209-212. WILSON, L. D., AND L. PORRAS. 1983. The ecological impact of man on south Florida herpetofauna. University of Kansas Museum of Natural History Special Publication 9:1-89. WOJNOWSKI, D., AND E. SELEMPO. 2005. Hemidactylus mabouia (tropical house gecko). predation. Herpetological Review 36(2):178-179. WOOD, K. V., J. D. NICHOLS, H. F. PERCIVAL, AND J. E. HINES. 1998. Size-sex variation in survival rates and abundance of pig frogs, Rana grylio, in Northern Florida wetlands. Journal of Herpetology 32(4):527-535. ZAMPROGNO, C., AND R. L. TEIXEIRA. 1998. Habitos alimentares da lagartixa de parede Hemidactylus mabouia (Reptilia, Gekkonidae) da planicie litoranea do norte do Espirito Santo, Brasil. Revista Brasileira de Biologia 58(1):143-150. 219

PAGE 220

220 BIOGRAPHICAL SKETCH Gregg Klowden was raised in the Chicago s uburb of Highland Park. In addition to his parents and two sisters, his house was filled with all sorts of furry, fuzzy, feathery, and slithering influences. He spent many camp filled summ ers riding horses in Wisconsin from where counselor Tom Harkin encouraged him to go on Mi nnesota Boundary Waters canoe trips. It was here and on many family camping trips that Gregg developed his zest for the outdoors that eventually carried him to the mountains of Boulder, Colorado for college. It was a non-major biology course his freshman year that turned his attention towards a career in science. Combined with his love of animals he decided life as a ve terinarian was for him. He returned to his home state for his senior year at the University of Illinois where he also a ttended veterinary school. After one year he decided to sw itch directions, but what to do? To help him think, he hit the road for a couple months in Europe which led him to Egypt, Guatemala, Belize, Honduras and then to Alaska for a 4 month stint working on a fi sh processing ship. Soon after, he entered the University of Florida where he completed his Ma sters of Science degree in the Department of Wildlife Ecology and Conservati on studying boa constrictors, Boa constrictor and rainbow boas, Epicrates cenchria in the forests of Peru and in museums of the United States. In Gainesville he befriended Cathy Olson with whom he shared many walks and talks and now his life. Cathy and Gregg were married in 2003 at one of their favor ite north Florida destinations, the Spirit of the Suwannee Music Park. They reside in southwest Florida in Port Charlotte with their eight furry, fuzzy, and f eathery children, Darby, Grover, Gus, Bernie, Zydeco, B2G2, Pedro, and Lola. Gregg plans to share his love of all things wild teaching college.