PLEISTOCENE AMPHIBIANS AND REPTILES FROM THE LEISEY SHELL PIT, HILLSBOROUGH COUNTY, FLORIDA Peter A. Meylan 1 ABSTRACT Amphibians and reptiles are common among the vertebrate remains recovered from the Leisey Shell Pit This study was undertaken to document the diversity ofthese groups and is not a detailed account of all available Leisey specimens. For the sake of completeness, an effort was made to identify every element that could add to the taxonomic list. A total of 20 reptiles and 4 amphibians has been identified from the late early Irvingtonian Iiisey 1 A and 3A localities. The fossils from the adjacent Leisey 2, which is chronologically mixed (early and late Pleistocene), includes three additional tan, and these are discussed in the systematic accounts. None of the Leisey sites possesses a diverse herpetofauna, especially compared to Inglis 1 A The Leisey l A herpetofauna is dominated by three tan-Alligator, Trachemys, and Hesperotestudo--while Lnisey 3A is dominated by aquatic natricine snakes. RESUMEN Es comun encontrar restos de anfibios y reptiles entre los restos de vertebrados recuperados del dep6sito de conchuelas de Leisey. Este estudio se Ilev6 a cabo con el objeto de documentar la diversidad de anfibios y reptiles en estos dep6sitos, sin ser una detallada cuenta de todos los especimenes disponibles para estos depdsitos. A bien de ser lo inds complete posible, se desarrol16 un esfuerzo por identificar cada elemento que pudiera acrecentar la lista taxon6mica. Se identific6 un total de 20 reptiles y 4 anfibios, provenientes de las localidades Leisey lA y 34 pertenecientes al Irvingtoniano temprano. Incluidos en los fbsiles encontrados en el dep6sito adyacente 1£isey 2, que se encuentran mezclados cronol6gicamente (Pleistoceno temprano y tradio), se identificaron tres taxones adicionales, siendo 6stos discutidos crt la relaci6n sistemdtica. Ninguno de los sitios Leisey posee una herpetofauna diversa, en especial cuando se la compara con Inglis l A La herpetofauna de Leisey 1 A es dominada por tres taxones-Alligator, Trachemys yHesperotestudo-mientras que Leisey 3A es dominada por serpientes acuiticas natricinas. 1 The author is an Assistant Professor, Natural Sciences Conegium, Eckerd College, St Petmbug, F[ 33711, and Research Associate, Department of Vertebrate Paleontology, Amencan Musemn of Nattual History, Central Pwk West at 79th St, New York. NY 10024. MEYLAN, P.A 1995. Pleistocene amphibians and reptiles from the Leisey Shell Pit Hillsborough County, Florida. Bull. Florida Mus. NaL Hist. 37 PL 1(9):273-297. 274 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, Fr. I, No. 9 INTRODUCTION Although the fossil record of reptiles and amphibians in Florida has been broadly outlined during the last fifty years of study, several important temporal gaps remain. The record is basically a Neogene one. There are a few scattered records of reptiles from the Eocene limestones that underlie most of Florida and a single Oligocene herpetofauna (I-75) that remains undescribed. The early Miocene fauna is well understood on the basis of Thomas Farm and a series of smaller sites. However, the most serious gap in the herpetofaunal record of Florida exists between those occurrences and the Clarendonian (Love Bone Bed). The Love Bone Bed and a series of other sites document the reptiles and amphibians through the later Miocene into the Pliocene, but there is a second major gap in the record in the Blancan. The herpetofauna of the earliest Pleistocene is well documented by Inglis IA, but until now there has existed a third significant gap in the record, between the earliest Imingtonian Inglis IA fauna and the numerous Rancholabrean faunas with their abundant reptile and amphibian remains. Thus, the Leisey herpetofauna (late early Irvingtonian, Table 1) fills an important gap in the herpetological fossil record for the State of Florida and the southeastern United States. ACKNOWLEDGEMENTS My participation in the study of Lciscy material has been encouraged and facilitated by Richard C. Hulbert Jr., Gary S. Morga* and S. David Webb. I am especially grateful to Gary for his patient help during numerous visits to the Florida Museum of Natural History and with arrangements for figures. I am also grateful to Linda D. Chandler for drawing figure 1 and Rhoda J. Bryant for typing and editing the manuscripL METHODS The diagnostic features used to identify material have been cited in the species accounts. All measurements are given in millimeters. Terminology for snake vertebrae follows Auffenberg (1963b) and Meylan (1982). Terminology for anuran ilia follows Auffenberg (1956), Tihen (1962), and Lynch (1966); that for the turtle shell follows Zangerl (1969). All specimen numbers refer to the Florida Museum of Natural History (UF) Vertebrate Paleontology Collection. Abbreviations are: CL, centrum length; NAW, neural arch Width; NH neural spine height; NL neural spine length; POPR, length of vertebra from front edge of prezygapophes to rear edge of postzygapophyses; PRI)R, width of vertebra to rear edge of prezyapophyses; SCL, straight carapace length; SVL, snout-vent length. MEYLAN: AMPHIBIANS & REPTILES FROM LEISEY SHELL Prr 275 Table 1. Amphibians and reptiles from Leisey 14 Leisey 34 and Leisey 2. Taxa followed by one asterisk (*) are earliest known occurrences at Leisey lA or JA Those followed bytwo asterisks (**) are absent from the modern fn,ina of Hillsborough Co., Florida. Taxa followed by three asterisks (***) are found only in Leisey 3 (see Morgan and Hulbert this vol.). Occurrence Ininglonian (mixed age?) Species Leisey lA Lzisey 3 A Lcisey 2 Stren sp. X X X Siren lacertina X X Amphiuma sp. X Bufo terrestris X cf Rana sp. X X Alligator mississippiensi, X Ophijaurus compressus* X 6. Tropidophi:*, ** X X X X X X X X X X X X X X X Nerodia sp X d 77,amnophu sp. X X ct Nerodia fasciata X Regina sp. cf. Regina alleni* cf. Elaphe obsoleta X Lampropeltis getulus X Coluber or Masticophia Drymarchon corals X Farancia cf. Agkistrodon piscivorous X d Crotalus adamanteus X X Sistrums mdianus X 4 Carena caretta X Chelonia mydas*** Chetydra serpentina*- Macroclemys temminck ** X X Hesperotestudo crassisculata" X X Hesperotestudo mlynarskii** Gopherus polyphemus Terrapene carolina X Trachemys scripta X X X X X X X X X X Pseudemys sp. Deirochelys reticularia X Kinosternon sp. X Apaloneferox X X 276 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, PT. I. No. 9 SYSTEMATIC PALEONTOLOGY URODELA SIRENIDAE Siren lacertina Linneaus 1766 Referred Specimens.- Leisey 2: UF 125089, four vertebrae; Leisey 3A: UF 124685, four vertebrae. The vertebrae are of the form typical of the Sirenidae (Goin and Auffenberg 1955). The centra are amphicoelus, and they have well developed neural spines that are Y-shaped in dorsal view. Ventrally they have well developed central keels with large foramina on either side. These large sirenid vertebrae are referred to Siren lacertina, the larger of the two currently recognized living species, on the basis of size alone. Siren sp. Refemd Specimens.- Leisey lA: UF 84012, 124666, 124688, 124689, four vertebrae; Leisey 2: UF 125090, 3 vertebrae; Leisey 3A: UF 124660, 14 vertebrae; UF 124661, 6 vertebrae; and UF 124665, an atlas. These vertebrae represent a small sirenid salamander. The smallest Leisey sirenid vertebrae have nearly straight lower margins to their centra. Goin and Auffenberg (1955) suggested that Siren can best be distinguished from Pseudobranchus by the shape of the lower margin of the centrum. In Siren intermedia and S. lacertina this margin is straight or nearly so. In Pseudobranchus it is distinctly concave. As pointed out by Goin and Auffenberg (1955), the vertebrae of the extant Siren species can be distinguished only on the basis of the very large sizc of some S lacertina. The small size of these vertebrae makes it impossible to identify them to species. AMPHIUMIDAE Amphiuma sp. Referred Specimen.- Leisey 2: UF 124088, one vertebra. This vertebra is much like those referred to the Sirenidae, being amphicoelus with well developed transverse processes and neural spines. Unlike that of sirenid vertebrae, the neural spine of Amphiuma divides distally and is thus T-shaped rather than Y-shaped in dorsal view. MEYLAN: AMPHIBIANS & REPTILES FROM LEISEY SHELL PIT 277 ANURA BUFONIDAE Bufo cf. B. terrestris (Bonnaterre 1789) Referred Specimens.- Leisey lA: UF 80773 and 83240, two ilia; Leisey 3A: UF 124664, ilium. These ilia lack a dorsal crest but have a well developed but low dorsal prominence that is directed dorsally. There is no protuberance on the dorsal prominence. This combination of characters distinguishes Bufb ilia from those of other anurans. In Bufb the ilial prominence is on the dorsal edge of the ilium, and although it may have a roughened area on it, it lacks a protuberance. These ilia strongly suggest those of Bufb terrestris in having a low mal prominence with the anterior and posterior slopes subequal (Tihen 1962). RANIDAE cf. Rana Linneaus 1758 Refemd Specimens.- I.eisey lA: UF 124690, ilium; Ikisey JA: UF124662 and 124663, partial right ilia. These fragmentary ilia show the remains of the large dorsal iliac crest typical of members of this genus. These specimens are too poorly preserved to allowfurther identification. CROCODILIA ALLIGATORIDAE Alligator mississippiensis gaudin 1803) Referred Specimens.- Leisey lA: UF 68908, 68909, 80023, 80143, 80619, 80769, 81136, and 81137, articulars and/or surangulars; UF 68907, 80302, and 80669, dentaries; UF 80075 and 80352, premaxillaries; UF 64388, 68912, and 68942, maxillaries; UF 68943, two parietals; UF 84017, squamosal; UF 81092, jugal; UF 84016, lacrimal; plus many limbs, vertebrae, and osteoderms; Lcisey 3A: UF 102522, most of a subadult skull and associated right and left mandibles. Alligator mississippiensis is a very common component of the Leisey fauna. The material listed here is only a sample of the diversity of the isolated elements available. The material referred to this taxon was carefully compared to Croco*lus acutus. There are historical records for C acutus less than 50 miles south of Leisey along the Gulf coast in Sarasota County (LeBuff 1957), as well as an unconfirmed Pleistocene record from Citrus County, Florida (Neill 1971). In every case the Lcisey elements were identical to A/ligator and differed significantly from C acutus. 278 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, PT. I, No. 9 All A lligator dentaries from Leisey have very short symphyses, spanning a distance ofonly two to three alveoli. The symphysis of C acutus is longer, usually four or five alveoli. The surangulars include the posterior portion of a very large mandibular fenestra. In C acutus this fenestra is quite small. In the available articulars the foramen nervi chorda tympani is about one-quarter of the way across the articular condyle as in A/ligator and not on the medial edge of the articular as in Croco*lus acutus. The premaxillae are short and have essentially transverse sutures for the maxillae. These are long and angular in C acutus. Furthermore, in none of these specimens is there evidence of the large dentary teeth of C acutus that pierce the premaxillae. The jugal is more expanded lateral to the orbit than in C acutus and is also concave to flat rather than convex. The lacrimal bone has a broadly opening lacrimal duct, unlike the acute deep opening of this duct in C acutus. SQUAMATA LACERTILIA ANGUWAE Ophisaurus compressus Cope 1898 Referred Specimens.- Leisey lA: UF 84013, a thoracic vertebra; Leisey 3A: UF 124683, a caudal vertebra. The single thoracic vertebra has a very flat centrum typical of the family Anguidae. The centrum is 2.8 mm long and 1.8 mm wide. The length to width ratio of the vertebral centrum will successfully differentiate the three well known living species of Ophisaurus from North America (Meylan 1982, fig. 4; 0. mimicus, Palmer 1987, not included). The length to width ratio of 1.75 is beyond two standard errors of the mean for the same ratio for 66 body vertebrae of Ophisaurus attenuatus and for 99 body vertebrae of 0. ventralis. It is quite close to the mean for living 0. compressus (1.74). The caudal vertebrae of 0. compressus can be distinguished from those of all North American congeners, including the recently described 0. mimicus, by the presence of a fused fracture plane or autotomy septum. A fused fracture plane is clearly preserved in the Leisey caudal. SERPENTES BOIDAE TROPIDOPHIINAE cf. Tropidophis Bibron 1840 Figure 1, A-E Referred Specimen.- Leisey 3A: UF 124637, one vertebra. MEYLAN: AMPHIBIANS & REPTILES FROM LEISEY SHELL PIT 279 A single vertebra from the Leisey fauna (Fig. 1) apparently represents a small boid snake of a group that previously is unrecorded from North America north of Mexico. UF 124637 is a small vertebra, with a square centrum (CL=3.5, NAW=3.6). It is wider across the prezygapophyses (PRPR=5.2) than its length from pre- to postzygapophyses (POPR=4.3). It has a short neural spine that covers only the posterior half of the neural arch and is very low (NL= 1.6, NH=O.6). The neural arch is slightly depressed; the centrum posteriorly constricted. Accessory processes were present but apparently poorly developed. A weakly projecting hypapophysis extends ventrally from this thoracic vertebra about 0.5 mm along the posterior half of the length of the centrum. The general shape of the vertebra, particularly the width, and the short neural spine suggest the family Boidae. The small size and development of the hypapophysis suggest the subfamily Tropidophiinae (Bogert 1964 1968b; Underwood 1976). The vertebra is most similar to those of Exihboa and some species of Tropidophis, especially T haetianus and T. feicki. Among described North American fossils, the Leisey vertebra resembles Huberophis from the Eocene of Georgia (Holman 1977), It differs from this species in having a thinner neural spine and a better defined haemal keel. However, the vertebrae are remarkably similar in lateral view. The Leisey specimen also shows some likeness to members of the genus Geringophis (Holman 1976, 1982). Two members of this genus of small boids have ventrally expanded haemal keels and neural spines limited to the posterior half of the neural arch. These genera have been placed in the subfamily Erycinae. However, it is possible that they belong to the Tropidophiinae, not the Erycinae. COLUBRIDAE NATRICINAE cf. Thamnophis Fitzinger 1843 Referred Specimens.- Leisey 1: UF 84014, 124691, vertebrae; Leisey 2A: UF 125097, two vertebrae; Leisey 3A: UF 124642, one vertebra, UF 124643, two vertebrae, UF 124644, three vertebrae, UF 124680, five vertebrae. These small vertebrae are long and narrow (the centra are much longer than wide). They have well developed hypapophyses with a long, narrow base. This hypapophysis shape suggests a natricine rather than a viperid snake. The very narrow shape of the centrum suggests a garter snake, rather than a water snake. Nerodia sp. Referred Specimens.- Leisey 2: UF 125096, one vertebra; Ikisey 3A: UF 124677, three vertebrae. 280 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, Fr. I, No. 9 Nerodia cf. N. fasciata GAnneaus 1166) Referred Specimens.- Leisey 2: UF 125096, one vertebra; Leisey 3A: UF 124646, two vertebrae; UF 124647, seven vertebrae. These relatively large vertebrae have well-developed, broad hypapophyses and centra that are nearly as wide as long. They appear too large and square to represent Thamnophis or Regina. The neural spine is about as high as long which is typical of the Nerodiajasciata/sipedon complex (see Meylan 1982). Regina sp. Referred Specimens.- Leisey 3A: UF 124640, 2 vertebrae, UF 124679, 4 vertebrae. Regina cf. R. alleni (Garman 1814) Referred Specimeni- Leisey 3A: UF 124641, one vertebra, UF 124678, 2 vertebrae. Vertebrae representing members of the genus Regina can be recognized by their small adult size (30-40 cm SVL), short sigmoid-shaped hypapophysis and neural spines that overhang anteriorly and posteriorly (Meylan 1982). These vertebrae are larger than those of Storeria and Virginia, and not as long and narrow as 77,amnophis. Regina alleni is the only member of the genus with long thin accessory processes of the type seen in UF 124641. Another of the vertebrae has a gutter on the dorsal edge of the neural spine, as described by Auffenberg (1963b) for this species. LAMPROPELTINAE cf. Elaphe obso/eta (Say 1823) Referred Specimens.- Leisey lA: UF 83082, a mid-body vertebra; Leisey 3A: UF 124681, one vertebra. These are essentially perfect mid-dorsal vertebra with square centra and neural arches only slightly wider than long across the zygapophyses. The neural spine is tall and not significantly undercut anteriorly or posteriorly. The haemal keet is narrow and straight. Subcentral ridges are moderately developed. Accessory processes are well developed but not longer than the prezygapophyseal facet. The zygosphene is nearly as wide as the neural arch; it is straight in dorsal view and roundly convex in anterior view. There are no epizygapophyseal spines. MEYLAN: AMPHIBIANS & REPTILES FROM LEISEY SHELL PIT 281 J A B r-»[~A - 1 f/ \/C r D Figure 1. UF 124637, a thoracic vertebra of d Tropidophis from Leisey 34 in (A) dorsal, (B) ventral, (C) lateral, and (D) anterior views. Scale bar = 5 mm. The absence of a hypapophysis, square vertebral centrum, lack of epizygapophyseal spines, and short laterally directed accessory processes suggest that these vertebrae represent one of the constricting colubri(is sometimes recognized as the subfamily Lampropeltinae (Smith et al. 1977; Meylan 1982) or tribe Lampropeltini (Dowling and Duellman 1974; Dowling and Fries 1987). The absence of well developed subcentral ridges and the presence of tall neural spines suggest that Lampropeltis getulus is not represented. The large size of the vertebrae suggests that one of the other smaller North American Lampropeltis is not represented. Although they could represent Elaphe guttata or Pituophis melanolucas, features of the zygosphene and neural spine (Meylan 1982) are most like Elaphe obsoleto, to which these vertebrae are tentatively referred. 282 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, Fr. L No. 9 Lampropeltis gefula (Linneaus 1766) Referred Specimens.- Leisey 2. UF 125095, two vertebrae; Leisey 3A: UF 124648, one vertebra. The vertebrae are referred to the Lampropeltinae based on criteria discussed under E/aphe obsoleta. The large size and very distinct subcentral ridges of these vertebrae allow their assignment to this species. COLUBRINAE Coluber sp. or Masticophis sp. Referred Specimens.- Leisey 3A: UF 124638, two vertebrae, UF 124639, one vertebra, UF 124682, one vertebra. These vertebrae are much longer than wide and lack hypapophyses. They have long narrow haemal keels and may represent either of these two genera. Drymarchon corals (Fitzinger 1843) Referred Specimens.- Leisey 2: UF 125093, two vertebrae. The vertebrae of Dc=archon are immediately recognizable by their large size and the beveled front edge of the neural spine that occurs in most populations (Meylan 1982). XENODONTINAE Farancia sp. Refemd Specimens.- Ikisey JA: UF 124645, two vertebrae. These vertebrae are typical of Farancia, being square across the zygapophyses with a markedly depressed neural arch. The accessory processes are short and stout and laterally directed. The haemal keel is strongly developed. VIPERIDAE Viperidae gen. et sp. indet. Referred Specimens.- Leisey 2: UF 125092, one fang; Leisey 3A: UF 124651, three vertebrae. MEYLAN: AMPHIBIANS & REMILES FROM LEISEY SHELL Prr 283 cf. Agkistrodon piscivorous (Lac€pue 1189) Referred Specimens.- Leisey lA: UF 86908, five vertebrae; UF 124635, one vertebra, UF 124669, two vertebrae; Leisey 3A: UF 124636, one vertebra. Agkistrodon piscivorus (Lac6~de 1189) Referred Specimens.- Leisey lA: UF 124692, one vertebra; Leisey 3A: UF 124687, one vertebra. cf. Crotalus adamanteus Beauvois 1799 Referred Specimens.- Leisey lA: UF 84015, 86962, and 124670, three vertebrae; Leisey 2: UF 125091, three vertebrae. All the above viperid vertebrae are large with thick and strongly developed hypapophyses. The centra are slightly wider than long, and the neural arches are much wider than long measured across the zygapophyses. Such wide vertebrae with thick strongly developed hypapophyses that are not expanded at the base are typical of the Viperidae. Holman (1963) suggested that the absence of pits lateral to the cotyle can be used to assign individual vertebrae to Crotalus rather than Agkistrodon, but there is some variation in this feature (Meylan 1982). Vertebrae with deep pits are tentatively referred to Agkistrodon; those without deep pits are tentatively referred to Crotalus. The fang is that of a solenoglyphus snake with only proximal and distal openings. Sistrurus miliarius (Unneaus 1166) Referred Specimens.- Leisey 3A: UF 124676, two vertebrae. These vertebrae represent very small adult viperid snakes. The neural canal is small relative to cotyle diameter, and they have a long, straight narrow hypapophysis. TESTUDINES CHELONHDAE cf. Caretta caretta (Unneaus 1158) Referred Specimen.- Leisey lA: UF 82717, a left hyoplastron found in the upper shell bed just above the main bone layer. 284 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, Yr. I, No. 9 This single left hyoplastron represents a cheloniid marine turtle with a very complete plastron. The dermal plastral callosity covers the medial processes of this element all the way to the midline. Based on the configuration of the hyoplastrog the entoplastron was small and elongate, and the hyoplastra met on the midline behind it. The element is 260 mm wide, which is probably too large to be Lepidoche/ys; it is from a more completely ossified plastron than those of either Cheionia or Eretmochelys. As UF 82717 was derived from a stratigraphically higher horizon than the main bone layer, it is of younger age and strictly spenking Coretta should not be considered a member of the Leisey Shell Pit local fauna. Its probable age is middle to late Pleistocene (Morgan and Hulbert this volume). Chelonia mydas (Linncaus 1758) Referred Specimens.- Leisey 3: UF 124674, a dentary. The referred dentary has a moderately wide triturating surface with a symphyseal ridge that joins a broad, well developed lingual ridge. At their junction is a large triangular tubercle. Che/onia, Lepidochelys, and Eretmochelys may have a large pyramidal tubercle on the symphysis, but only in Che/onia are there deep pits on either side of a well developed symphyseal ridge, a narrow shelf medial to the labial ridge, and a weakly developed symphyseal hook. The dentary represents a small individual (* 30 cm SVL) of the size known to frequent inshore waters of Florida's Gulf coast (Carr and Caldwell 1956). CHELYDRIDAE Chelydra serpentina (Linneaus 1758) Referred Specimens.- Leisey 3: UF 135677, a neural. The single neural is half-moon-shaped, very wide (almost four times as wide as long), with a very weakly developed midline keel. The neurals of snapping turtles are particularly wide and often lack the regular 6- or 8-sided shapes seen in most turtles. The midline keel is developed too weakly for this element to represent Macroclemys. Macroclemys cf. M. temmincki groost 1835) Referred Specimens.- Leisey lA: UF 84005, a costal fragment; and UF 81198, a peripheral; Leisey 18: UF 124675, a right dentary; Leisey 2: UF 125099, partial associated shell and skeleton; and uncatalogued parts of two or three individuals; Leisey 3A: UF 116093, a left dentary. MEYLAN: AMPHIBIANS & REPTILES FROM LEISEY SHELL PIT 285 The two dentaries are massive and broad, the two halves of the jaw form a broad symphysis that sweeps gradually upward into a strong, sharp hook. The hook in the dentary of Chelydra is much lower (Gaffney 1975, figs. 4 and 5). The costal is incomplete but shows a triangular, thickened region lateral to its medial edge. This is part of the strong, parasagittal carapacial keels that are typical of Macroclemys. The peripheral is about 40 mm long and has a deep pit for insertion of a large rib end. It could represent Chely#a or Macroclemys. The recent discovery of a snapping turtle in the Blancan of Florida with a Chelydra-like skull but a strongly keeled Macroclemys-like shell (Haile 7C; Hulbert et al. 1989) suggests that any assignment of a snapping turtle to genus on the basis of keeled shell fragments alone must be considered tentative. The Leisey records for Macroclemys lie far south of its current distribution (Fig. 3; see Discussion). TESTUDINIDAE Hesperotestudo Williams 1950 Bramble (1971) questioned the naturalness of the genus Geochelone as used by Williams (1950), Loveridge and Williams (1957), and Auffenberg (1974). He felt that it included taxa that were ancestral to gopher tortoises (Gopherus) as well as living Geochelone. Further studies of these tan (Crumly 1982, 1984) support Bramble's contention. Crumly (1984) identified the presence of a lateral surangular process as an important derived feature suggesting monophyly of advanced tortoises, including all living species of Geochelone (the type is the living species, Geochelone elegans Schopff 1792) and such genera as Testudo, Indotestudo, and Malacochersus. This process is unknown among North American fossil forms. The current trend towards the restriction of scientific names to monophyletic groups would suggest that the application of the name Geochelone to any North American tortoise would be inappropriate, since it would make Geochelone paraphyletic. Bramble (1971) and Auffenberg (1974) have clearly indicated that the name Hesperotestudo (usually used as a subgenus) applies to those North American tortoises in which the dermal ossicles are extremely well developed and the proximal caudal vertebrae have expanded lateral processes. Preston (1979) previously applied this name at the generic level in his review of late Pleistocene turtles from the mid-continental United States. The North American forms currently referred to Geochelone should be called Hesperotestudo in order to promote a better understanding of the interrelationships among land tortoises. It has long been suggested that therd are two lineages of Hesperotestudo in the Pleistocene of Florida. These lineages have been separated on the basis of size, the development of the caudal buckler, the visibility of growth rings, depth of the anal notch, and other details of shell morphology (Auffenberg 19634 1988). 286 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL. 37, PT. I, No. 9 Using these criteria, these large (subgenus Caudochelys) and small (subgenus Hesperotestudo) lines can be recognized in the Leisey fauna. Hesperotestudo (Caudochelys) crassiscutata (Leidy 1889) Figure 2 Referred Specimens.- Leisey lA: UF 64395, parts of five costals; UF 65005, peripheral; 80593, much of a plastron (510 mm midline length); - UF 84300, epiplastron; UF 65052, 65053, 65054, 68928, partial plastra; UF 80676, entoplastron; UF 80593, 83602, 84007, 84066, 84299, 84301, 84311, isolated xiphiplastra; UF 81925, nuchal; UF 64396, 68933, 81480, 68919, femora; UF 84024, 84025, armor; UF 64398, 65051, 88126 ,caudal vertebrae; Leisey 2: UF 125065, left epiplastron; UF 125064, neural; Leisey 3: UF 130008, proximal humerus; Leisey 3A: UF 102513, a very large humerus. The name Hesperotestudo crassiscutata is based on Testudo crassiscutata Leidy from the Pleistocene of Florida. Although this lineage of large land tortoises has been traced from the Miocene through the Pleistocene of Florida (Auffenberg 1963a), no species other than H. crassiscutata has been described from the Pleistocene of Florida, and it seems likely that this is the species represented by most of the Hesperotestudo material in the fauna. Hespertestudo cf. H. mlynarskii (Autrenberg 1988) Referred Specimens.- Leisey lA: UF 80675, much of a plastron with parts of a carapace; UF 80897, 81132, epiplastra; UF 80351, entoplastron; UF 80166, 80361, hyoplastra; UF 83090, xiphiplastron; UF 80461, nuchal; UF 83091, pygal; Leisey 2: UF 142235, left epiplastron (may represent this species or H. incisa). Hesperotestudo mlynarskii was recently described from the late Irvingtonian Coleman 2A locality in Sumter County, Florida (Auffenberg 1988). It is considered to differ from Rancholabrean H. incisa in part by having a shallower anal notch (Auffenberg 1988, fig. 3). A single small xiphiplastron (UF 83090) falls in the range of variation known for H. mlynarskii and well below that known for H. incisa. It is based on this single element that the Leisey representative of the turgida-complex tortoise is tentatively referred to this recently described form. Gopherus polyphemus (Daudin 1803) Referred Specimens.- Leisey lA: UF 80559, 80796, and 81485, epiplastra; UF 64394, 68939, 80363, 80458, 81007, 81063, xiphiplastra; UF 68900, nuchal; UF 82605, two peripherals; UF 84302, pygal. MEYLAN: AMPHIBIANS & REPTILES FROM LEISEY SHELL PIT 287 A B C 6,4 Figure 2. UF 64398, 65051, and 88126, caudal vertebrae from a large Hesperotestudo cf H. crassiscutata. These vertebrae all show the broad flat transverse processes of the caudal vertebrae which appear to be diagnostic for the genus. Scale bar = 30 mm for A and B, 20 mm for C. The nuchal of Gopherus is the most immediately identifiable element. This genus is unique among testudinoids in having a ventral strut on the nuchal against which the neural spine ofthe first thoracic abuts (Bramble 1982). This structure is evident in the nuchal (UF 68900) from Leisey. The epiplastra and xiphiplastra are like those of other tortoises in having gular and anal projections. However, these elements, as well as the peripherals 288 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, Fr. I, No. 9 and pygal thought to represent this species, are too large to represent Hesperotestudo mlynarskii and too thin to represent H. crassiscuttata of this size. EMYDIDAE Terrapene carolina (Linneaus 1758) Refemd Specimen&- Leisey lA: UF 80022, most of the anterior lobe of the plastron; UF 80898, partial carapace; UF 81027, the central portion of the posterior lobe of a plastron; UF 82441, parts of a shell; UF 83709, puts of both hyoplastra with hinge; UF 85405, much of a carapace; Leisey 2: UF 125066 and 125067, partial carapaces; UF 125068 and 125069, partial plastrai UF 125070, ilium. All the shell material exhibits fusion between the elements. A hinge between the hyo- and hypoplastra is evident in the preserved plastra, and the bridge peripherals were not sutured to the plastron. Several of these specimens represent large box turtles, up to approximately 230 mm straight carapace length. The posterior peripherals are not extremely recurved, and there is a prominent dorsal keel in UF 80898 and 85405. There can be no doubt that this material represents a box turtle of the genus Terrapene. The peaked shape of the shell and presence of a dorsal keel suggests that T. ornata is not represented. Although it is comparable in size, it does not have the strongly flared posterior peripherals that are typical of Terrapene carohna putnami or T. c. ma/or (Milstead 1969). It has the well developed dorsal keel of T. c bauri, but it appears to represent a form somewhat larger than this subspecies, therefore it is referred simply to Terrapene carolina. Trachemys scripta (Schoepff 1792) Referred Specimens.- This taxon is abundant throughout Leisey lA and also is present in Leisey 2. It is represented in the UF collection by hundreds of isolated elements, as well as several partial shells. Leisey lA: UF 81146, 81488, 82440, and 82612. This material represents a moderately large emydid with a primitive neural pattern (4<6<6<6<6<6<6<6), short rib heads, plastral buttresses reaching costals one and five and inframarginal series divided by contact between marginal and plastral scales. All the carapacial elements are covered by elongate rugosities. There is well developed nuchal scute underlap, and the posterior peripherals are doubly scalloped, but not recurved as in T idahoensis (Jackson 1988). The epiplastra have a straight anterior edge that is perpendicular to the midline and has a small anterior denticle at its lateral extreme. The gular scutes overlap the epiplastron for well over one-half of their dorsal length (56-98°/4 mean = 70.4% f MEYLAN: AMPHIBIANS & REFnLES FROM LEISEY SHELL Prr 289 12.8; N = 11 epiplastral). The entoplastra are slightly wider than long (L/W = 0.86* 0.06, N = 9). The sculpturing of the carapacial elements and the degree of nuchal scute underlap and gular scute overlap is typical of Trachemys scripta. The referred epiplastra are un%ke those of the Pseudemys floridana and P. rubriventris groups which are smoothly rounded anteriorly, lack a well developed denticle, and have much shorter gular scute overlap. Pseudemys sp. Refemd Specimens.- Leisey IA: UF 82758, an epiplastron; Ikisey 3: UF 135681, left hyoplastron. This single emydid el)iplastron is unlike that of Trachemys in having a smoothly rounded anterior margin and very short scute overlap. It could represent either a P. jloridana or P. rubriventris group member. Deirochelys sp. Referred Specimens.- Leisey 2: UF 125071, partial costal; UF 125072, partial peripheral. Jackson (1978) described the diagnostic lateral displacement of the ribheads in this taxon which allows identification of isolated costals. The weak ornamentation, typical of Deirochelys, is seen in both the costal and peripheral. Kinosternon sp. Referred Specimens.- Leisey lA: UF 84920, a nuchal, Leisey 2: UF 125073, nuchal; UF 125081, 125082, two hyoplastron; UF 125083-125086, four hypoplastra; UF 125075-125077, three costals; UF 125074, one neural; UF 125078-125080, three peripherals; UF 125087, one epiplastron, Leisey 3A: UF 124652, a partial nuchal; UF 124684, the left half ofa dentary. The plastral elements of Kinosternon can be recognized by their small size, presence of hinged joints and absence of dorsally projecting plastral buttresses on the hyo- and hypoplastra. The width of the plastral elements suggests that Kinostemon rather than Stemotherus is present. The nuchals have a very narrow area covered by marginal scutes, a vefy small nuchal scute, and broad coverage by the first costal scutes. 290 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL- 37, PT. I, No. 9 TRIONYCHIDAE Apalone ferox (Schneider 1783) Referred Speciment- Ikisey lA: UF 80549, 80662, two costals; UF 81005, epiplastron; UF 81006, nuchal; UF 81031, xiphiplastron; UF 81032, fused hyo- hypoplastra; UF 81064, costal; UF 81091, 81100, two epiplastra; UF 81101, two costals; UF 81141, costal; UF 81763, first left costal; UF 83078, left maxilla; UF 83437, right maxilla; UF 83720, fused hyo-hypoplastra; UF 83843, fused first and second neurals; UF 83859, fused hyo-hypoplastra; UF 83860, right hypoplastron; UF 84004,7th costal; UF 84022, left humerus; UF 84023, cervical five or six; Leisey 2: UF 125060, a skull; UF 125062, fned hyo-hypoplastron; UF 125098, a pleural; UF 125063, a humerus; UF 125061, a neural; Leisey 3: UF 135682-135684, partial hyo-hypoplastron; Leisey 3A: UF 102468, 142236, and 142237, partial costals. Apalone ferox is the only living New World softshell in which the hyo- and hypoplastra are normally fused to one another in the adults. Only one small unfused trionychid hypoplastron was found in this fauna. At least three other fused hyo-hypoplastra were collected. Other features of this trionychid material is consistent with this identification (Meylan 1987). The other two living North American forms, A. spingera and A. mutica, have callosities on the epiplastra as adults. Three adult epiplastra from Leisey (UF 81005, 81091, and 81100) lack such callosities. The seventh costal (UF 84004) has rib attachments for thoracic vertebrae seven and eight fused to it, indicating that the eighth costal was fused with the seventh. This occurs only in the subtribe Platypeltini, to which A. ferox belongs. Furthermore, the maxillae (UF 83078 and 83437) belong to the same individual and reveal that the maxillae met above the premaxillae, as in all softshells, but did not meet on the midline of the palate, which is also a derived feature of this subtribe. DISCUSSION The herpetofauna from Leisey differs little from that of Hillsborough County, Florida today (Table 1). It is essentially modern. Only the two species of Hesperotestudo are now extinct. Two other elements of the fauna, Macroclemys and the tropidophine snake, are no longer found in this area. At present Macroclemys is found no farther south than the drainage of the Suwannee River (Fig. 3). However, there is evidence that it was found throughout much of peninsular Florida in the early Pleistocene (Table 2). The most likely explanation for the marked reduction in the range of Macroclemys during the Pleistocene is the drowning of its riverine habitats by marine transgressions MEYLAN: AMPHIBIANS & REMILES FROM LEISEY SHELL PIT 291 14 i 8 5 7 ,6 4 2 3 Figure 3. Recent range and fossil occurrences for Macroclemys temmincki in Florida. The Recent range of this species is crosshatched. Fossil localities from which this species has been recovered are: (1) FL Green Mine 13. (2) St Petersburg Times, (3) Leisey Shell Pit. (4) Oldsmar 1, (5) Wekiva Spring, (6) Rock Springs, (7) Oklawaha I, and (8) Buzzard Island. The age, material collected museum number, river drainage, and other details of these occurrences are given in Table 2. Table 2. Records from peningular Florida for fossil Macroctemys temmincki outside of its modern range.* 292 B U LLE TIN FLO R ID A M U SEU M N ATU R A L H ISTO R Y VO L„ 37, P r. No. 9 River Max el. L,ocality Land Mammal Age Catalogue No. Element drainage (in m) Ft Green Mine 13 Hemphillian UF 35829 pyga Peace 40 Polk Co. UF 38035 neurals UF 90367 St Petersburg Times late Blancan UF 69683-69690 peripherals 15 Pinellas Co. Leisey Shell Pit Irvingtonian UF 84005, 81198 costal, peripherals Little Manatee 38 Hillsborough Co. 124675, 116093 dentaries Oldsmar 1 early Rancholabrean UF 135629 shell 3 Pinellas Co. Wekiva Spring Rancholabrean UF 2508 mandible Wekiva 19 kg Co. UF 9852 epiplastra, peripherals Leisey 2 Rancholabrean UF 125099 partial skeleton Little Manatee 38 Rock Spring Rancholabrean UF 135685 bridge peripheral St Johns 20 Orange Co. Oklawaha 1 Rancholabrean UF 135686, 135688 dentary, peripheral St Johns 20 135689,135687 peripheral, neural Boinard Island Rancholabrean UF 137924 dentary St Johns 20 St. Johns River near San Mateo • Specimms froIn Macasphalt (-Warren Biolhers) Shell Pit. West Palm Beach Site . Hay's (1908) report o f four peripherals *probably from Peace Creek beds' ES not included. MEYLAN: AMPHIBIANS & REF'TILES FROM LEISEY SHELL PIT 293 Table 3. Preferred habitats of extant species in the Leisey lA herpetofauna. The number ofelements representing each taxon is listed below the Florida plant communities in which each species is f6und regularly. Data are largely from Carr (1940). Longleaf Xeric Mesic Lcntic Lob.c pine, hanunock2 hammock Flatwoods waters3 waters4 1 1 4 Bu/6 terrestria 1 1 1 cf Rana 1 1 1 1 1 1Alligator mississippiensia 19+ Ophisaurus compreisus 1 cf Thamnophis 2 2 2 d Elaphe obsoleta 1 cf. Agkistrodon piactvorous 9 9 9 d Crotalus adamanteus 3 3 3 3 Macroclemys temminch 2 GophemspoOphemus 13 13 Terrapene carolina 6 6 6 Trachemys scripta 100+ Pseudemys 1 1 Kinosternon 1 Apaloneferox 23+ Total for Habitat 18 26 22 26 137+ 24+ 1 = high pine 2 = upland hammock 3 = swamps, ponds 4 = fiven (Pritchard 1989). Webb (1974) pointed out that the ecological requirements of certain species make them particularly susceptible to changes in sea level and thus useful in identifying eustatic changes in Florida. Macroclemys temminckii falls into this category. It is restricted to the main channels of rivers and streams and apparently moves overland infrequently. This species would be eliminated from any basin that was completely submerged by saltwater during a major transgression. Any attempt to determine the extent of transgression necessary to flood any late Neogene or Pleistocene river basins in Florida is complicated by the observation that the northern part ofthe peninsula may have been uplifted 36-41 m during the Pleistocene (Opdyke et at. 1984). Thus, the paleoelevation of the basins for which we have records of Macroclemys (Table 2) cannot be determined accurately. It is, nevertheless, possible to make some observations about the past distribution of Macroclemys in Florida. 294 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL- 37, PT. I. No. 9 A late Miocene low sea level stand between 6.7 and 5.2 Ma (Haq et al. 1987) would be the best explanation for the expansion of this turtle's range into peninsular Florida. Low sea level would have promoted downstream coalescence of major rivers and streams on the Gulf Coast. This regression would explain the Hemphillian record from Polk County, which is in the Peace River drainage. A subsequent transgression and high sea level stand between 5.2 and 3.0 Ma would have eliminated all southern peninsular Macroc/emys populations. Following this major early Pliocene sea level rise subsequent regressions apparently were sufficient to allow Macroclemys to re-enter a number of drainage systems along the west coast of Florida (such as the Wekiva and Little Manatee rivers) but, as far as is known, not sufficient to allow the re-establishment of this species in the Peace River drainage. Transgressions during the Pleistocene may have been responsible for the extinction of Macroc/emys from some of the lower elevation drainages (Oldsmar, Wekiva). However, no transgression that would be sufficient to cause extinction of the Leisey population (Little Manatee River) is known (Haq et al. 1987). It is interesting that during the Rancholabrean, Macroclemys occurred in the St. Johns and Oklawaha rivers (Table 2). It is absent from this system today and could have been eliminated only by a sea level rise of about 20 m. Paleoecology.- Since only two of the 26 taxa found in the herpetofauna of Leisey lA and 3A are extinct accurate reconstruction of the paleoenvironment of these sites is feasible. If we assume that the habitat requirements of the members of the herpetofauna have not changed significantly in the last 1.5 million years, then habitat reconstruction is affected most seriously by problems of transportation, sorting, and sampling. The sampling is admittedly biased, because many elements of the most common spedes (Alligator, Trachemys, Hesperotestudo, Apalone) were not studied and do not appear under referred specimens. However, as large samples of each are reported, and they dominate the reconstruction of the paleoenvironment, it seems unlikely that a complete treatment of all material would change the outcome reported here. Transportation and sorting also are not likely to be serious problems. Disarticulated shells of individual turtles were found in adjacent squares in many cases at Leisey lA (Pratt and Hulbert this volume), and many articulated skeletons were found at Leisey 3A. These associations and the absence of water-worn fossils suggest that the faunas are largely autochthonous. I have followed the methodology employed in Meylan (1982), where the number of occurrences for each taxon is scored for each of six major habitat types in which it occurs commonly (Tables 3 and 4). In this case the absolute number of elements rather than a minimum number of individuals is used. Across the bottom of each table the number of times that a species common to a given habitat type occurs in the Leisey lA and 3A faunas is summed. MEYLAN: AMPHIBIANS & REFnLES FROM LEISEY SHELL PIT 295 Table 4. Preferred habitats of extant species in the Leisey 3A herpetofauna. The number ofelements representing each taxon is lietert below the Florida plant communities in which each species is regularly found Data are largely from Carr (1940). I,ongleaf Xeric Mesic Lentic Lotic pinel hammock~ hammock Flatwoods watcrs~ waters4 Siren sp. 21 Siren lacertina 4 Bufbterresms 1 1 1 1 1 c£ Rana 2 2 2222 Alligator miss,Imppiensis 15+ Ophijaums compresms 1 cf 77,amnophis 11 11 11 6. Nerodia 3 et Nerodia fasciata 9 ct Regina 6 cf. Regina allent 3 ct Elaphe obsoleta 1 Lampropelns getulus 1 1 Coluber or Masticophis 4 4 4 4 4 Farancia 4 c£ Agkistrodon piscivorous 2 2 2 Sistrums milianus 2 2 2 Macroclemys temmincki 1 Apaioneferox 1 Kinosternon 2 Total for Habitat 9 10 20 26 87+ 2 1 = high pine 2 = upland hammock 3 = swamps. ponds 4 = rivers 5 This is an ar~culated struil with mandibles. For both sites, species that are typical of lentic freshwater systems predominate. However, there is a clear difference between the two sites. In Leisey \A Trachemys scripta, Apatone ferox, and Alligator mississippiensis are the most common species in the fauna. These are all freshwater species that can be found in a variety of habitats. They are generally not abundant in large rivers but require open water. In Leisey 3A the most common species are Siren, Thamnophis, Nerodia, and Regina. Again, these are all forms associated with lentic freshwater systems. However, none is an open water species. Two species that normally require some open water Wiligator andApalone) are represented (this is an artifact of curation, not actual rarity). Thus, Leisey 3A is more likely representative of a slough filled with emergent vegetation, while Leisey lA was probably assembled in 296 BULLETIN FLORIDA MUSEUM NATURAL HISTORY VOL 37, Fr. L No. 9 a body of open water such as an oxbow lake or a pond. Both sites contain the lotic species Macroctemys temmincki and, therefore, probably were connected to a permanent river, at least periodically. It is also clear that upland species are more abundant in Leisey lA than in Leisey 3A. Gopherus, Terrapene, and Crotalus are found in the former but not in the latter. Furthermore, 1he two species of Hesperotestudo (not included in Tables 3 and 4) are common in Leisey lA (see species accounts) but represented by a single humerus in Leisey 3A. Thus, Leisey must have been adjacent to an upland community, most likely high pine or xeric hammock, to account for the large number of highly terrestrial turtles. Two species of Florida's Recent herpetofauna, Regina alleni and Ophisaurus compressus, are reported from Leisey and predate any previous reports. Of the three species of glass lizards found in peninsular Florida, 0. ventrWis and 0. attenuatus are described from localities predating Leisey. The only previous report of O. compressus is from the Rancholabrean ofHaile 2B (Auffenberg 1955). LITERATURE CITED Auffenbeg W. 1955. Glass lizards (Ophisaurus) in the Pleistocene and Plioccne of Florida. Herpetologica 11(2):133-136. 1956. Remarks on some Miocene anurans from Florida, with a description of a new species ofHyla. Breviora 52:1-11. . 19631 Fossil testudinine turtles of Florida genera Geochelone and Floridemys. Bult. Florida State Mus., Biol. Sci. 7(2):53-97. 1963b. Fossil makes of Florida. Tulane Stud. Zool. 10:131-126. 1974. Checklist offossillandtortoises. Bull. Florida Sta. Mus., Biol. Sci. 18(3):121-251. . 1988. A new species of Geochelone (Testudinata: Testudinidae) from the Pleistocene of Florida (U.S.A). Acta Zool. Cracov. 31:591-604. Bogert C. M. 19681 The variations and affinities of the dwarf boas of the genus Ungahophis. Amer. Mus. Nov. 2340:1-26. 1968b. Anew genus and species of dwarfboafronisouthern Mexico. Amer. Mus Nov. 2354:1-38. Bramble, D. 1971. Functional morphology, evolution, and paleoccoloy of gopher tortoises. Ph.D. Diss., Univ. California, Berkeley. . 1982. Scaptochely:: Generic revision and evolution of gopher tortoises. Copeia 1982:852-867. Carr, A F., and D. K Caldwell. 1956. The ecology and migrations of sea turtles I. Results of field work in Florida, 1955. Amer. Mus. Nov. 1793:1-23. Crumly, C. R. 1982. A cladistic analysis of Geochelone using cranial osteology. J. Hers). 16:215-234. . 1984. The evolution of land tortoises (family Testudinidae). Ph. D. Dias., Rutgers State Univ., New Brunswick, New Jersey. Dowling H. G., and W. E. Duellman. 1974. Systematic hers,etology: A synopsis of families and higher categories. HISS Publ. Herpetol. 7:1-240. Dowling, H. G., and I. Frics. 1987. A taxonomic study of the ratsnakes VIH. A proposed new genus for Elaphe triaspis (Cope) Herpetologica 43:200-207. Gamwy, E. W. 1975. Phylogeny of the chelydrid turtles: A study of shared derived characters ofthe Skull. Field. Geol. 33:137-178. Goin, C., and W. Auffenberg 1955. The fossil salamanders of the family Sirenidac. Bull. Mus. Comp. Zool. 113(7):497-514. Haq, B. U., J. Hardenbol, and P. R. Vail. 1987. Chronology of fluctuating sea levels since the Triassic. Science 235:1156-1167. MEYLAN: AMPHIBIANS & REFI'ILES FROM LEISEY SHELL Prr 297 Holman, J. A 1963. Late Pleistocene amphibians and reptiles of'the Clear Creek and Ben Franklin local faunas of Texas. So. Methodist Univ., J. Grad. Res. Cen. 31:152-167. 1976. Snakes ofthe Gering Formation (lower Miocene) of Nebraska. Herpetologica 32:88-94. . 1977. Upper Eocene snakes (Reptilia, Serpentes) from Georgia. J. Herp. 11:141-145. . 1982. Geringophis (Serpentes: Boidae) from the middle Ohgocene of Nebraska. Herpetologica 38:489-492 Hulbert, 1 C., Jr., G. S. Morgan, and A R Poyer. 1989. Associated skeleton of megathere, pampathere, tai}ir and turtles from the latest Pliocene or earliest Pleistocene of north=central Florida. J. Vert Paleon. 9(3) Supplement:26A (abstract). Jackson, D. R. 1978. Evolution and fossil record of the chicken turtle Deirochelys, with a re-evaluation of the genus. Tulane Stud Zool Bot. 20:35-55. . 1988. A re-examination of fossil turtles of the genus Trachemys (Testudines: Emydidae). Herpetologica 44:317-325. LeBufC C. R. 1957. The range of Croco*ha acums along the Florida Gulf coast Herpetologica 13:188. I«overidge. A, and E. E Williams 1957 Revision of the African tortoises and turtles of the suborder Cryptodira. Bull. Mus.Comp. Zool. 115:163-557. Lynch, J. D. 1966. Additional treefrogs (Hylidae) from the North American Pleistocene. Ann. Carnegie Mus. 11:265-271. Meylan, P. A. 1982. The squamate reptiles ofthe Inglis lA fauna (Irvingtonian:Citrus Co., Florida). Bull. Florida State Mus., Biol. Sci. 27(3):145. 1987. The phylogenetic relationships of soft-shelled turtles (family Trionychidae). Bull. Amer. Mus. Nat. Hist 186:1-101. Milstead, W. W. 1969. Studies on the evolution of box turtles (genus Ten-apene). Bull. Florida State Mui, Biol. Sci. 14(1):1-113. Neill W. T. 1971. Thelastoftheruling reptiles,alligators crocodiles and their kin Columbia Univ. Press, New York, 486 pp. Opdyke, N. D., D. P. Spangler, D. L Smith, D. S. Jones. and R. C. UndquisL 1984. Origin of the epetrogenic uplift of Pliocene-Pleistocene beach ridges in Florida and development of the Florida karst Geology 12:226-228. Palmer, W. 1987 A new species of glass lizard (Anguidae:Ophiwurus) from the southeastern United States. Hers)etologica 43(4):415-423. Prestog k E. 1979. Late Pleistocene cold-blooded vertebrate faunas from the mid-continental United States. I. Reptilia; Testudines, Crocodilia Univ. Michigan Mus. Zool., Paleon- Pap. 19:1-53. Pritchard, P. C. H. 1989. The alligator snapping turtle: Biology and conservation. Milwaukee Pub. Mus., Milwaukee, Wisconsin, 104 pp. Smith, H. M., R. B. Smith, and H. L Sawin. 1977. A summary ofsnake classification (Reptilia, Serpentes). J. Herp. 1:115-121. Tihen, J. A 1962. A review ofNew World fossil bufonids. Amer. Midl. Nat 68:1-50. Undenvood, G. 1976. A sys:ematic anal>,is ofboid snakcs. Pp. 151-175 in A crA Bellairs and C. B. Cox, eds. Morphology and Biology of Reptiles. Iinnean Society Symposium Volume. Webb, S. D. 1974. Chronology of Florida Pleistocene mammals. Pp. 5-31 in S. D. Webb, ed. Pleistocene Mammals off:lorida Univ. Presses Florida, Gainesville. Williams, E. E. 1930. Variation and selection in the cervical central articulations of living tuttles Bull Amer. Mus. Nat. Hist 94:309-561. Zangert, R 1969. Theturtle shell. Pp. 311-339 in C. Gans, A d'A Bellairs, and T. Parsons, eds. Biology oflhe Reptilia. Vol. 1. Academic Press, Izndon and New York.