Habitat Occurrence of Florida s Native Amphibians and Reptiles

Transcription

1 Habitat Occurrence of Florida s Native Amphibians and Reptiles Technical Report No. 16 Kevin M. Enge February 1997 Florida Game and Fresh Water Fish Commission 620 South Meridian Street Tallahassee, Florida

2 Habitat Occurrence of Florida s Native Amphibians and Reptiles Technical Report No. 16 Kevin M. Enge Florida Game and Fresh Water Fish Commission Route 7, Box 3055, Quincy, Florida February 1997

3 ii FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 Suggested citation: Enge, K. M Habitat occurrence of Florida s native amphibians and reptiles. Tech. Rep. No. 16. Florida Game and Fresh Water Fish Comm., Tallahassee. 44 pp + vi.

4 HABITAT OCCURRENCE OF FLORIDA S NATIVE AMPHIBIANS AND REPTILES Enge iii ABSTRACT Florida has a diverse herpetofauna of at least 144 native species, with the highest species diversity in the upper Apalachicola River basin and a progressive decline southward along the peninsula. The relative abundance of amphibians and reptiles in a particular habitat is determined, in part, by each of the following: topography, fire frequency, hydroperiod, soil, water quality and source, and juxtaposition of habitats in the landscape. Microclimate, microhabitat (e.g., gopher tortoise [Gopherus polyphemus] burrows), and past or present human disturbance may also help determine herpetofaunal distribution patterns. Fire is an important factor in maintaining many habitats and in determining the relative abundance of amphibian and reptile species, which are often adapted to fire. Habitat use by a given species may be seasonal, and population densities may fluctuate greatly depending upon reproductive success and mortality, which for many amphibian species are heavily influenced by rainfall. Tables are provided that give the relative abundance of various native amphibian and reptile species in 31 Florida habitats based on drift-fence surveys, literature, observations, and educated guesses. The tables are intended to provide land managers with a list of species that can be expected to occur on a tract of land with particular habitats and to enable researchers to ascertain the effectiveness of their herpetofaunal inventory efforts. The herpetofauna of xeric upland habitats is often adapted to fire, frequently uses burrows, and often contains sand-swimming reptile species. Mesic upland habitats may have a diverse herpetofauna, and salamanders may be a dominant component along streams, especially in the Panhandle. Mesic flatland habitats may contain a diverse herpetofauna ranging from xeric-adapted to aquatic species in different seasons. Wet flatland and floodplain wetland habitats often have dissimilar herpetofaunal assemblages, especially of aquatic species, because of differences in the dominant vegetative cover (i.e., trees or grasses) and hydroperiod duration and water depth. Basin wetland habitats are important breeding sites for many amphibian species because of their ephemeral nature and scarcity of aquatic predators. Although seepage slope habitats in the Panhandle tend to be small in size, they are important reservoirs for rare species such as the Florida bog frog (Rana okaloosae) and pine barrens treefrog (Hyla andersonii). The highly fragmented rockland habitats of South Florida support populations of many state-listed taxa, especially in the Lower Keys. Coastal habitats have low herpetofaunal species richness because of their harsh environmental conditions and scarcity of fresh water. Offshore islands have fewer species than the mainland, but populations of some species can be very dense.

5 iv FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 ACKNOWLEDGMENTS I would like to thank Dirk Stevenson and Paul Moler for reviewing the habitat occurrence tables and providing input on the text. I also am indebted to Florida Game and Fresh Water Fish Commission personnel who provided me with unpublished herpetofaunal survey data: Brian Millsap, Stephen Stiegler, Tim Towles, and Kristin Wood. Employees with the Florida Department of Environmental Protection who graciously provided unpublished drift-fence data were Bert Charest, Terry Hingtgen, Dan Pearson, and Mathew Wingate. Additional herpetofaunal survey data were obtained from George Dalrymple, Bruce Means, and Karl Studenroth. David Cobb, Don Wood, David Cook, and 3 anonymous reviewers provided comments on drafts of this manuscript.

6 HABITATOCCURRENCE OF FLORIDA S NATIVEAMPHIBIANS AND REPTILES Enge v TABLE OF CONTENTS ABSTRACT iii ACKNOWLEDGMENTS iv INTRODUCTION Patterns of Species Richness Temporal Variation in Habitats and Habitat Use Effects of Fire METHODS Regions Habitat Types Relative Abundance RESULTS AND DISCUSSION Herpetofauna of Xeric Uplands Herpetofauna of Mesic Uplands Herpetofauna of Mesic Flatlands Herpetofauna of Wet Flatlands Herpetofauna of Floodplain Wetlands Herpetofauna of Basin Wetlands Herpetofauna of Seepage Wetlands Herpetofauna of Rockland Habitats Herpetofauna of Coastal Habitats Herpetofauna of Aquatic Habitats LITERATURE CITED APPENDIX

7 HABITAT OCCURRENCE OF FLORIDA S NATIVE AMPHIBIANS AND REPTILES Enge 1 INTRODUCTION This report attempts to characterize the native herpetofaunal assemblages of Florida s major habitat types (excluding streams, ponds, lakes, and marine communities) using data from drift-fence studies, published literature, and personal observations. Little information is available on the herpetofauna of some habitats, so educated guesses are made regarding the typical herpetofauna in order to provide a basis for further research or discussion. The relative abundance of native amphibian and reptile species in various habitats in different regions of Florida are provided in tabular form. The tables are intended to provide land managers with a list of species that can be expected to occur on a tract of land with particular habitats and to enable researchers to ascertain the effectiveness of their herpetofaunal inventory efforts. The tables will be updated as new data become available and comments are received from knowledgeable field biologists. The information provided in the text is intended to familiarize persons with some of the published information on the herpetofaunal use of habitats in Florida. Scientific names of most amphibian and reptile species mentioned in the text are given in the appendix. Several authors have developed classification schemes to reflect the relative abundance of Florida s amphibian and reptile species in different habitats (e.g., Florida Game and Fresh Water Fish Commission 1976b; Layne et al. 1977; Ashton and Ashton 1988, 1991a,b), the importance of habitats to species (e.g., Duever et al. 1979), or lists of expected or observed species in different habitats (e.g., Carr 1940, Duellman and Schwartz 1958, Harima 1969, Auffenberg 1981, Wilson and Porras 1983, Wolfe et al. 1988). Most of these lists were based on personal observations, published literature, and guesswork, but subsequent research and drift-fence surveys have provided additional information on many of the more cryptic species and poorly known habitats. Drift-fence surveys, which have been conducted in Florida since 1975, provide information on habitat use and relative abundance of herpetofauna (see Enge 1997a for a summary of Florida studies), allowing modification of previous tables. Few drift-fence surveys of wetland habitats in Florida were conducted prior to 1990 because of problems with installing fences in standing water, especially in unconsolidated substrates. However, use of silt fencing, a woven polypropylene material used to control sediment runoff from construction sites, has since permitted drift-fence surveys of many of Florida s wetland habitats (Enge 1997c). Florida supports a diverse native herpetofauna of at least 27 anuran, 26 salamander, 2 crocodilian, 27 turtle, 1 amphisbaenid, 17 lizard, and 44 snake species. There are probably herpetofaunal taxa yet to be discovered in Florida, and the herpetofauna of many habitats remains relatively poorly studied. Since 1970, populations of 4 recognized species the seal salamander (Means and Longden 1970), many-lined salamander (Christman and Kochman 1975), pine barrens treefrog (Christman 1970), and carpenter frog (Stevenson 1970) have been documented in Florida. In the past 10 years, 6 previously unknown or unrecognized species have been described from Florida: Apalachicola dusky salamander (Means and Karlin 1989), southeastern slimy salamander (Highton et al. 1989), southern dwarf siren (Moler and Kezer 1993), Florida bog frog (Moler 1985), Escambia map turtle (Lovich and McCoy 1992), and mimic glass lizard (Palmer 1987). The taxonomy of many taxa has changed. For example, the 2 subspecies of river cooter were recently given species status, as were the 2 subspecies of Florida cooter (Seidel 1994, 1995), although this taxonomic revision has been challenged (Jackson 1995). At least 2 new salamander species remain undescribed (Palis and Jensen 1995; P. E. Moler, Florida Game and Fresh Water Fish Comm., pers. commun.). Based on range maps by Conant and Collins (1991), there are 37 currently recognized native herpetofaunal taxa (i.e., species and subspecies) endemic to Florida and another 10 native taxa nearly endemic to Florida. Patterns of Species Richness A continental pattern of increasing species richness with decreasing latitude is evident in amphibians, reptiles (Kiester 1971, Arnold 1972, Schall and Pianka 1978), land birds (MacArthur and Wilson 1967), and mammals (Simpson 1964). Snake species richness in the Northern Hemisphere increases with decreasing latitude, and when the effect of latitude is removed, most of the variation in snake species diversity can be accounted for by differences in prey species richness, such as anurans and lizards (Arnold 1972). The greatest species richness of freshwater turtles (13 species) in Florida occurs in the Escambia River basin in the western Panhandle; other basins in the Panhandle and northcentral Florida contain 12 species (Iverson and Etchberger 1989). Northern Florida, therefore, has the third richest turtle fauna in the world, which can be attributed partly to the warm, temperate,

8 2 FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 maritime climate with abundant rainfall (Iverson and Etchberger 1989). The greatest herpetofaunal species richness in North America north of Mexico occurs in the upper Apalachicola River basin, which has the greatest topographic relief (and a correspondingly high diversity of habitats) in the state (Means 1977). The high species richness of anurans, turtles, and snakes in this river basin may be a function of Florida s warm climate, high humidity, and abundant precipitation, but the latter 2 climatic variables may also account for the low species richness for lizards (Means 1977). Geographic variation in lizard species density is correlated with mean annual sunfall (i.e., number of clear days over extensive areas), which may explain why deserts have high lizard species density (Bogert 1949, Schall and Pianka 1978). The major center of salamander distribution and greatest species density are found in the Appalachian Mountains (Kiester 1971, Porter 1972), but an unexpectedly large number of salamanders occurs in the Apalachicola River basin, which may reflect the system of ravines with cooler microclimates and a diverse assemblage of swampdwelling and aquatic species (Means 1977). The Panhandle is situated such that it supports species from 5 geographic areas of endemism: north, Atlantic Coastal Plain, Gulf Coastal Plain, peninsular Florida, and western United States (Neill 1957a, Goin 1958, Means 1977, Auffenberg 1982). The number of native amphibian and reptile species declines as one progresses southward through Florida s peninsula (Duellman and Schwartz 1958, Kiester 1971, Means and Simberloff 1987, Auth 1989). The peninsula effect of reduced species diversity on peninsulas versus equal-sized mainland regions has also been observed for mammals (Simpson 1964) and land birds (MacArthur and Wilson 1967). The Upper Keys are less species-rich than the Lower Keys (Auffenberg 1982, Auth 1989), perhaps because of mainland species historically moving overland to the Lower Keys (which are composed of Miami oolite, the same foundation as the southern mainland and islands of Florida Bay) prior to land subsidence. The Upper Keys were not yet present because they are based on a coral reef, the Key Largo formation, that apparently grew in the Pamlico Sea (Neill 1957a). The Keys have less than 1/3 as many amphibian and reptile species as the mainland, which may reflect the reduction in surface fresh water and habitat types, and the higher extinction rates characteristic of islands (Auth 1989). The occurrence of amphibians in the Keys is probably affected by the lower precipitation compared to the rest of Florida, and the absence of a pronounced wet season (Chen and Gerber 1990). The decreasing herpetofaunal species richness through Florida s peninsula could be attributed to a reduction in topographic and habitat variation (Means and Simberloff 1987) instead of to a peninsula effect. The low, flat, extensive wetlands comprising the Everglades-Big Cypress region have the lowest herpetofaunal species diversity of any equally sized area on Florida s peninsula, perhaps because of the lack of upland species (Means and Simberloff 1987). The diversity of trees, which are used by many terrestrial herpetofaunal species, decreases in southern Florida, and most tropical hardwoods occur in stands with a patchy distribution. The lack of acid wetlands and large streams in southern Florida also helps explain why some amphibian and reptile species do not occur there (Means and Simberloff 1987). Fourteen amphibian species that depend upon winter rains to provide suitable breeding and larval habitat, and 6 species that require 9 20 months in the aquatic larval stage, are absent from southern Florida (Means and Simberloff 1987). Species density in Florida by county (excluding marine turtles and the American crocodile), also reflects a peninsula effect for anurans, salamanders, turtles, lizards, and snakes (Auth 1989). The highest herpetofaunal species density (107 species) occurs in the Panhandle counties of Jackson, Gadsden, Liberty, and Wakulla, and the lowest species density (61 species) is in continental Monroe County (Auth 1989). Maximal amphibian species density occurs in the Panhandle (Auth 1989). Salamander species density declines much more precipitously than anuran species density in the peninsula (Auth 1989). The patterns of turtle and snake maximum species densities match better than the pattern for lizards; the region of maximum species density for lizards extends much farther southward through the peninsula along the well-drained soils of the Lake Wales Ridge (Auth 1989). Although long-term climatological and geological factors (e.g., erosion, land subsidence, sea level changes) control the general distribution of species and habitats, ecological events of the last 40,000 years likely determine current herpetofaunal distributions (Auffenberg 1982). The amount of insolation, mean annual temperatures, maximum and minimum temperatures, growing season lengths, and mean annual and seasonal precipitation all may affect the distribution of amphibian and reptile species. Temperature extremes, rather than mean annual temperatures, are likely more important to amphibians and reptiles because they can result in

9 HABITAT OCCURRENCE OF FLORIDA S NATIVE AMPHIBIANS AND REPTILES Enge 3 death or reproductive incapacitance (Auffenberg 1982) and affect plant communities (e.g., frost killing mangrove trees). Florida s humid, subtropical climate is characterized by a cool, dry season and a warm, rainy season that may include tropical storms and hurricanes. Because Florida is a long peninsula with a north-south orientation, there is a southward decline in annual temperature range and an increase in seasonal rainfall variation (Chen and Gerber 1990). The more temperate conditions of North Florida (especially the Panhandle), with its cooler weather and relatively high winter precipitation, have important implications for the distribution of many amphibian species. Nineteen of Florida s 26 salamander species are restricted to the northern peninsula and/or Panhandle. The number of species within a given habitat can be expected to increase with the structural complexity of the habitat and sometimes with an increase in primary productivity (MacArthur 1965). Plant diversity may lead to increased herbivore diversity and secondarily to increased diversity of carnivores such as anurans, lizards, and snakes. It has been postulated that diversity tends to increase with ecological succession; however, in Florida and elsewhere, herpetofaunal species diversity does not appear to increase with ecological succession (Auffenberg 1982). Herpetofaunal species diversity is, instead, often higher in earlier successional stages or subclimaxes than in climax communities because the more open canopy and greater insolation result in higher food productivity and more diverse thermal gradients (Auffenberg 1982). Temporal Variation in Habitats and Habitat Use Habitats are dynamic, and so are their herpetofaunal assemblages. Long-term changes in herpetofaunal relative abundance are often related to successional community changes (Christman et al. 1979, Fitch 1982, Mendelson and Jennings 1992). As succession occurs, populations of some species may decline or disappear, and new species may colonize the now suitable habitat. Some herpetofaunal species attain their highest population densities in early-successional habitats created by disturbances such as fires, hurricanes, and logging (Christman et al. 1979, Greenberg et al. 1994). The relative abundance of amphibians and reptiles in a particular habitat is determined, in part, by each of the following factors: topography (i.e., elevation, slope aspect and steepness), fire frequency, hydroperiod, soil (fertility, particle size, ph, and depth), water quality and source, and juxtaposition of other habitats in the landscape. Differences in density and structural characteristics of the overstory, understory, and ground cover may also affect the relative abundance of herpetofaunal species in a particular habitat, partly by influencing the microclimate (i.e., temperature, light intensity, relative humidity, wind). The relative abundance of herpetofauna in a habitat can be changed by human activities such as pollution or use of pesticides (Wilson and Porras 1983, Harris and Vickers 1984), wetland drainage or ditching (Vickers et al. 1985), logging and site preparation (Christman et al. 1979, U.S. Fish and Wildlife Service 1980, Enge and Marion 1986, Anderson and Tiebout 1993, Greenberg et al. 1994), land clearing for agriculture (White 1983, Humphrey et al. 1985, Dalrymple 1988), habitat fragmentation (Stout and Corey 1995, Miller and Schaefer 1993, McCoy and Mushinsky 1994), road construction (Hellman and Telford 1956, Cristoffer 1991, Enge 1992), harvest for pets or food (Enge 1993), prescribed burning (Campbell and Christman 1982b; Mushinsky 1985, 1992), and suppression of natural fires. Habitat fragmentation may preclude the presence of viable populations of species with large home ranges, such as the eastern indigo snake (Moler 1982, Timmerman 1994). Herpetofaunal relative abundance also may be affected by animal disturbances, such as rooting feral hogs (Sus scrofa), grazing cattle, and the introduction of nonnative species. At least 30 species of nonnative amphibians and reptiles have established populations in Florida (McCoid 1995), and some of these particularly the cane toad (Bufo marinus), Cuban treefrog (Osteopilus septentrionalis), brown anole (Anolis sagrei), and spectacled caiman (Caiman crocodilus) may affect the relative abundance of native amphibians and reptiles through competition or predation (King and Krakauer 1966, Rossi 1981, Wilson and Porras 1983, Campbell and Gerber 1996). Litter-layer disturbance and predation by the nonnative nine-banded armadillo (Dasypus novemcinctus) (Fitch et al. 1952, Auffenberg and Iverson 1979, Jones 1992, Carr 1994, Aycrigg et al. 1996) and predation by red imported fire ants (Solenopsis invicta) (Mount 1981) may be particularly significant in affecting herpetofaunal relative abundance. Seasonal activity patterns and habitat use of herpetofauna vary immensely. The biphasic or complex life cycle of most amphibians (Wilbur 1980) results in larval and postmetamorphic amphibians of the same species often utilizing different habitats. Florida amphibians that breed in wetlands with long hydroperiods can extend their breeding season throughout the year, allowing them

10 4 FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 to temporally segregate their use of wetlands and reduce larval competition (Heatwole 1982). Upland amphibians travel to wetlands to breed, and their larvae are restricted to wetland habitats for several weeks to >1 year depending on the species. Breeding amphibians may remain in wetland habitats for only 1 or a few days in the case of explosive breeders (e.g., eastern spadefoot). Although the breeding season of barking treefrogs lasts 4 5 months in northern Florida, most males spend only 2 3 days in breeding choruses (Murphy et al. 1993). Treefrogs and other terrestrial amphibians often move to wetlands at night and return to more upland habitats during the day to forage and hide. Aquatic turtles and some aquatic snakes (e.g., eastern mud snake) typically move to upland habitats for short periods of time to deposit eggs (Semlitsch et al. 1988, Gibbons et al. 1990). In fall, adult eastern mud turtles sometimes travel >1 km from aquatic habitats to overwinter (Bennett et al. 1970). Aquatic turtles (e.g., yellowbelly slider, cooters, and chicken turtle) and aquatic snakes (e.g., black swamp snake) also may move overland between wetlands, especially to escape drought conditions (Gibbons et al. 1983; Dodd 1992, 1993). Habitat use by snakes may vary depending on their sex and reproductive condition, foraging and digestive state, ecdysis, disease and injury, social relationships, and learning (Reinert 1993). Snakes may travel through atypical habitats during the breeding season, especially male snakes making long-distance movements in search of mates. Activity patterns of snakes vary among species, sexes, and age classes (Dalrymple et al. 1991). In a study of striped crayfish snakes in water hyacinths in South Florida canals, Godley (1980) found that the mean density of snakes was 30 times greater in winter than in summer, because snakes moved from surrounding flooded marshes to water hyacinths in fall in response to fluctuations in water level, prey availability, and predator pressure. Winter refugia are not a limiting factor in most of Florida, and snakes need not move long distances to communal hibernacula, as sometimes occurs in higher latitudes (Parker and Brown 1980, Gregory 1984). Upland snakes may periodically use wetland habitats to obtain water or exploit food resources. For example, corn snakes used wet prairies and the marshy borders of wetlands during the summer, and xeric uplands as migratory routes and winter-use areas (Franz 1995). Other upland snake species that frequently forage in or around wetlands are the racer, eastern hognose snake, and eastern indigo snake. Aquatic snake species are sometimes found in upland habitats. Black swamp snakes seasonally moved through upland habitat separating a lake and temporary pond; the pond was utilized even during years when it lacked water because of drought conditions (Dodd 1993) Population densities of some species, particularly ones with rapid population turnovers like many anurans and lizards, may exhibit marked annual fluctuations. For example, during drought periods, amphibians dependent on ephemeral breeding ponds may not successfully reproduce for 1 year (Pechmann et al. 1989, Dodd 1992, Murphy et al. 1993), which could subsequently affect predator populations. After successful reproductive episodes, large numbers of newly metamorphosed amphibians (and to a lesser extent, neonate reptiles) may temporarily create a high relative abundance of some species (e.g., eastern spadefoot) in a habitat (Neill 1957b, Enge and Marion 1986), but mortality and dispersal will thin out this influx of individuals. Annual drying of wetlands is important for the reproductive success of some amphibian species, because populations of aquatic predators (e.g., other amphibians, invertebrates, fishes) are reduced. Short hydroperiods will only result in metamorphosis of amphibian species with short larval periods, and late-breeding species may not even get a chance to breed (Pechmann et al. 1989). The annual variation in amphibian reproductive success is illustrated by a drift-fence study at a wetland in South Carolina, where annual numbers of successfully metamorphosing juvenile amphibians were positively correlated with hydroperiod length, ranging from as few as zero to as many as 75,644 individuals of 15 species (Pechmann et al. 1989). Herpetofaunal populations using marshes and other ephemeral wetlands in southern peninsular Florida are subjected to more pronounced seasonal fluctuations in water tables than farther north. The high water levels present in South Florida marshes during the summer rainy season, which extends from May through October, decline through the winter and disappear in spring as evapotranspiration increases (Kushlan 1990). Normal hydroperiods and water depths have been altered in the Everglades ecosystem by an elaborate water management network of canals and levees south of Lake Okeechobee. The resulting hydrologic regimes are inconsistent with, and less predictable than, the natural seasonal pattern and have changed plant communities and their faunal use in the Everglades (Loftus et al. 1990). In marshes in northern peninsular and Panhandle Florida, higher winter precipitation means water is often present through early spring. Amphibian populations that rely on

11 HABITAT OCCURRENCE OF FLORIDA S NATIVE AMPHIBIANS AND REPTILES Enge 5 ephemeral wetlands may have a higher frequency of successful annual recruitment of young into the population in northern Florida than in southern Florida. Amphibians and reptiles are adapted for dealing with low-energy environments and are able to survive periods of prolonged shortages of water, food, or oxygen that would kill birds, mammals, and many fishes (Pough 1980, 1983). Effects of Fire Fire is an important factor in altering and maintaining many of Florida s habitats and in determining their herpetofaunal assemblages. Florida has one of the highest frequencies of lightning strikes of any region in the United States (Abrahamson et al. 1984). Lightning and man-made fires may either ensure the continued existence of many fire-dependent habitats, or they may radically alter the vegetative composition and structure of other habitats. Florida habitat types that typically burn under natural conditions at intervals of 10 years are sandhill, upland pine, pine rockland, mesic flatwoods, wet flatwoods, dry prairie, wet prairie, depression marsh, basin marsh, and swale (Florida Natural Areas Inventory 1990). It is often difficult to identify habitat types in the field because human exclusion of fire has altered natural fire frequencies and the corresponding appearance of these habitats. Many wildlife species have evolved to cope with fire, especially in habitats where fires are frequent (Komarek 1969, Ernst et al. 1995). However, glass lizards appear to suffer substantial mortality from fire (Babbitt and Babbitt 1951, Means and Campbell 1981), although they are often most abundant in firedependent habitats. Fires in wetlands with saturated soils or standing water may result in minimal mortality of herpetofauna, whereas fires during drought conditions may result in extensive mortality of a variety of herpetofauna (Wade et al. 1980). There are limited refugia from fire in non-inundated marshes, and fires can consume organic soil. Such destructive peat fires will kill even burrowed animals and alter the subsequent vegetative community. The reduced vegetative evapotranspiration after fire may result in longer hydroperiods and deeper water levels in some habitats like herb bogs, which may make them more suitable as breeding sites for amphibians, such as the pine barrens treefrog (Means and Moler 1979, Means 1990). Amphibians and reptiles have several options to escape fire: ascend trees, burrow in the soil, retreat to underground burrows, hide under surface cover, take shelter in water, or move to unburned areas. The patches of habitat that do not burn during a fire are important in providing refugia both during and after a fire. Surface cover, including snags, stumps, and fallen logs, often provide inadequate refugia during fires. However, the resulting burned-out stumpholes and patches of mineral soil where fallen logs have burned provide important microhabitats to some species (Means 1985, Stout et al. 1988). Unburned patches of litter provide detritivores and their predators (e.g., amphibians and lizards) with food and provide escape cover for terrestrial herpetofauna (Christman 1983). The reduced cover and altered microclimate (i.e., higher insolation, more extreme temperatures, higher surface wind, decreased relative humidity) make burned areas unsuitable for some herpetofaunal species, particularly ones preferring relatively cool, shady, and moist conditions. The reduced cover after fire also may lead to increased predation on surface-active amphibians and reptiles, particularly from avian predators. Xeric-adapted species that benefit from the more open conditions created by fire include the six-lined racerunner (Mushinsky 1985), Florida scrub lizard (Lee 1974), and gopher tortoise (Landers and Speake 1980, Auffenberg and Franz 1982, Diemer 1986). Aquatic turtles may preferentially use burned upland areas to nest (Apthorp 1993, Whitehouse 1993). Populations of species that are dependent on a litter layer or woody debris (e.g., ground skink, southeastern fivelined skink) may temporarily decline after a fire (Means and Campbell 1981, Mushinsky 1992), whereas populations of other species (e.g., Florida crowned snake) may remain unchanged (Mushinsky and Witz 1993). The timing and severity of fires may be critical to the survival of some species. Fast-moving ground fires during the growing season are usually the most beneficial. Natural fires, which normally occur from lightning strikes during the summer, differ from cool-season prescribed fires by leaving irregular, unburned patches in low spots and numerous fire shadows (e.g., behind fallen logs; Christman 1983).

12 6 FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 METHODS Regions For tabulation purposes, Florida was subdivided into the Panhandle, northern peninsula, and southern peninsula regions (Fig. 1). These regions were roughly based on herpetofaunal species distributions, which are often determined by physiography, climate, habitat, and past environmental or geophysical conditions. Species ranges were determined from existing range maps (Ashton and Ashton 1988, 1991a,b; Conant and Collins 1991), and the regions of occurrence of native species are given in the appendix. Many species reach their range limits near these regional boundaries. There is a minimum decline of 5 species between adjacent counties in the Panhandle and northern peninsula regions and of 8 species between adjacent counties in the northern peninsula and southern peninsula regions (Auth 1989). The maximum decline in mainland Florida of 14 herpetofaunal species occurs between Volusia and Brevard counties (Auth 1989). Although the herpetofauna of the Keys is different from that of the mainland, the Keys were not considered as a separate region because of their small area and the differences in herpetofaunal composition between the Upper Keys and Lower Keys (see Auffenberg 1982, Lazell 1989). Some species were not included in a region if their occurrences were peripheral. Habitat Types The habitat classification scheme and descriptions used were developed by the Florida Natural Areas Inventory (1990), wherein habitats are grouped according to their Natural Community Groups: xeric uplands, mesic uplands, rocklands, mesic flatlands, wet flatlands, seepage wetlands, floodplain wetlands, basin wetlands, and coastal (includes coastal uplands and floral-based estuarine). Xeric upland communities (sandhill, scrub, and xeric hammock) occur on very dry, deep, welldrained hills or ridges of sand with xeric-adapted vegetation. Sandhill habitat occurs on rolling hills and typically consists of widely spaced longleaf pines (Pinus palustris) with a sparse understory of deciduous oaks (e.g., turkey oak [Quercus laevis], bluejack oak [Q. incana], sand post oak [Q. margaretta]) and a fairly dense ground cover of wiregrass (Aristida spp.), other grasses, and herbs. Scrub habitat occurs on sand ridges or old dunes with deep, fine sand and consists of a closed to open canopy of sand pines (P. clausa). There is an understory of dense clumps or thickets of scrub oaks (e.g., sand live oak [Q. geminata], myrtle oak [Q. myrtifolia], Chapman s oak [Q. chapmanii], scrub oak [Q. inopina]) and other shrubs (e.g., saw palmetto [Serenoa repens], rosemary [Ceratiola ericoides], rusty lyonia [Lyonia ferruginea]), and sparse ground cover interspersed with patches of barren sand and ground lichens (Cladonia spp.). The natural fire frequency of sandhill is probably every 2 5 years, whereas scrub typically burns catastrophically once every years. In the absence of fire, these communities succeed to xeric hammock, which is either a scrubby, dense, lowcanopy, oak-dominated forest (e.g., live oak, sand live oak, laurel oak) with little understory (typical shrubs include saw palmetto and sparkleberry [Vaccinium arboreum]), or a multi-storied forest of tall trees with an open or closed canopy. Mesic upland communities (slope forest, upland hardwood forest, upland mixed forest, and upland pine forest) consist of a diverse mixture of broadleaved and needle-leaved temperate woody species on dry to moist sand hills with varying amounts of clay, silt, or organic material. Slope forest habitat is a well-developed, closed-canopied forest of upland hardwoods on steep slopes or ravines, often with seepage streams at the base. Upland hardwood forests in the northern Panhandle and upland mixed forests in the northern and central peninsula are climax communities of well-developed, closedcanopy forests of primarily hardwood trees on rolling hills. Both habitats have similar tree species composition (e.g., southern magnolia [Magnolia grandiflora], pignut hickory [Carya glabra], sweetgum [Liquidambar styraciflua], Florida maple [Acer barbatum], swamp chestnut oak [Quercus michauxii], spruce pine [Pinus glabra]), but upland hardwood forests contain some more northern species (e.g., American beech [Fagus grandifolia], shortleaf pine [P. echinata]). Upland pine forests have widely spaced longleaf pines (replaced by shortleaf and loblolly pines [P. taeda] in old field situations) and southern red oaks (Q. falcata) with few understory shrubs and a dense ground cover of grasses and herbs. This habitat occurs on the rolling clay hills of extreme northern Florida and is maintained by fires at 3- to 5-year intervals. Mesic flatlands (dry prairies, scrubby flatwoods, and mesic flatwoods) have flat, moderately welldrained, sandy substrates with a mixture of organic material, often with a hard pan. Dry prairie habitat, which normally burns every 1 4 years, is a

13 HABITAT OCCURRENCE OF FLORIDA S NATIVE AMPHIBIANS AND REPTILES Enge 7 Fig. 1. The 3 regions of Florida used to examine the geographic distribution, habitat occurrence, and relative abundance of native amphibian and reptile species.

14 8 FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 moderately drained to poorly drained, nearly treeless plain with a dense ground cover of grasses, herbs, saw palmetto, and shrubs. Scrubby flatwoods habitat is often intermingled with mesic flatwoods but occurs on slightly elevated relictual sandbars and dunes. Scrubby flatwoods are characterized by widely scattered slash (Pinus elliottii) and longleaf pines with a sparse, shrubby understory of scrub oaks and saw palmetto, and numerous areas of barren white sand. Mesic flatwoods habitat has drainage characteristics similar to dry prairies but consists of an open-canopy forest of widely spaced pine trees with little or no understory and a dense ground cover of herbs and shrubs. The most common plant associations of mesic flatwoods are longleaf pine/wiregrass/runner oak (Quercus pumila) and slash pine/gallberry (Ilex glabra)/saw palmetto. The natural fire frequencies of mesic flatwoods and scrubby flatwoods are probably every 1 8 years and 8 25 years, respectively. Wet flatlands (wet flatwoods, wet prairies, marl prairies, and hydric hammocks) are flat with poorly drained sand, marl, or limestone substrates. Wet flatwoods habitat consists of scattered slash and/or pond (Pinus serotina) pines or cabbage palms with either a thick, shrubby understory and very sparse ground cover, or a sparse understory and a dense ground cover of hydrophytic herbs and shrubs. The natural fire frequency is probably every 3 10 years, and the flatwoods are inundated for 1 month during the rainy season. Wet prairie habitat is a treeless plain with a sparse to dense ground cover of wiregrass, toothache grass (Ctenium aromaticum), maidencane (Panicum hemitomon), spikerush (Eleocharis spp.), and beakrush (Rhynchospora spp.). It is seasonally inundated or saturated for days annually and burns every 2 4 years. Marl prairie habitat has a freshwater marl substrate and consists of a sparsely vegetated, seasonal marsh on South Florida flatlands along the interface between deeper wetlands and coastal or upland communities where limestone is near the surface. Stunted pond cypress (Taxodium ascendens) or mangrove trees are often widely scattered among the sedges and grasses. Hydric hammock is a welldeveloped hardwood (e.g., diamond-leaf oak [Quercus laurifolia], red maple, swamp bay [Persea palustris], sweetbay [Magnolia virginiana], water oak [Q. nigra], southern magnolia) and cabbage palm forest with a variable understory often dominated by palms and ferns. It occurs on low, wet sites where limestone may be near the surface; the saturated soil is briefly inundated following heavy rains. Floodplain wetlands (bottomland forests, floodplain forests, floodplain swamps, strand swamps, and swales) are flat and have alluvial sand or peat substrates that are periodically flooded by rivers. Bottomland forest habitat is rarely inundated (not annually) and consists of a low-lying, closedcanopy forest of tall, straight trees (e.g., water oak, live oak, red maple, sweetgum, loblolly pine, spruce pine) with either a dense, shrubby understory and little ground cover, or an open understory and a ground cover of ferns, herbs, and grasses. Floodplain forest habitat is a hardwood (e.g., overcup oak [Quercus lyrata], water hickory [Carya aquatica], diamond-leaf oak, swamp chestnut oak) forest occurring on drier soils at slight elevations (i.e., on levees, ridges, terraces) within the floodplain, primarily along alluvial rivers in the Panhandle. The forest is flooded most years for 2 50% of the growing season, and the understory is open and parklike or dense and nearly impenetrable. Floodplain swamp habitat is flooded most of the year and consists of buttressed hydrophytic trees (e.g., bald cypress [Taxodium distichus], tupelo) and generally a very sparse understory and ground cover. Strand swamp is a shallow, forested, usually elongated depression or channel dominated by bald cypress. The peat and sand soils, which occur over limestone, are inundated days annually. Swale habitat is a marsh situated in a broad, shallow channel with flowing water (primarily in the Everglades-Big Cypress region) and is characterized by emergent grasses (especially sawgrass [Cladium jamaicense]), sedges, and herbs up to 3 m tall. Limestone underlies the peat or sand soil, which usually has a sheet of water flowing over it 250 days per year. Light ground fires typically occur every 1 5 years. Basin wetlands (basin marshes, basin swamps, depression marshes, and dome swamps) consist of wetland woody and/or herbaceous vegetation in shallow, closed (except during high water) basins with peat or sand substrates that are usually inundated. Basin marsh habitat is an herbaceous or shrubby wetland situated in a relatively large, irregularly shaped basin (formerly a shallow lake) with a hydroperiod around 200 days per year and fires every 1 10 years. Basin swamp habitat is a forest of hydrophytic trees (e.g., blackgum [Nyssa biflora], pond cypress, bays [Persea spp.], slash pine) and shrubs situated in a relatively large, irregularly shaped basin not associated with rivers. It has a hydroperiod of days annually and fires every years. Depression marsh habitat consists of herbaceous vegetation in a shallow,

15 HABITAT OCCURRENCE OF FLORIDA S NATIVE AMPHIBIANS AND REPTILES Enge 9 usually rounded depression in sand substrate. It has a variable hydroperiod (drying most years and sometimes inundated <50 days) and burns frequently. Dome swamp is a shallow, forested (predominately pond cypress, blackgum, and slash pine), usually circular depression that generally presents a domed profile because smaller trees grow in the shallower water along the periphery. It typically occurs in sandy flatwoods and in karst areas and has a hydroperiod of days annually, with water remaining the longest in the deeper center of the dome. Fires occur relatively frequently around the margins of domes and occasionally burn through the interior. Seepage wetlands (baygall and seepage slope) occur on sloped or flat sands or peat with high moisture levels maintained by downslope seepage. Baygall habitat is a dense, tall forest of generally straight-boled hardwoods (e.g., sweetbay, swamp bay, loblolly bay [Gordonia lasianthus]) with a mostly open understory of shrubs and ferns and a floor carpeted by sphagnum moss (Sphagnum spp.), often interlaced with convoluted tree roots. Baygalls occur in peat-filled seepage depressions at the base of sandy slopes, at the edge of floodplains, or in flat areas with high lowland water tables. Seepage slope habitat is a shrub thicket (shrub bog) or boggy meadow (herb bog) on, or at the base of, a slope where downslope seepage maintains saturated, but rarely inundated, soil conditions (small pools and rivulets are common, however). Typical plants include pond/slash/longleaf pine, titi (Cyrilla racemiflora), fetterbush (Lyonia lucida), myrtleleaved holly (Ilex myrtifolia), and insectivorous plants. Herb bogs are maintained by fires at least every 5 years, whereas shrub bogs only burn every years. Rockland communities (pine rockland and rockland hammock) are either low, generally flat limestone outcrops with tropical vegetation, or are exposed limestone (through karst activities) with tropical or temperate vegetation. Pine rockland habitat is an open-canopy forest of slash pines with a patchy understory of tropical and temperate shrubs and palms, and a variable ground cover of grasses and herbs. It now occurs primarily along the Miami Rock Ridge and Lower Florida Keys and is maintained by fires every 3 10 years. In the absence of fire, it eventually succeeds to rockland hammock. Rockland hammock habitat is a closed-canopy hardwood forest of diverse tree species, such as live oak, gumbo limbo (Bursera simaruba), wild tamarind (Lysiloma bahamense), stoppers (Eugenia spp.), pigeon plum (Coccoloba diversifolia), false mastic (Mastichodendron foetidissimum), and poisonwood (Metopium toxiferum), that is often protected from fire by surrounding wetlands. Of the 4 coastal habitats considered, 2 habitats (coastal strand and maritime hammock) are included under coastal upland communities found along shorelines subject to high-energy waves. Coastal strand habitat is a dense thicket of salt-tolerant shrubs (e.g., saw palmetto, sand live oak, cabbage palm, myrtle oak, yaupon [Ilex vomitoria], sea grape [Coccoloba uvifera], Spanish bayonet [Yucca aloifolia]) and prickly pear (Opuntia stricta) occurring on stabilized, wind-deposited dunes. The shrubs, which are often dwarfed and pruned by salt spray-laden winds, are located usually between the beach dunes and maritime hammock or scrub. Maritime hammock is a narrow band of hardwood (e.g., live oak, cabbage palm, red bay [Persea borbonia]) forest with a dense, wind-pruned canopy that occurs on old coastal dunes. The other 2 habitats (tidal marsh and tidal swamp) are included under floral-based estuarine communities and are found along coastlines of low wave energy. Tidal marsh is an expanse of grasses, rushes, and sedges along coastlines and river mouths. Black needlerush (Juncus roemarianus), smooth cordgrass (Spartina alterniflora), and sawgrass often form dense, uniform stands, depending on tide levels and elevation. Tidal swamp is a dense, low forest of buttonwood (Conocarpus erecta) and red (Rhizophora mangle), black (Avicennia germinans), and/or white (Laguncularia racemosa) mangroves occurring along relatively flat, intertidal and supratidal shorelines in southern Florida. Disturbed cover types (agricultural areas and grasslands, urban and barren lands, shrub and brush lands, exotic plant communities) represent ca. 42% of Florida s present landscape (Kautz et al. 1993). The herpetofauna of these human-altered habitats was not characterized, but populations of some species may be higher in ruderal habitats (e.g., sugar cane fields) than in natural habitats (Wilson and Porras 1983, Enge 1993). Relative Abundance Relative abundances of species within a taxonomic suborder or order (i.e., anuran, salamander, turtle, crocodilian, lizard, snake) are designated in the tables as common, uncommon, or rare. The Florida worm lizard, an amphisbaenid, is included under lizards. Relative abundance includes both a species occurrence and a population density component, but the categories of relative abundance cannot be quantified and vary among species. A species designated as common in a particular habitat

16 10 FLORIDA GAME AND FRESH WATER FISH COMMISSION TECHNICAL REPORT NO. 16 would be expected to occur on most tracts of land containing that habitat and to be present at population densities higher or equivalent to other habitats where it is listed as common. An uncommon species in a particular habitat would occur on only some tracts of land containing that habitat and would usually occur at population densities lower than in habitats where it is listed as common. However, the population density of an uncommon species may temporarily be high, such as during an influx of newly metamorphosed anurans, aggregation of breeding adults, or periods of flooding or drought. Rare species are infrequently recorded in the habitat being considered and usually occur at low population densities when present. When the range of a species does not encompass an entire region, the relative abundance given refers to habitats occurring within the range of the species. Relative abundances cannot be compared between taxonomic orders. For example, an anuran species listed as uncommon in a certain habitat may occur at higher population densities than a snake species listed as common. If wetlands are present in xeric or mesic uplands, the relative abundance of many amphibian and semiaquatic reptile species will be greater than indicated in the tables. In addition, aquatic turtles and snakes that do not appear in the tables may utilize these upland habitats to deposit eggs (Semlitsch et al. 1988, Mushinsky and Wilson 1992). Aquatic species (e.g., pig frog, eastern mud snake, black swamp snake) that transiently use upland habitats while traveling between wetlands are not included under upland habitats in the tables unless they were frequently recorded by several driftfence surveys (e.g., chicken turtle in sandhill habitat). The occurrence and relative population density of species in habitats are based on published and unpublished drift-fence surveys; personal observations and communications; species accounts from field guides (e.g., Mount 1975; Ashton and Ashton 1988, 1991a,b; Conant and Collins 1991), books (e.g., Lazell 1989, Moler 1992), and scientific studies (e.g., Florida Game and Fresh Water Fish Commission 1976a,b); and the aforementioned habitat occurrence lists or tables. In cases where information on a habitat was scarce or absent, educated guesses were made on the occurrence and abundance of species based on information from similar habitats and species habitat requirements. When assigning relative abundance values to species, I tried to take into account the biases associated with drift-fence data, which tend to underrepresent arboreal species (e.g., treefrogs, rough green snakes, rat snakes), fossorial species (e.g., Florida worm lizard, sand skink), and large turtles and snakes (Campbell and Christman 1982a; Gibbons and Semlitsch 1982; Dodd 1991). There are also biases among drift-fence studies because of differences in methodology, including array design, fencing material, number and types of traps, and trapping seasons and duration. Several published drift-fence studies provided information on multiple habitats in the Panhandle (U.S. Fish and Wildlife Service 1980), northern Florida (Wood 1993), central Florida (Florida Game and Fresh Water Fish Commission 1976a; Joiner and Godwin 1992a,b,c,d,e), and southern Florida (Dalrymple 1988, Timmerman et al. 1994), but most studies focused on only 1 or 2 habitats. Numerous drift-fence surveys have been conducted in flatwoods (U.S. Fish and Wildlife Service 1978, Labisky et al. 1983, White 1983, Enge and Marion 1986, Palis and Jensen 1995), sandhill (U.S. Fish and Wildlife Service 1978, Campbell and Christman 1982b, Mushinsky 1985, Dodd and Charest 1988, Stout et al. 1988, Dodd 1992, Franz et al. 1995, Palis and Jensen 1995, Stout and Corey 1995), and scrub (Christman et al. 1979, Campbell and Christman 1982b, Christman 1988, Greenberg et al. 1994, Mushinsky and McCoy 1995) habitats. Some published literature exists on the herpetofauna of floodplain wetland (Wharton et al. 1982) and coastal communities (Neill 1958, Blaney 1971, Odum et al. 1982). Some published drift-fence data exist for tidal marsh (Joiner and Godwin 1992a,Wood 1993), dome swamp (Vickers et al. 1985), upland pine and upland hardwood forests (Means and Campbell 1981), upland hardwood or slope forest (Enge 1997b), upland mixed forest (Miller and Schaefer 1993), and bottomland forest (Means and Studenroth 1994). Information on snake relative abundance in swale habitat came partly from a roadcruising survey (Bernardino and Dalrymple 1992).