xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001992355
007 cr mnu|||uuuuu
008 090317s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002392
Population densities of the Cuban treefrog, Osteopilus septentrionalis and three native species of Hyla (Hylidae), in urban and natural habitats of southwest Florida
h [electronic resource] /
by Teresa Piacenza.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 39 pages.
Thesis (M.S.)--University of South Florida, 2008.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: The Cuban treefrog, Osteopilus septentrionalis, is an invasive species in Florida that may be negatively impacting adult and larval native treefrog species through competition via direct predation. The purpose of this study was to determine the abundance and distribution of O. septentrionalis in urban, semi-urban and natural habitats. The distribution and abundance of native treefrogs within the study area, Hyla cinerea, Hyla femoralis and Hyla squirella, were also estimated and compared to that of O. septentrionalis. Treefrogs were captured using PVC pipes with two internal diameters, 1.9 cm and 4.45 cm, hung on three tree types, Pinus, Quercus and Sabal. Distance to water, time of year and monthly rainfall were also considered as a potential influence of treefrog captures. O. septentrionalis was found in sites with both urban and natural habitats in frequencies far greater than those of native treefrogs.Results indicate that O. septentrionalis are captured more frequently in sites with urban habitat; however recapture rates and estimates of population size indicate that the population sizes of O. septentrionalis may actually be larger at sites with natural habitat. O. septentrionalis are found significantly more frequently in Quercus. H. femoralis and H. squirella were found significantly more frequently in PVC pipes with an internal diameter of 4.45 cm. A weak, although significant, positive correlation was found between distance to water and treefrog abundances for O. septentrionalis and H. squirella. Native treefrogs were found less frequently in PVC pipes that also contained O. septentrionalis; this result is perhaps the most interesting because it may indicate that native treefrogs will avoid habitat or refuges where O. septentrionalis are present. If O. septentrionalis is a real threat to species of treefrogs in Florida, these results should be of concern.Findings suggest that PVC pipes may give biased estimates of treefrog densities. Available refuge, temperature and precipitation all appear to have an effect on how attractive PVC pipes are to treefrogs.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Gordon Fox, Ph.D.
t USF Electronic Theses and Dissertations.
Population Densities of the Cuban Treefrog, Osteopilus septentrionalis and Three Native Species of Hyla (Hylidae) in Urban and Natural Habita ts of Southwest Florida by Teresa Piacenza A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Biology College of Arts and Sciences University of South Florida Major Professor: Gordon Fox, Ph.D. Gary Huxel, PhD. Henry Mushinsky, PhD. Peter Stiling, Ph D. Date of Approval: April 2, 2008 Keywords: H. cinerea, H. femoralis, H. squirella invasive, PVC Copyright 2008, Teresa Piacenza
Acknowledgements My experience as a graduate student in the Biology department at USF has been one of the most enjoyable times in my life. I have many people to th ank for this. I would sincerely like to thank my advisor, Gor don Fox, for all of his patience and guidance throughout the construction of this thesis a nd for never turning me away the countless times I popped into your office for help. Thank you also to my committee, Gary Huxel, Henry Mushinksy and Peter Stiling. Special th ank you to Gary for bringing me to Baja, Mexico, my first and favorite field site. I w ould also like to thank the Biology department for providing a teaching assistantship throughout my time as a graduate student. This support has allowed me to conduct research a nd successfully complete my degree. Thank you to my lab mates and my fellow graduate st udents who have made my time spent here thought-provoking and a ton of f un. Thank you to my first an d most wonderful lab mate Celina for your friendship and advice about biology and everythi ng else. Thank you to my family and my friends for the love and the support that you ha ve given me in all forms. A very special thank you to Matt fo r the endless good advice you give and for the good example you set. Thanks for always motiv ating me to do my best. Thank you to my dog Charlie whose unconditional love and companionship has given me an incredible amount of comfort throughout stressful times, unbeknownst to him.
i Table of Contents List of Tables ii List of Figures iii Abstract iv Introduction 1 Methods 7 Study Sites 7 Trapping Methods 10 Statistical Analysis 12 Results 14 Treefrog Density Estimates 14 Correlation of Natives and Osteopilus septentrionalis 18 Effects of Site, Tree Type and PV C Diameter on Treefrog Captures 19 Distance to Water Correlation 21 Seasonal and Precipitation Eff ects on Treefrog Captures 22 Discussion 2 7 Treefrog Density Estimates 27 Correlation of Natives and Osteopilus septentrionalis 30 Effects of Site, Tree Type and PV C Diameter on Treefrog Captures 30 Distance to Water Correlation 31 Seasonal and Precipitation Eff ects on Treefrog Captures 31 Invasion of Florida by Osteopilus septentrionalis 32 Conclusion 3 5 References 36
ii List of Tables Table 1 Area, tree density and land-use de scriptions of study sites 9 Table 2 Number of trees chosen at each site 11 Table 3 Total captures, recaptures and % recaptures of tr eefrogs 14 Table 4 Vertebrates found in PVC pi pes during study. 15 Table 5 Correlation of O. septentrionalis and native treefrogs 19 Table 6 Site, tree type and PVC pipe eff ects on treefrog captures 20 Table 7 Number of treefrog captures in PVC pipes with internal diameters of 1.91cm and 4.45 cm. 21 Table 8 Correlation of treefrog captures and distance to water 21 Table 9 Correlation of total captures and monthly precipitation 24 Table 10 Number of treefrog captures at diffe rent times of the year 25
iii List of Figures Figure 1 Osteopilus septentrionalis 3 Figure 2 Current distribution of O. septentrionalis in Florida 4 Figure 3 Aerial map of sites used in study 8 Figure 4 Tree densities of each study site 10 Figure 5 Total captures and recaptures of H. cinerea, H. femoralis, H. squirella and O. septentrionalis 16 Figure 6 Recapture rates of H. cinerea, H. femoralis, H. squirella and O. septentrionalis 17 Figure 7 Population size estimatio n using Schnabel method for H. cinerea, H. femoralis, H. squirella and O. septentrionalis 18 Figure 8 Correlation of treefrog ca ptures and distance to water for H. cinerea, H. femoralis, H. squirella and O. septentrionalis. 22 Figure 9 Monthly precipitation and total tree frog captures 23 Figure 10 Total treefrog captures versus m onthly precipitation 24 Figure 11 The deviation of observe d treefrog captures from the expected treefrog captures at differe nt times of year 26 Figure 12 Dispersal of non-native he rpetofauna, endemic to islands, in Florida counties 34
iv Population Densities of the Cuban Treefrog, Osteopilus septentrionalis and Three Native Species of Hyla (Hylidae) in Urban and Natural Habita ts of Southwest Florida Teresa Piacenza ABSTRACT The Cuban treefrog, Osteopilus septentrionalis is an invasive species in Florida that may be negatively impacting adult a nd larval native treefrog species through competition via direct predation. The purpos e of this study was to determine the abundance and distribution of O. septentrionalis in urban, semi-urban and natural habitats. The distribu tion and abundance of native tr eefrogs within the study area, Hyla cinerea, Hyla femoralis and Hyla squirella, were also estimated and compared to that of O. septentrionalis. Treefrogs were captured using PVC pi pes with two internal diameters, 1.9 cm and 4.45 cm, hung on three tree types, Pinus, Quercus and Sabal. Distance to water, time of year and monthly rainfall were also considered as a potential influence of treefrog captures. O. septentrionalis was found in sites with both urban and natural habitats in frequencies far greater than thos e of native treefrogs. Results indicate that O. septentrionalis are captured more frequently in sites with urban habitat; however recapture rates and estimates of population size indicate that the population sizes of O. septentrionalis may actually be larger at sites with natural habitat. O. septentrionalis are found significantly more frequently in Quercus H. femoralis and H. squirella were found significantly more frequently in PVC pipes with an internal diameter of 4.45 cm. A weak, although significant, positive correlation was found between distance to water and
v treefrog abundances for O. septentrionalis and H. squirella. Native treefrogs were found less frequently in PVC pipes that also contained O. septentrionalis; this result is perhaps the most interesting because it may indicate that native treefrogs will avoid habitat or refuges where O. septentrionalis are present. If O. septentrionalis is a real threat to species of treefrogs in Florida, these result s should be of concern. Findings suggest that PVC pipes may give biased estimates of treef rog densities. Availabl e refuge, temperature and precipitation all appear to have an effect on how attractive PVC pipes are to treefrogs.
1 Introduction The invasion of natural communities by exotic species is a major threat to biodiversity, second only to habitat destruc tion (Williamson, 1999). This threat may arise through predation, competition for resources disruption of trophic dynamics, habitat modification, hybridization with natives or as vectors for disease (Simberloff, 1997; Greenless, 2007). Habitats which are suscepti ble to invasion by exotics include those which have been disturbed by urban development, habitats with a tropical climate or island habitats. Florida possesses all of these ch aracteristics, and therefore is particularly vulnerable to invasions; the southe rn peninsula of Florida is e ffectively an island as it is surrounded on the eastern, western and southern sides by water and on the northern side by frost. Additionally, habitats of Florida ar e favorable targets for both terrestrial and aquatic invaders because of the abundance of wetland habitats. The introduction of nonindigenous species is facilitate d by shipping ports and airports in the cities of Miami and Tampa (Simberloff, 1999). The Cuban treefrog, Osteopilus septentrionalis Figure 1, is an i nvasive species in Florida which may be negatively influencing populations of native treefrogs. The native range of O. septentrionalis includes Cuba, Isla de Juvent ud, the Cayman islands and the Bahamas (Meshaka, 2001). Other than Florida, the exotic range of O. septentrionalis includes Puerto Rico, St. Croix, St. Thomas, St. Maarten (Powell et al., 1992; Meshaka, 2001), Anguilla (Townsend et al. 2000; Mesh aka, 2001), Necker Island (Meshaka, 2001),
2 Antigua (Lindsay and Cooper 2008) and Curacao (Van Buurt, 2007). The colonization of Key West, Florida by O. septentrionalis was first reported by Barbour in 1931. According to Barbour (1931) O. septentrionalis had been present on Key West long before his report and was most likely introdu ced to the island via passenger boats and ferries carrying shipments of Cuban produ ce. In 1952 there was a report of several O. septentrionalis on mainland Florida in Miami (Schwartz, 1952). O. septentrionalis was first reported in Hillsborough county, wher e the present study takes place, in 1996 (Meshaka, 1996). Oliver (1950) re ported that the Cuban anole, Anolis sagrei, was most likely introduced to Tampa via shipments of lumber and vegetable produce imported from Cuba; it is likely that O. septentrionalis were also introduced in this way. Presently the distribution of O. septentrionalis reaches Duval County as is shown in Figure 2. It is believed that O. septentrionalis will continue to disperse alo ng the Gulf coast of Florida into adjacent coastal states and eventually into Mexico; breeding populations may begin to colonize Georgia, however cool winter temper atures will limit dispersal further along the East coast of the United States (M eshaka, 2001; Soma, 2007). Meshaka (2001) predicts that O. septentrionalis will continue to disperse th roughout the Caribbean as well as Central and South America.
3 Figure 1. Osteopilus septentrionalis. Photo by Teresa Piacenza.
4 090180 45Miles Counties containing O. septentrionalis Counties not containi ng O. septentrionalisAuthor: Teresa PiacenzaFigure 2. Current distribution of O. septentrionalis in Florida. The successful colonization of Florida by O. septentrionalis can be attributed to several characteristics. Members of O. septentrionalis have high fecundity because of large clutch sizes, short larval periods and th e ability to reach sexua l maturity in a short period of time. Additionally O. septentrionalis are capable of breeding year round in counties including and south of Hillsbo rough (Meshaka, 2001). Known indigenous Florida predators of O. septentrionalis include Thamnophis sauritus, Thamnophis sirtalis, Elaphe obsolete, Coluber constrictor (Love, 1995; Meshaka and Ferster, 1995; Meshaka, 1997) and Corvus brachyrhynchos (Butterfield et al., 1997). However, a noxious skin secretion produced by O. septentrionalis may potentially be making them unpalatable to
5 many other potential predators, which may contribute to successful colonization. O. septentrionalis also have the ability to tolerate a ra nge of physical c onditions; Meshaka (2001) reports that larvae are able to surv ive in breeding waters with temperatures varying from 12-41C and adults will remain active at night during periods of low temperatures. Finally, little adaptation was required of O. septentrionalis while invading because of the striking similarities of habita t in their native and nove l ranges (Butterfield et al., 1997; Meshaka, 2001) O. septentrionalis primarily inhabits mesophytic forests and also mangrove forests (Meshaka, 2001). It is difficult to assess the threat that O. septentrionalis poses to populations of native Florida treefrogs. O. septentrionalis have been shown to compete with native anurans through direct adult (Meshaka, 2001; Wyatt, 2004) a nd larval predation (Smith, 2005a; Smith 2005b; Smith 2006). Dietary and habitat preferences of O. septentrionalis overlap with the preferences of native treef rogs, however these resources are not limited and therefore competition is of little consequence. Although it has been shown that O. septentrionalis can have a negative impact on individua ls of native treefrogs, currently, it is unknown whether the presence of this invader is affecting populations of natives. The purpose of this study is to de termine how population densities of O. septentrionalis vary in natural and urban habitats using PVC refugia. It has been reported that O. septentrionalis will readily inhabit urban areas (Meshaka 2001). However, before now, an intense sampling effort in urban areas has not been performed. I was interested to know if O. septentrionalis would be found in similar dens ities in natural and urban habitats and also if these de nsities coincide with the de nsities of native treefrogs. Additionally I hope to determine if internal PVC pipe diamet er, tree type or distance to
6 water will influence treefrog capture rates. PVC refugia have been shown to be an effective method for capturing treefrogs (B oughon and Staiger, 2001; Zacharow et al. 2003; Bartareau 2004). This trap method is partic ularly useful in Fl orida; treefrogs are able to escape from poles if temperatures get too high or the environment becomes too dry, therefore eliminating trap mortality.
7 Methods Study Sites Nine sites including natural, urban and semi -urban habitats were chosen for this study, Figure 3. I chose the USF Tampa campus as the location for my three urban sites. The plant communities at these sites have been altered by urban development and are surrounded on all sides by roads or parking lots. Trout Creek Park, part of the Hillsborough County owned Wilderness Park, was used as a natural site. Trout Creek is composed of mesic and scrubby flatwood plan t communities. The University of South Florida Ecological Research Area (USF Eco Area) was used as the location for three additional natural sites which are also co mposed of mesic and scrubby flatwood plant communities. The USF Eco Area is located on the Hillsborough River and is north and northeast of the USF Tampa Campus. The rema ining two sites chosen for this study are located at the USF golf course and the USF Riverfront Park. The plant communities at these sites have been also altered by urba n development, however, they are bordered by natural habitat. The USF golf course is bor dered by a wetland. The USF Riverfront Park is bordered by a wetland on one side and by hardwood forest on the other. For the purpose of this study, these sites will be considered semi-urban. To further characterize each of the nine sites chosen, I estimated tree density using the T-squared sampling procedure (Besag 1973, Krebs 1999). Additionally, the land use descriptions for each site
8 were obtained from the Federal Geographic Da ta Committee. Table 1 gives the area, tree density and land use description for e ach of the nine sites. o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o oo o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o USF West USF Central USF East Golf course Eco West Eco Central Eco East Riverfront Park Trout Creek 0 0.7 1.4 0.35 Miles o Trees used in study USF Tampa campusFigure 3. Aerial map of sites used in study.
9 site area (km2) trees / meters2land use description Eco Central 0.099 0.118 Upland forests, hardwood conifer mixed Eco East 0.015 0.042 Upland forests, hardwood conifer mixed Eco West 0.015 0.072 Upland forests, hardwood conifer mixed. Wetland hardwood forest, steam and lake swamps (bottom land). Golfcourse 0.414 0.003 Urban and built-up, recreational Riverfront Park 0.052 0.005 Urban and built-up, institutional Trout Creek 0.033 0.074 Upland forests, hardwood conifer mixed. Wetland hardwood forest, steam and lake swamps (bottom land). USF Central 0.046 0.005 Urban and built-up, institutional USF East 0.025 0.020 Urban and built-up, institutional USF West 0.107 0.003 Urban and built-up, institutional Table 1. Area, tree density and landuse descriptions of study sites.
10 Sites E c o Cent ra l E co E a s t Eco West Trout Cre e k Go lfc ou rse Riverf r ont U SF Central US F E ast US F West Trees per m 2 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Figure 4. Tree densities of each study site. Trapping Methods To attract treefrogs I hu ng PVC pipes on trees over an aluminum nail which was placed approximately 2 meters above the ground. Two 76cm long poles with different internal diameters, 1.9cm and 4.45cm, were hung on each tree. I capped the bottom of each pipe so that they would hold water after a rain. I drilled two holes in each pipe, one 3 cm from the top to fit over the nail and one 8cm from the bottom so that the entire pipe wouldnÂ’t fill with water. At each site I chose 20-21 trees, either Pinus, Quercus or Sabal. I determined the latitude and longitude of each tree using Tr imble, GeoExplorer CE, GPS equipment. The numbers of each tree type used at eac h site are listed below in Table 2.
11 Tree type Quercus Sabal Pinus Eco Central 10 0 10 Eco East 7 6 7 Eco West 7 7 7 Golfcourse 7 7 7 Riverfront Park 7 7 7 Trout Creek 7 7 7 USF Central 19 1 0 USF East 8 4 8 USF West 19 0 1 Table 2. Number of trees chosen at each site. After installing the PVC pipes at each site I waited a minimum of 2 weeks before beginning data collection in or der to give the treefrogs e nough time to find the pipes. I collected data from October 2006 through September 2007. I checked pipes twice a month, usually every other week. There were some months that pipes were only visited once; because of this sampling error I used the average number of treefrogs caught each month in my data analysis. I removed treefrogs by gently shaking them out of the pipes. All vertebrates found inside the pipes were identif ied by species and recorded. A ll new treefrog captures were marked using toe clips. A single toe, speci fically the back right outermost toe, was clipped to distinguish recapt ures from new captures.
12 Statistical Analysis To compare the number of treefrogs capture d at each site, on different tree types and in PVC pipes with different internal diam eters, I ran a 3 way analysis of variance (ANOVA), using a type IV sums of squares (SS), for each species. A type IV SS was used to account for empty cells, i.e. the three genera of trees used in this study were not present at all nine sites. Be fore running the analysis I tran sformed the data by taking the square root of total captures. I chose this analysis in order to test for a significant difference in the number of treefrogs at each site, for each tree type, in PVC pipes with small or large internal diameters, and also for site by tree type in teractions, site by PVC diameter interactions, tree type by PVC diameter interactions and fina lly site by tree type by PVC diameter interactions. Sites with less than two treefrog captures were not included in the analysis. I estimated the population size of each tr eefrog species at ea ch site using the Schnabel method (Schnabel 1938, Krebs 1999). In order to determine whether the population size of each species of treefrogs di ffered significantly at each site I used the 2 goodness of fit test, using the mean frequency as my expected value. Sites where zero treefrogs were captured were not included in the analysis. To determine if O. septentrionalis are found more frequently in PVC pipes that also contain native treefrogs I ran a Pearson correlation test for each sites. The Pearson correlation test will indicate if there is any correlation between the number of O. septentrionalis and the number of native treefrogs found in a given PVC pipe on a given day. Specifically, a positive correlation will indicate that O. septentrionalis are found
13 more often in PVC pipes where native treefr ogs can also be found. Only PVC pipes that contain at least one treef rog on a given day were used for the analysis. To test for a correlation between treefr og abundance and distance to water I used the Pearson correlation test for each species of treefrog. Distance to water was calculated by marking each PVC pipe using Trimble, GeoE xplorer CE, GPS equipment. I converted these data into a vector point feature in ArcMap, ArcGIS9. I obt ained a hydrologic map of Hillsborough County from the Florida Geogr aphic Data Library and overlaid this feature on to my GPS data feature. Small ponds and ditches that were not shown on the hydrologic map were selected in ArcMap, Ar cGIS and added to the data file. Using ArcMap, ArcGIS9 I spatially joined the 2 featur es in order to calcul ate the distance from each PVC pipe to the nearest body of water. To test for a relationship between capture s of each species of treefrog and the time of year the capture took place, I created a contingency tabl e and performed a G-test. The G-test is similar to a 2 test, however it is less sensitive to sample size. This test will show if different species of treefr ogs are behaving similarly in different times of the year. I correlated monthly treefrog captures w ith monthly rainfall measurements, taken from the National Climatic Data Center with the Pearson correlation test.
14 Results Treefrog Density Estimates During this study I captured 2,086 treefrogs including 1,305 recaptures. O. septentrionalis were captured the most fre quently (n = 1,449) followed by H. squirella (n = 424), H. cinerea (n = 160) and H. femoralis (n = 53). The modified numbers of total captures, recaptures and % re captures are shown in Table 3. In addition to treefrogs, seven other types of vertebrates were also f ound in PVC pipes. These are listed in Table 4. H. cinerea H. femoralis H. squirella O. septentrionalis total recap % recap total recap% recap total recap% recap total recap% recap Eco Central 0 0 0.00 16 12 0.72 4 2 0.50 57 28 49.56 Eco East 0 0 0.00 3 1 0.33 0 0 0.00 79 36 45.86 Eco West 4 2 0.43 9 5 0.53 0 0 0.00 16 4 25.00 Golf Course 87 57 0.65 0 0 0.00 11 8 0.76 153 112 73.11 Riverfront 3 1 0.40 0 0 0.00 8 6 0.75 49 31 63.27 Trout Creek 2 2 0.75 5 2 0.33 8 4 0.47 5 1 20.00 USF Central 1 0 0.00 0 0 0.00 94 31 0.33 228 164 71.87 USF East 0 0 0.00 0 0 0.00 38 22 0.57 93 62 66.67 USF West 0 0 0.00 0 0 0.00 78 28 0.36 137 89 65.20 Table 3. Total captures, recaptures and % recaptures of treefrogs.
15 Organism Total captures Osteopilus septentrionalis 1468 Hyla squirella 450 Hyla cinerea 162 Hyla femoralis 54 Anolis carolinensis 96 Anolis sagrei 196 Hemidactylus garnotii 84 Eumeces inexpectatus 6 Diadophis puctatus 1 Lampropeltis triangulum 1 Thamnophis sauritus 1 Table 4. Vertebrates found in PVC pipes during study. The number of total captures and recaptures for each treefrog sp ecies at each site is shown in Table 3. H. cinerea were captured almost exclusiv ely at the USF golf course and not at all at the natural sites Eco Central and Eco East or the urban sites USF East and USF West. H. femoralis were only captured at the natural sites and in very low numbers. The majority of H. squirella were captured at the urba n USF campus sites and not at all at the natural sites Eco East or Eco West. O. septentrionalis were captured at all sites, most frequently at the urban USF ca mpus sites and at the semi-urban USF golf course but less freque ntly at Trout Creek.
16 Figure 5. Total captures and recaptures of H. cinerea, H. femoralis, H. squirella and O. septentrionalis. EC = Eco Area Central, EE = Eco Area East, EW = Eco Area West, G = Golf course, R = Riverfr ont, T = Trout Creek, UC = USF Central, UE = USF East, UW = USF West. Recapture rates for each sp ecies of treefrog at each site are shown in Figure 6. H. cinerea had high recapture rates at the U SF golf course and at Trout Creek. H. femoralis had high recapture rates at Eco Area Central and Eco Area West. H. squirella had high recapture rates at USF East. The recapture rates of O. septentrionalis have an interesting pattern. Rates of recapture are high at all urban and semi-urban sites while are much lower at the natural sites. H. cinerea ECEEEWTGRUCUEUW 0 20 40 60 80 100 O. septentrionalis ECEEEWTGRUCUEUW # of captures 0 50 100 150 200 250 H. squirella Sites ECEEEWTGRUCUEUW # of captures 0 20 40 60 80 100 Urban Natural Semi-urban Recaptures H. femoralis Sites ECEEEWTGRUCUEUW 0 2 4 6 8 10 12 14 16 18
17 H. cinerea ECEEEWTGRUCUEUW % recapture 0 20 40 60 80 H. squirella Sites ECEEEWTGRUCUEUW % recapture 0 20 40 60 80 O. septentrionalis Sites ECEEEWTGRUCUEUW 0 20 40 60 80 H. femoralis ECEEEWTGRUCUEUW 0 20 40 60 80 Figure 6. Recapture rates of H. cinerea, H. femoralis, H. squirella and O. septentrionalis. EC = Eco Area Central, EE = Eco Area East, EW = Eco Area West, G = Golf course, R = Riverfront, T = Trout Creek, UC = USF Central, UE = USF East, UW = USF West. The results of the Schnabel population estimates are shown in Figure 7. The largest population estimate of H. cinerea was found at the USF Golf course. H. femoralis have a very small population estimate (n < 7) at all sites. The larges t population estimates of H. squirella were found at USF Central, USF West and Riverfront Park The largest population estimates of O. septentrionalis were found at the U SF golf course, Eco Central and Eco East. A 2 goodness of fit test was used to compare the population sizes of each treefrog species at each site. Sites wh ere treefrogs were not captured were not included. No significant differen ce in population size was found for H. femoralis ( 2 =
18 O. septentrionalis Sites ECEEEWTGRUCUEUW Population estimate 0 20 40 60 80 100 120 140 H. cinerea EWTGR Population estimate 0 20 40 60 80 100 120 140 H. femoralis ECEEEWT 0 5 10 15 20 25 H. squirella Sites EETGRUCUEUW Population estimate 0 20 40 60 80 100 120 140 Figure 7. Population size estima tion using Schnabel method for H. cinerea, H. femoralis, H. squirella and O. septentrionalis. Note the difference in scale. Error bars represent the upper and lower confidence limits using a 95% confidence interval. EC = Eco Area Central, EE = Eco Area East, EW = Eco Area West, G = Golf course, R = Riverfront, UC = USF Central, UE = U SF East, UW = USF West. 0.085, df = 3, P = 0.80). A significant diffe rence in population size was found for H. cinerea, H. squirella and O. septentrionalis. Correlation of Natives and Osteopilus septentrionalis The results of the Pearson correlati on test indicate that captures of O. septentrionalis and native treefrog in PVC pipes are ne gatively correlated at all sites, Table 5. These results are significant at th e 0.01 level for Eco Central, Eco West, Golf course, Trout Creek, USF East and USF West and at the 0.05 level for Riverfront Park.
19 Site r = P = Eco Central -0.585 0.01 Eco East -0.086 0.367 Eco West -0.846 0.01 Golf course -0.412 0.01 Riverfront -0.209 0.05 Trout Creek -0.844 0.01 USF Central -0.34 0.061 USF East -0.412 0.01 USF West -0.334 0.01 Table 5. Correlation of O. septentrionalis and native treefrogs Effects of Site, Tree Type and PVC Diameter on Treefrog Captures The results of the 3-way ANOVA can be found in Table 6. There was a significant difference in treefrog captures at each site for H. cinerea, H. squirella and O. septentrionalis but not for H. femoralis. A significant difference in treefrog captures among tree type was found for H. squirella and O. septentrionalis but not for H. cinerea or H. femoralis. H. femoralis and O. septentrionalis were found significantly more frequently in PVC pipes with an internal diameter of 4.45cm. There was no significant difference in treefrog captures in PV C pipes with different diameters for H. cinerea or H. squirella. The number of treefrog captures in PVC pi pes of each internal diameter can be found in Table 7. The interaction between site and tree type was significant for O. septentrionalis ; all other interactions were not significant.
20 SS df MS F P H. cinerea site 6.44 4 1.61 5.06 0.001 tree 1.23 2 0.61 1.93 0.149 diameter 0.95 1 0.95 2.99 0.086 site tree 0.50 6 0.08 0.26 0.953 site diameter 0.98 4 0.25 0.77 0.546 tree diameter 0.11 2 0.05 0.17 0.843 site tree diameter 0.47 6 0.08 0.25 0.961 Error 46.13 145 0.32 H. femoralis site 0.59 3 0.20 1.30 0.278 tree 0.34 2 0.17 1.13 0.326 diameter 1.59 1 1.59 10.57 0.001 site tree 1.05 5 0.21 1.40 0.230 site diameter 0.12 3 0.04 0.28 0.844 tree diameter 0.20 2 0.10 0.67 0.514 site tree diameter 0.21 5 0.04 0.28 0.926 Error 21.67 144 0.15 H. squirella site 21.76 6 3.63 8.52 0.000 tree 3.60 2 1.80 4.23 0.016 diameter 0.51 1 0.51 1.20 0.274 site tree 5.34 10 0.53 1.26 0.257 site diameter 2.82 6 0.47 1.10 0.360 tree diameter 0.32 2 0.16 0.38 0.686 site tree diameter 1.42 10 0.14 0.33 0.972 Error 105.58 248 0.43 O. septentrionalis site 54.90 8 6.86 10.16 0.000 tree 11.10 2 5.55 8.22 0.000 diameter 2.52 1 2.52 3.73 0.054 site tree 15.40 13 1.18 1.75 0.050 site diameter 3.95 8 0.49 0.73 0.664 tree diameter 1.38 2 0.69 1.02 0.363 site tree diameter 2.69 13 0.21 0.31 0.991 Error 220.27 326 0.68 Table 6. Site, tree type and PVC pi pe effects on treefrog captures.
21 PVC pipe diameter 1.91cm 4.45cm H. cinerea 29 158 H. femoralis 7 48 H. squirrela 228 253 O. septentrionalis 767 881 Table 7. Number of treefrog captures in PVC pipes with internal diameters of 1.91cm and 4.45cm. Distance to Water Correlation The results of the Pearson correlation wh ich compared total numbers of treefrogs captured in each PVC pipe with distance to water are shown in Table 8 and Figure 8. A significant positive correlation was found for H. squirella and O. septentrionalis. N r P H. cinerea 182 -0.130 0.078 H. femoralis 182 -0.074 0.322 H. squirella 182 0.172 0.02 O. septentrionalis 182 0.151 0.041 Table 8. Correlation of treefrog captures and distance to water.
22 O. septentrionalis Distance to water (km) 0.00.20.40.60.81.0 0 10 20 30 40 50 60 70 H. cinerea 0.00.20.40.60.81.0 Total captures 0 10 20 30 40 H. femoralis 0.00.20.40.60.81.0 0 2 4 6 8 10 H. squirella Distance to water (km) 0.00.20.40.60.81.0 Total captures 0 10 20 30 40 50 Distance to water (km) Distance to water (km) r = -0.130 P = 0.078 r = -0.074 P = 0.322 r = 0.172 P = 0.02 r = 0.151 P = 0.041 Figure 8. Correlation of treefrog cap tures and distance to water for H. cinerea, H. femoralis, H. squirella and O. septentrionalis. Seasonal and Precipitation Effe cts on Treefrog Captures Figure 9 shows the total rainfall and total treefrog captures for each month. Rainfall was the highest during the months of June through September. Small peaks in rainfall occurred in the months of Decembe r and April. There a ppears to be little variation in total captures of H. cinerea and H. femoralis each month regardless of rainfall. Total captures of H. squirella peaked in February and A ugust. Total captures of O. septentrionalis peaked in the months of Ma rch through May. Monthly treefrog captures are plotted against monthly precipita tion in Figure 10. A significantly negative correlation was found between total captures and monthly precipitation for O.
23 septentrionalis (r = -0.782, P = 0.003) as shown in Table 9. No correlation was found for H. cinerea, H. femoralis or H. squirella. Month Oct 0 6 N o v 06 Dec 06 Jan 07 F e b 0 7 Mar 0 7 A pr 07 May 07 Jun 0 7 J u l 07 Aug 07 Sep 07 log 10 + 1 (total captures) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 H. cinerea H. femoralis H. squirella O. septentrionalis Oct 06 Nov 06 D e c 0 6 Jan 07 Fe b 07 M a r 07 Ap r 07 M a y 0 7 Ju n 07 J ul 07 Aug 0 7 S ep 0 7 Precipitation (cm) 0 5 10 15 20 25 Figure 9. Monthly precipitation and total treefrog captures
24 Precipitation 0510152025 Log10+1(Total captures) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 H. cinerea H. femoralis H. squirella O. septentrionalis Figure 10. Total treefrog captures versus monthly precipitation. N r P H. cinerea 12 -0.447 0.145 H. femoralis 12 -0.344 0.274 H. squirella 12 0.339 0.281 O. septentrionalis 12 -0.782 0.003 Table 9. Correlation of total captures and monthly precipitation.
25 To determine whether treefrog species a nd season are related to each other a Gtest contingency table was created, Table 9. The G value exceeds the critical value at P = 0.001. This indicates that treefrog species and the time of year they are captured are not independent of one another. The deviations of each observed value from the expected value is shown in Figure 10. JanuaryMarch AprilJune JulySeptember OctoberDecember H. cinerea 42 42 27 51 H. femoralis 24 8 5 18 H. squirrela 151 68 184 47 O. septentrionalis 473 485 175 363 df = 1 P < 0.001 2 = 233.54 Table 10. Number of treefrog captures at different times of the year.
26 H. cinereaH. femoralisH. squirellaO. septentrionalis ln(observed/expected) -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 January-March April-June July-September October-December Figure 11. The deviation of observe d treefrog captures from the expected treefrog captures at different times of year.
27 Discussion Treefrog Density Estimates The results show that O. septentrionalis have successfully colonized the urban, semi-urban and natural sites of this fi eld study. Among all of the sites sampled, the distribution of O. septentrionalis overlaps with that of nativ e treefrog species including H. cinerea, H. femoralis and H. squirella. Moreover, as you can see from Figure 4, O. septentrionalis is not merely present in habitats where native treefrogs reside; the frequencies of O. septentrionalis are greater than the frequencie s of natives at every site. Total captures of O. septentrionalis are higher at the urban site s USF Central, USF East, USF West and the semi-urban USF golf cour se. One potential explanation for these results is that O. septentrionalis is largely present in urban and semi-urban habitats, while less so in natural habitats. However, another explanation for these results may be that the PVC pipes used in this experiment are attrac ting treefrogs in urban areas more so than they would in natural areas reflecting the lack of refuge in urban habi tats. Urban habitats lack refuge that natural habitats provide and so PVC pipes may be more attractive in urban habitats. Therefore, low numbers of captures at natural sites may be misleading. Furthermore, PVC pipes may not be a good i ndicator of the actual treefrog abundances because the refuge provided by the PVC pipe may only be attractive when little other refuge is available. It is im portant to remember that this method of trapping is different from other methods of trapping, i.e. pitfall traps or live tr aps such as Sherman traps,
28 because the organism has the ability to stay or leave. The number of captures may not be an estimate of abundance but a estimate of th e number of organisms that choose to take refuge in a PVC pipe. It is possible that very high recapture rate s of treefrogs are an indicator of a small population size. Although total cap tures at a certain site may be high this may not be an indication of the actual popul ation size if the same indivi dual is being captured over and over again. Site fidelity has been observed in treefrogs in previous studies (Irvin et al., 2007; Pittman et al. 2008) Figure 5 shows that O. septentrionalis had high total captures at the urban sites USF Central, USF East USF West and the USF golf course (semiurban) and lower total captures at the natural sites Eco Cent ral, Eco East, Eco West and Trout Creek. Figure 6 shows that O. septentrionalis has very high rates of recaptures (6573%) at the sites with high total captures a nd lower rates of recaptures (20-50%) at the sites with lower total captures. An exception to this trend for O. septentrionalis is Riverfront Park. These results strengthen the idea that O. septentrionalis may not necessarily be more abundant in urban habita ts than in natural habitats. Moreover, the Schnabel population estimate (Figure 7), whic h accounts for recapture rates, indicates that the larges t populations of O. septentrionalis can be found at Eco Central, Eco East and the USF golf course; the smallest populat ions can be found at USF Central, USF East, USF West and Eco West. The Schnabel population estimate of Trout Creek may be misleading because of the low number of captu res (n = 5). With the exception of Eco West and the USF golf course these results may indicate that O. septentrionalis can be found in natural habitats more frequently than in urban habitats.
29 The results for total captures and recapt ures for the native treefrogs do not follow the same trend as O. septentrionalis. Although H. squirella have the highest total captures at the urban site USF Central, Figur e 5, they have the lowest recapture rate, Figure 6. Additionally, the results of the Sc hnabel population estimate indicate that the largest population of H. squirella can be found at USF Central. H. squirella have previously been associated w ith urban areas; moreover they pr efer to breed in temporary ponds and ditches (Ashton and Ashton, 1988). With in USF Central there is a small ditch that holds water for most of the year; the presence of this pot ential breeding pool may account for high captures of H. squirella at this site. H. femoralis are found most frequently and have high recapture rates at the natural site Eco Area Central, Figures 5 and 6. The Schnabel population estimates for H. femoralis may be misleading because of the low number of captures in this species. H. cinerea were captured almost exclusively at the USF golf course. I believe that this is because the golf course is bordered by a wetland which includes several permanent ponds. H. cinerea is the only treefrog in this study that prefers to breed in permanent ponds, and, not coincide ntally the USF golf course is the only site used in this study that includes a permanent pond.
30 Correlation of Natives and Os teopilus septentrionalis Native treefrogs were found less frequen tly in PVC pipes th at also contained O. septentrionalis, Table 7. This result is perhaps the most interesting because it may be indicating that native treefrogs w ill avoid habitat or refuges where O. septentrionalis are present. Another possibility is that natives are taking refuge in PVC pipes containing O. septentrionalis but leave pipes to escape predation; or maybe natives were not able to escape quickly enough and were consumed by O. septentrionalis. Effects of Site, Tree Type and PVC Pi pe Diameter on Treefrog Captures The results of the 3-way ANOVA show that there is an interaction between tree type and site for O. septentrionalis Table 6. This indicates th at the habitat location of Quercus, Sabal and Pinus may be influencing their ability to attract treefrogs. Moreover, this suggests that the sites may differ in ability to attract tr eefrogs due to differences in plant community, Table 2. No interaction was found between tree type and pipe diameter, Table 6. This suggests that pipe diameters do not differ in the ability to attract treefrogs when hung on different tree types. All species of treefrog were found more frequently in PVC pipes with an internal diameter of 4.45 cm. These results were significant for H. femoralis and O. septentrionalis. This may be because of the larger volume of water large PVC pipes retain after a rain. Treefrogs are perhaps attracted to the pipes for the offered moisture and not necessarily the size. In addition it is possible that the small PVC pipes used are excluding very large individuals of H. cinerea or O. septentrionalis Previous studies have shown that H. cinerea and H. squirella are captured more frequently in 1.91cm PVC pipes (Zacharow et al. 2003).
31 H. squirella and O. septentrionalis were found significantly more frequently in PVC pipes hung on Quercus Table 6. Although not significant H. cinerea were found more frequently in Sabal and H. femoralis were found more frequently in Pinus. Boughton et al. (2000) examined the effects of tree type on treefrog capture rates using hardwood trees, including Quercus, and softwood trees, including Pinus. Their results indicate that H. cinerea and H. squirella were captured significantly more frequently on hardwood tree species. Distance to Water Correlation Although significant, the posit ive correlation of treefrog captures and distance to water for H. squirella and O. septentrionalis was very weak, Table 8 and Figure 8. However a positive correlation is interesting because it suggests more than one possibility. One explanation of these results is that as the distance to water increases the number of treefrogs also increases. Anothe r explanation is that as distance to water increases the PVC pipes become more attractiv e to treefrogs. After a moderate rain, water that has collected in PVC pipes would remain for at least 2 weeks. The moisture offered by PVC pipes placed far from water may be what is attracting the treefrogs. Seasonal and Precipitation Effe cts on Treefrog Captures The results of the G-test indicate that the species of treefrogs examined in this study are behaving differently at different times of the year The time of y ear with the highest frequency of captures was different for each species of treefrog as shown in Table 11. However the time of year with the s econd highest frequency, January through March, was the same for H. cinerea, H. squirella, and O. septentrionalis. H. femoralis were captured the most frequently during this time period. Additionally, the total number of
32 treefrogs captured in January through March was greater than any other time period. This pattern suggests that PVC pi pes are more attractive to tr eefrogs during cooler months. Previous studies have also concluded that treefrogs will seek refuge in PVC pipe and other types of shelter in cold weather (Goin 1958, Zacharow et al. 2003) During the months of July through September H. cinerea, H. femoralis and O. septentrionalis were captured the least frequently. H. squirella were captured the most frequently during this time period. These results ar e considered further in the discussion below. A negative correlation between treefrog captu res and rain fall does not necessarily indicate that O. septentrionalis are more abundant during dry m onths. It is more probable that PVC pipes become more attractive to O. septentrionalis in the drier months. This may be because PVC pipes collect water af ter a rain and retain moisture for a considerable time. As time after a hard rain increases total captures of O. septentrionalis also increase. In months with high rainfall O. septentrionalis may no longer rely on moisture offered by the PVC pipes and may find refuge elsewhere; this may explain dips in total captures in months with or after hi gh rainfall, Figure 9. No pattern seems to exist between total captures and rainfall for H. cinerea, H. femoralis or H. squirella. Peaks in total captures during the winter months may be explained by temperature or rainfall. Invasion of Florida by Os teopilus septentrionalis The invasion of Florida by O. septentrionalis is interesting because it is an example of an island species who has invade d a mainland. It is ge nerally accepted that islands are susceptible to invasion. Moreover sp ecies endemic to islands are vulnerable to predation and are also poor competitors as a result of evolving a habitat with few natural predators and competitors (Lazell, 2005). These generalizations do not apply to O.
33 septentrionalis ; here is an island species who is an excellent competitor and evader of predators. O. septentrionalis has dispersed beyond the insulate d southern peninsula and into parts of Central and North Florida. 18 of the 36 non-native herpetofauna in Florida are also endemic to islands. However the majority of these species have not dispersed very far and most have remained within one county as can be seen in Figure 11. O. septentrionalis, Anolis sagrei and Eleutherodactylus planirostris do not, however, follow this trend and have successfully colonized la rge portions of southern, central and north Florida (Butterfield et al., 1997). A. sagrei and E. planirostris first colonized Florida in the late 1800's however O. septentrionalis was not reported on mainland Florida until 1952. The dispersal of O. septentrionalis has happened relatively quickly. Simberloff (1997) describes two methods of dispersal; diffusion and jump dispersal. Diffusion dispersal occurs as an organism slowly expands its distri bution through natural emigration to neighboring habitats. Jump dispersal may occur when the dispersing organism is brought to a new location, far from the habitat in which it originally colonized. A good example of jump dispersal in treefrogs is indivi duals who will hitch rides on cars while taking refuge in door gaps (personal observation). Perhaps one more reason why O. septentrionalis is such a successful colonizer is that it is dispersing through Florida by both diffusion and jump dispersal.
34 Species A. ameiva A. c hl o roc yan u s A. cristat e llus A c y bo t es A. distichus A. equ e st ri s A. ga rma ni A. sagrei B. con st rict o r E p la n iro st ri s G a l bo gu l ari s I i g ua n a L. c ari n at u s L. pe rso natus L. schreibersi O. septentrionalis S. arg us S. elegans Number of Florida counties 0 10 20 30 40 50 60 Figure 12. Dispersal of non-native herpetofauna endemic to islands, in Florida counties. Data was obtained from Strangers in Paradise (Butterfield et al. 1997). The number of counties in which O. septentrionalis are present have been m odified to the current 2008 distribution.
35 Conclusions Perhaps the most important finding in this study is that PVC refugia, when used in estimates of treefrog density, are potentially pr oducing biased results. Available refuge, temperature and precipitation all appear to have an effect on how attractive PVC pipes are to treefrogs. If the presence of O. septentrionalis in Florida is a real th reat to populations of native treefrogs then the results of this study are concerning. Future studies of O. septentrionalis using PVC pipe refugia should co nsider using PVC pipes of a 4.45cm internal diameter or larger which are associated with Quercus or other hardwood trees. Additionally the capture biases mentioned above should be considered in future estimates of treefrog densities. A much longer term, and larger scale, study is needed to determine if the presence of O. septentrionalis is negatively influencing the popul ations of native treefrogs in Florida. Campbell et. al (2007) has proposed a removal study of O. septentrionalis from a public park in Hillsborough County. Their study ma y lead to a better understanding of the effects of the invasion by O. septentrionalis. However, due to the exceptional ability of O. septentrionalis to colonize as well as the conti nued immigration from native and novel ranges, this invader is undoubtedly taking permanen t residence in Florida.
36 References Bartareau, T.M. PVC Pipe Diameter Influe nces the Species and Sizes of Treefrogs Captured in a Florida Coastal Oak Scrub Community. 2002. Herpetological Review 35(2): 150-152. Barbour, T. 1931. Another Introduced Frog in North America. Copeia 3:140 Behler, J. L., and F. W. King. 1979. Th e Audubon Society Field Guide to North American Reptiles and Amphibians. Besag, J., Gleaves, J.T. 1973. On the detecti on of spatial pattern in plant communities. Bulletin of the International Statistical Institute 45:6. Boughton, R. G., J. Staiger, and R. Franz. 2000. Use of PVC pipe refugia as a sampling technique for hylid treefrogs. Ameri can Midland Naturalist 144:168-177. Butterfield, B.P., W.E. Meshaka and C. Guyer. 1997. Nonindigenous Amphibians and Reptiles. In D. Simberloff, D.C. Schmitz and T.C. Brown (Eds.), Strangers in Paradise (pp. 123-138). Wash ington, DC: Island Press. Campbell, K.R., T.S. Campbell, S.A. Johnson. 2007. Morris Bridge Wellfield/ Flatwoods Park long-term amphibian monitoring project, Hillsborough County, Florida. Annual Meeting of the Florid a Academy of Sciences abst ract, Biological Sciences (BIO) C: Bio C-06. Conant, R., and J. T. Collins. 1998. A Field Guide to Reptiles and Amphibians Eastern and Central North America. H oughton Mifflin Company, New York.
37 Goin, O. B. 1958. A comparison of the nonbreeding habits of two treefrogs, Hyla squirella and Hyla cinerea Quarterly Journal of the Fl orida Academy of Sciences 21:12. Greenlees, M.J., G.P. Brown, J.K. Webb, B.L. Phillips and R. Shine. 2007. Do invasive cane toads ( Chaunus marinus ) compete with Australian frogs ( Cyclorana australis )? Austral Ecology 32:900-907. Irvin, P., T.S. Campbell, K.R. Campbell, S.A. Johnson. 2007. Population-level effects of introduced Cuban treefrogs ( Osteopilus septentrionalis) on native treefrog assemblages in isolated wetlands. Annua l Meeting of the Florida Academy of Sciences abstract, Biological Sciences (BIO) C: Bio C-08. Johnson, S.A. 2007. The Cuban treefrog ( Osteopilus septentrionalis ) in Florida. University of Florida, Institute of Food and Agricultural Sciences. Krebs, C. J. 1999. Ecological Methodol ogy. 2nd edition. Benjamin/Cummings, Menlo Park. Lazell, J. 2005. Island Fact and Theory in Nature. University of California Press, Berkeley. Lindsay, K. and B. Cooper. 2008. Osteopilus septentrionalis becomes established on Antigua, West Indies. App lied Herpetology 5(1):96-98. Love, W. B. 1995. Osteopilus septentrionalis (Cuban Treefrog). Predation. Herpetological Review 26:201-202. Meshaka, W. E. 1996. Vagility and the Fl orida Distribution of the Cuban Treefrog ( Osteopilus septentrionalis ). Herpetological Review 27:37-40.
38 Meshaka, W. E. 1997. Osteopilus septen trionalis (Cuban treefrog). Predation. Herpetological Review 28:147. Meshaka, W. E. 2001. The Cuban Treefrog in Florida. University Press of Florida, Gainesville. Meshaka, W. E., and B. Ferster. 1995. Two sp ecies of snakes prey on Cuban treefrogs in southern Florida. Florid a Field Naturalist 23:97-98. Oliver, J.A. Anolis sagrei in Florida. Copeia 1:55-56. Pittman, S.E., A.L. Jendrek, S.J. Price, M. E. Dorcas. 2008. Habitat Selection and Site Fidelity of Cope's Gray Treefrog ( Hyla chrysoscelis ) at the Aquatic-Terrestrial Ecotone. Journal of Herpetology In press. Powell, R., R.J. Passaro, and R.W. Hende rson. 1992. Noteworthy herpetological records from Saint Maarten, Netherlands, Antille s. Caribbean Journal of Science 28:234235. Salinas, F. V. 2006. Breeding behavior and colonization success of the Cuban treefrog Osteopilus septentrionalis Herpetologica 62:398-408. Schnabel, Z. E. 1938. The estimation of the total fish population of a lake. American Mathematician 45:5. Schwartz, A. 1952. Hyla septentrionalis Dumeril and Bibron on the Florida mainland. Copeia. 1952:117-118. Simberloff, D. 1997. The Biology of Invasions. In D. Simberloff, D.C. Schmitz and T.C. Brown (Eds.), Strangers in Paradise (pp. 3-17). Washington, DC: Island Press.
39 Smith, K.G. 2005. Effects of nonindigenous ta dpoles on native tadpol es in Florida: evidence of competition. Biologi cal Conservation 123: 433-441. Smith, K.G. 2005. An exploratory assessment of Cuban Treefrog ( Osteopilus septentrionalis ) tadpoles as predators of na tive and nonindigenous tadpoles in Florida. Amphibia-Reptilia 26:571-575. Smith, K.G. 2006. Keystone predators (eastern newts, Notophthalmus viridescens ) reduce the impacts of an aquatic invasi ve species. Oecologia 148:342-349. Soma, L.A. 2007. Osteopilus septentrionalis USGS Nonindigenous Aquatic Species Database, Gainesville, FL. Townsend, J.H., J.M. Eaton, R. Powell, J. S. Parmerlee Fr., and R.W. Henderson. 2000. Cuban treefrogs ( Osteopilus septentrionalis )in Anguilla, Lesser Antilles. Caribbean Journal of Science 36:326-328. vanBuurt, G. 2007. Breeding population of Osteopilus septentrionalis in Curacao. Applied Herpetology 4(4):390-391. Williamson, M. 1999. Invasions. Ecography 22: 5-12. Wyatt, J. L., and E. A. Forys. 2004. Conser vation implications of predation by Cuban Treefrogs (Osteopilus septentrionalis) on native hylids in Florida. Southeastern Naturalist 3:695-700. Zacharow, M., W. J. Barichivich, and C. K. Dodd. 2003. Using ground-placed PVC pipes to monitor Hylid treefrogs: Capture bias es. Southeastern Naturalist 2:575-590.