Journal of Vertebrate Paleontology 25(4):979–982, December 2005 © 2005 by the Society of Vertebrate Paleontology

NOTE

A REAPPRAISAL OF SOME PALEOGENE TURTLES FROM THE SOUTHEASTERN UNITED STATES PATRICIA A. HOLROYD1, JAMES F. PARHAM1,2, and J. HOWARD HUTCHISON1, 1Museum of Paleontology, University of California, Berkeley, California 94720, U.S.A; 2Evolutionary Genomics Department, Joint Genome Institute, 2800 Mitchell Drive, Walnut Creek, California 94598, U.S.A., [email protected]

North American Paleogene turtles are best known from the intermontane basins of the Western Interior, stretching along a north-south axis from Alberta and Saskatchewan to New Mexico (see e.g., Hutchison, 1982, 1992, 1998; Holroyd and Hutchison, 2000; Holroyd et al., 2001). Although its fossil record is poor by contrast, the southeastern part of the United States still is an important area of diversification for several continental turtle clades (e.g., Kinosternoidae, Dermatemydidae; Hutchison, 1982, 1991, 1998). Most of the Paleogene turtle faunas described from the eastern United States (i.e., east of the Mississippi River) are found in near-shore marine sediments and represent mixed assemblages that include both freshwater and saltwater forms. Trionychids and chelonioids are the most common taxa, while pleurodires, kinosternoids, and adocids are much more rare (e.g., Weems, 1988, 1999; Hutchison and Weems, 1999). Given these combinations of taxa, it is difficult to say with confidence whether any of these turtle assemblages from the eastern U.S. exclusively represents either freshwater or continental ecosystems. Reports of continental turtles from the Eocene and Oligocene of the southeastern U.S. are scarce. Three reports of testudinids include Hay’s (1899) description of Hadrianus schucherti from the upper Eocene Jackson Group of western Alabama, Patton’s (1969) referral of specimens from Florida’s Oligocene I-76 local fauna to the testudinid Floridemys, and Franz and Franz’s (2004) report of a Gopherus-like form from the Oligocene Chandler Bridge Formation of South Carolina. Holman and Case (1988) have reported middle Eocene trionychids and emydids from the Point “A” Dam site, Tallahatta Formation, Covington Co, Alabama, and Holman (2000) has reported middle Eocene trionychids and emydids from the Jim Woodruff Dam site, Gadsden Co., Florida. These claims are important, because they purport to be the first reported evidence of aquatic testudinoids in the Paleogene of the southeastern US. Here we reexamine these important records and suggest alternative interpretations of the age and affinities of some of these turtles. Abbreviations and Terminology—Institutional: MSUVP—Michigan State University Museum, East Lansing, Michigan. Osteological: C, costal (e.g., C1); P, peripheral (e.g., P2). All measurements are in millimeters (mm). The term “peripheral length” refers to the straight line dimension of the peripheral bone between its cephalic and caudal sutures at the perimeter of the shell as seen in dorsal view. Bone and scale terminology follows Zangerl (1969). Taxonomic framework follows Joyce et al. (2004). SYSTEMATIC PALEONTOLOGY TESTUDINES Linnaeus, 1758 CRYPTODIRA Cope, 1868 TRIONYCHIDAE Gray, 1825 TRIONYCHINAE Gray, 1825 Genus indet. Referred Material—MSUVP 1960A, right hypoplastron (Fig. 1); MSUVP 1960B, distal costal fragment (Holman, 2000:fig. 2A–B); MSUVP 1961, costal free rib fragment; MSUVP 1962, terminal costal rib fragment (Holman, 2000:fig. 2c), all from the Jim Woodruff Dam site; MSUVP 1186, two costal fragments and a neural, Point “A” Dam site (Holman and Case, 1988:fig. 1). Remarks—The most complete and informative element is the nearly complete plastral element MSUVP 1960A (Fig. 1). Originally referred to

Trionychidae indet. as a costal, the presence of a midline suture that is concave in shape and extensive free edges indicate that this specimen is the medial portion of a hypoplastron. The presence of an unpitted and tapered medial margin with a large, open pattern of pitting generally excludes the Plastomeninae and indicates a trionychine. Among North American Eocene trionychines, this specimen can also be excluded from the genus Axestemys, an extremely large pedomorphic taxon that lacks well-defined surface sculpture. Holman (2000) noted that MSUVP 1960A is similar to specimens referred to “Amyda? virginiana (Clark) by Hay (1908) . . . , a taxon that probably should be referred to the modern genus Apalone.”; Hutchison and Weems (1998:182, fig. 9) noted that the type material of Trionyx virginianus Clark 1895 is not generically or specifically diagnostic and that the concept of the species rests primarily on the shell figured by Lynn (1929:fig. 1, 2) and Weems (1988:fig. 9, 10). Weems (1988) referred the species to Aspideretes, and Hutchison and Weems (1998) referred it questionably to Aspideretes. This genus, based on extant species on the Indian subcontinent, is used as a form genus in North America for several Late Cretaceous and Paleogene species with a preneural, although Gardner et al. (1994) referred many of the North American Cretaceous “Aspideretes” to a new genus Aspideretoides. Apalone is restricted to those fossil and extant taxa lacking a preneural. In the absence of more diagnostic elements and a more thorough revision of North American Paleogene species variously attributed to “Aspideretes,” the generic attribution of trionychines from the southeastern United States remains unclear, although it is unlikely that any should be ascribed to Aspideretes. CHELONIOIDEA Baur, 1893 Referred Material—MSUVP 1964, 3 rd or 5 th neural (Holman, 2000:fig. 3a; Fig. 2); MSUVP 1965, proximal portion of a posterior costal (Holman, 2000:fig. 3b; Fig. 3); MSUVP 1966, peripheral fragment (Holman, 2000:fig 3c; Fig. 4), all from the Jim Woodruff Dam site, Florida. MSUVP 1205, four proximal costal fragments and a peripheral from the Point “A” Dam site, Tallahatta Formation, Alabama (Holman and Case, 1988:fig. 2a-e; Figs. 5, 6). Remarks—MSUVP 1964 is a well-preserved, isolated neural and is the only complete bone in the assemblage. Holman (2000) remarked that this specimen is similar to the third neural of a batagurid (⳱geoemydid) or emydid turtle (i.e., a non-testudinid testudinoid). Although described as hexagonal, the bone is more accurately characterized as casque-shaped with the anterior end indented to receive the posterior process of the preceding neural. This neural shape is primitive for cryptodires and is widely distributed among clades. The absence of a median dorsal keel precludes a number of taxa from consideration (e.g., chelydrids and kinosternoids), but this specimen is not immediately distinguishable from the neurals of stem cheloniid sea turtles (Pancheloniidae of Joyce et al., 2004) and shows remarkable resemblance to the neural ascribed to Puppigerus camperi by Weems (1999). The posterior placement of a straight vertebral scale sulcus renders it more like the 5th neural of a stem cheloniid turtle than the third neural of a geoemydid/emydid turtle, although we agree that it is similar to the condition found in the Miocene emydid Trachemys hilli (Cope, 1878). A referral of this element to Chelonioidea would also be consistent with the thinness and porosity of the bone. Holman (2000) noted the great width and depth of the vertebral sulcus, a feature that is not diagnostic of later-occurring marine turtles or testudinoids, but similar to that seen in certain Paleogene sea turtles,

979

980

JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005

FIGURE 1. Trionychidae, MSUVP 1960A from the Jim Woodruff Dam site. Hypoplastron in ventral view. Scale bar equals 1 cm.

FIGURE 3. Chelonioidea, MSUVP 1965. Proximal costal from the Jim Woodruff Dam site, in external (left), visceral (center), and right lateral view (right). Scale bar equals 1 cm.

e.g.,“Catapleura ruhoffi” Weems, 1988 (see taxonomic comment of Tong and Hirayama, 2002) or the material referred to Puppigerus by Weems (1999). MSUVP 1965 is a proximal fragment of a costal bone. Holman (2000) notes that it is unsculptured, as in the Bataguridae. Many emydid turtles also lack sculpture, as do most sea turtles and dermatemydids; therefore, this character is not diagnostic for any one group. Holman (2000) suggests that MSUVP 1965 is a right costal from the posterior part of the shell. Though not explicitly stated, this assessment is probably based on the curvature of the bone. While it is true that the posterior costals are curved, anterior costals can be as well, and so we refrain from designating a side for this specimen. The scale sulci represent the juncture of two pleural scales and one vertebral scale. The sulcus between the pleural scales is parallel to the midline axis of the bone, but is shifted either anteriorly or posteriorly so that it divides the costal into unequal portions. If we accept that this is a posterior right costal, then the sulcus is on the anterior third of the sixth costal. Unfortunately, this is not consistent with the common scalation pattern of any Paleogene testudinoid. Although variation exists, the typical pattern is to have the pleural scale sulcus on the posterior half of the sixth costal. Finally, the ventral surface preserves the base of the rib head in MSUVP 1965. Although the rib head itself is missing, enough is preserved to show that it was cylindrical and stout. Distal to the base of the rib head, the imbedded rib is evident as a low ridge. The costal fragments differ from corresponding elements of North American Eocene batagurids, emydids, testudinids, and kinosternoids in being highly cancellous with thin outer and inner surface laminae. The cylindrical rib head, the visibility of the imbedded rib, and the thinness and flatness of the bone itself are all more similar to the morphology of the posterior costals of pancheloniid sea turtles, and the preserved morphology is indistinguishable from that seen in Puppigerus (Moody, 1974; Schleich, 1988). MSUVP 1205 is a batch catalogued specimen containing three proximal costal fragments (Holman and Case, 1988:fig. 2a, b, d: Fig. 5) that preserve the juncture of the vertebral and pleural scales but are less complete than MSUVP 1965. The only character cited in favor of their assignment to Emydidae by Holman and Case (1988) is that their smooth

texture is like that of some species of the batagurid Echmatemys from the Eocene of Wyoming. These specimens bear no particular resemblance to Echmatemys, which is much thicker shelled, has thick external laminae, and typically displays much more deeply incised sulci. In any case, the smooth texture is not diagnostic for aquatic testudinoids. MSUVP 1966 is a fragment of a peripheral that Holman (2000) interpreted as a right peripheral from the anterior part of the shell of a batagurid or emydid. The peripheral fragment is strikingly smooth (polished) but has a pattern of finely incised vermiculate vascular grooves. This type of vascularization is typically absent in testudinoids, but occurs commonly in chelydrids, chelonioids, and pleurodires. The element has very thin surficial laminae, strongly cancellous internal morphology (Zangerl 1948a:pl. 1, figs. 1, 2), a vacuity for deep insertion of the rib into the peripheral, a sharp and unserrated free margin, and lacks growth annuli, consistent with marine turtles. The relatively equal length-towidth ratio is also consistent with Cretaceous and Paleogene chelonioids (see e.g., Zangerl, 1953:figs. 76, 88). The deep ventral coverage of the scales differs from that of known pleurodires (Zangerl, 1948b:fig. 5) and resembles that of pancheloniids. One of the most striking features of MSUVP 1965 and the costals in MSUVP 1205 is the thinness of the bone. It is much thinner than costals of most aquatic testudinoids. While some aquatic testudinoids have thin shells, these tend to be highly domed forms (e.g., Emys blandingii). The flatness of these costals precludes them from belonging to a high-domed turtle. MSUVP 1965 can be confidently referred to Chelonioidea based on the features above. The MSUVP 1205 costals also are referred to Chelonioidea, but with less confidence. All of these specimens lack features that would permit them to be referred to a genus within Chelonioidea.

FIGURE 2. Chelonioidea, MSUVP 1964 from the Jim Woodruff Dam site. Neural in external (left), visceral (center), and right lateral view (right). Scale bar equals 1 cm.

FIGURE 4. Chelonioidea, MSUVP 1966. Peripheral from the Jim Woodruff Dam site, in visceral (left) and external (right) views. Scale bar equals 1 cm.

TESTUDINES, indet. Remarks—In addition to the three costals listed above, MSUVP 1205 includes a nearly complete peripheral element (Holman and Case, 1988:fig. 5c; Fig. 6). The anterior edge of the internal surface is eroded, but the insertion for the costal rib is still present. The extent of the pit

NOTES

981

FIGURE 7. Testudines indet., MSUVP 1967. Indeterminate element in external (left) and visceral (right) view from the Jim Woodruff Dam site. Scale bar equals 1 cm.

specimen exhibits no features that would suggest it is a testudinoid xiphiplastron. The margins are abraded and rounded, and the presence or absence of sutures cannot be verified. There is a transverse ridge that exhibits a topographic resemblance to the dorsal scale margin of a testudinoid xiphiplastron. However, the ridge is covered by dense, shiny lamellar bone and is longitudinally striated as in trionychid ribs and plastral struts. The specimen is also very thick (4 mm) relative to its width (40 mm), suggesting that it is not a xiphiplastron. It is possible that this specimen is a highly abraded posterior costal of a trionychid. The element remains unidentified. DISCUSSION Age and Composition of the Jim Woodruff Dam Site

FIGURE 5. Chelonioidea, MSUVP 1205. Costals from Point “A” Dam site. Scale bar equals 1 cm.

shows that the costal rib was stout with a blunt posterior end. The peripheral tapers laterally and scute sulci are visible on the flat dorsal and convex ventral surfaces. The peripheral has a squarer outline than that of extant cheloniids, but also lacks any diagnostic features of aquatic testudinoids. The element for now remains unidentifiable. Holman (2000:fig.3d) identified MSUVP 1967 (Fig. 7) as a batagurid or emydid xiphiplastron. Other than its generally triangular shape, this

The Jim Woodruff Dam site lies just south of this dam along the Apalachicola River (see detailed map in Holman, 2000:fig. 1). The geologic context of the finds was not noted by the original collectors. Holman (2000) interpreted the age of the Jim Woodruff Dam site as middle Eocene, based on the combined presence of the whale Basilosaurus and the snake Palaeophis. However, the rocks along the Apalachicola River in this area have most recently been mapped as “Residuum on Oligocene sediments,” i.e, undifferentiated weathered sediments of lower and upper Oligocene age composed of sandy clays with inclusions of fossiliferous, silicified limestone (Huddlestun, 1993; Scott et al., 2001). Given the nature of the mapped sediments, there are different ways in which the fossils of the Jim Woodruff Dam site can be interpreted. First, it is likely that the fossils are a reworked assemblage, an idea supported by the abraded and polished nature of some of the material. Second, it is possible that these fossils represent an assemblage from more than one original source horizon that could vary in age from middle Eocene to Oligocene. Such a mixture would be consistent with the mixed marine and estuarine nature of the fossils, although as noted by Holman (2000), a combination of marine and continental organisms is not uncommon in estuarine environments or in nearshore marine environments (e.g., Holroyd et al., 1996). In any event, we do not feel confident that the fossils from this site constitute a local fauna because of their possible lack of contemporaneity. Also, we do not think the age of the specimens can be ascertained with certainty unless subsequent finds can establish the stratigraphic provenance of a similar combination of faunal elements. For these reasons, we interpret the Jim Woodruff Dam turtles as being a mixed assemblage that contains both marine chelonioids and freshwater trionychids. This mixture may reflect the original nearshore depositional setting, reworking of source sediments, or some combination of both factors. Characterization of Point “A” Dam Site and Jim Woodruff Dam Site Turtle Faunas

FIGURE 6. Testudines indet., MSUVP 1205. Peripheral from Point “A” Dam site in external, right lateral, visceral, and medial views. Scale bar equals 1 cm.

Based on our reappraisal of the Point “A” Dam site and the Jim Woodruff Dam site turtles, there are presently no reliable published records of aquatic testudinoids from the Paleogene of the southeastern United States. Instead, the Point “A” Dam site and the Jim Woodruff

982

JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 25, NO. 4, 2005

Dam site are more similar in taxonomic composition to other Paleogene sites on the Atlantic coastal plain (Weems, 1988; Hutchison and Weems, 1998; Weems, 1999). Our findings are also consistent with the depositional history of the area, because throughout most of the Paleogene, northern Florida and eastern Alabama were submerged under shallow seas that did not retreat until the late Eocene (e.g., Huddleston, 1993; Manning, 2003). Acknowledgments—We thank Alan Holman and the staff of the Michigan State University Museum for access to the Alabama and Florida turtles, Thierry Smith of the IRSNB for access to comparative material of European sea turtles, and an anonymous reviewer for helpful comments and editing. This work is LBNL-56274 and was performed under the auspices of the U.S. Department of Energy, Office of Biological and Environmental Research. This is UCMP contribution No. 1865. LITERATURE CITED Baur, G. 1893. Notes on the classification of the Cryptodira. American Naturalist 27:672–675. Clark, W. B. 1895. Contributions to the Eocene fauna of the Middle Atlantic Slope. Johns Hopkins University Circulars 15(121): 3–6. Cope, E. D. 1865. Third contribution to the herpetology of tropical America. Proceedings of the Academy of Natural Sciences of Philadelphia 1865:185–198. Cope, E. D. 1868. On the origin of genera. Proceedings of the Academy of Natural Sciences of Philadelphia 1868:242–300. Cope, E. D. 1878. Descriptions of new extinct Vertebrata from the upper Tertiary and Dakota formations. Bulletin of the United States Geological and Geographical Survey, Territories 4:379–396. Franz, R., and S. Franz. 2004. Gopher tortoise evolution: East vs West. A possible paradigm shift. 29th Annual Meeting and Symposium of the Desert Tortoise Council, Abstracts. Gardner, J. D., A. P. Russell, and D. B. Brinkman. 1994. Systematics and taxonomy of soft-shelled turtles (Family Trionychidae) from the Judith River Group (mid-Campanian) of North America. Canadian Journal of Earth Sciences 32:631–643. Gray, J. E. 1825. A synopsis of the genera of reptiles and Amphibia, with a description of some new species. Annals of Philosophy 10:193–217. Hay, O. P. 1899. Descriptions of two new species of tortoises from the tertiary [sic] of the United States. Proceedings of the United States National Museum 22:21–24. Hay, O. P. 1908. Fossil turtles of North America. Carnegie Institute of Washington, Publication 75: 1–568. Holman, J. A., and G. R Case. 1988. Reptiles from the Eocene Tallahatta Formation of Alabama. Journal of Vertebrate Paleontology 8: 328–333. Holman, J. A. 2000. First report of an Eocene reptile fauna from Florida, USA. Palaeovertebrata 30(1–2):1–10. Holroyd, P. A., and J. H. Hutchison. 2000. Proximate causes for changes in vertebrate diversity in the early Paleogene: an example from turtles in the Western Interior of North America. GFF 122:75–76. Holroyd, P. A., J. H. Hutchison, and S. G. Strait. 2001. Changes in turtle diversity and abundance through the earliest Eocene Willwood Formation: preliminary results. University of Michigan Papers in Paleontology 33:97–108. Holroyd, P.A., E. L. Simons, T. M. Bown, P. D. Polly, and M. J. Kraus. 1996. New records of terrestrial mammals from the upper Eocene Qasr El Sagha Formation, Fayum Depression, Egypt. Paléobiologie et évolution des mammifères du Paléogène; Volume jubilaire en homage à Donald E. Russell. Palaeovertebrata 25:175–192. Huddlestun, P. F. 1993. A revision of the lithostratigraphic units of the Coastal Plain of Georgia—The Oligocene. Georgia Geological Survey Bulletin 105:1–152. Hutchison, J. H. 1982. Turtle, crocodilian, and champsosaur diversity changes in the Cenozoic of the north-central region of western United States. Palaeogeography, Palaeoclimatology, Palaeoecology 37:147–164. Hutchison, J. H. 1991. Early Kinosternidae (Reptilia: Testudines) and

their phylogenetic significance. Journal of Vertebrate Paleontology 11:145–167. Hutchison, J. H. 1992. Western North American reptile and amphibian record across the Eocene/Oligocene boundary and its climatic implications; pp. 451–467 in D. R. Prothero and W. A. Berggren (eds.), Eocene-Oligocene Climate and Biotic Evolution. Princeton University Press, Princeton, New Jersey. Hutchison, J. H. 1998. Turtles across the Paleocene/Eocene epoch boundary in west-central North America; pp. 401–408 in M.-P. Aubry, S. G. Lucas, and W. A. Berggren (eds.), Late PaleoceneEarly Eocene Climatic and Biotic Events in the Marine and Terrestrial Records. Princeton University Press, Princeton, New Jersey. Hutchison, J. H., and R. E. Weems. 1998. Paleocene turtle remains from South Carolina. American Philosophical Society Transactions 88: 165–195. Joyce, W. G., J. F. Parham, and J. A. Gauthier. 2004. Developing a protocol for the conversion of rank names to phylogenetically defined clade names, as exemplified by turtles. Journal of Paleontology 78:989–1013. Linnaeus, C. 1758. Systema Naturae, Volume 1 (tenth edition). Laurentius Salvius, Holmia, 824 pp. Lynn, W. G. 1929. A nearly complete carapace of a fossil turtle Amyda virginiana (Clark). Proceedings of the United States National Museum 76(26):1–4. Manning, E. M. 2003. The Eocene-Oligocene transition in marine vertebrates of the Gulf Coastal Plain; pp. 366–385 in D. R. Prothero, L. C. Ivany, and E. A. Nesbitt (eds.), From Greenhouse to Icehouse: The Marine Eocene-Oligocene Transition. Columbia University Press, New York, New York. Moody, R. T. J. 1974. The taxonomy and morphology of Puppigerus camperi (Gray), an Eocene sea turtle from northern Europe. Bulletin of the British Museum (Natural History), Geology 25:153–186. Patton, T. H. 1969. An Oligocene land vertebrate fauna from Florida. Journal of Paleontology 43:543–546. Scott, T. M., K. M. Campbell, F. R. Rupert, J. D. Arthur, T. M. Missimer, J. M. Lloyd, J. W. Yon, and J. G. Duncan. 2001. Geologic Map of the State of Florida. Florida Geological Survey Open-File Report No. 80. Schleich, H. H. 1988. Eozäne Schildkrötenreste (Reptilia, Testudines) von St Pankraz am Haunsberg (Österreich). Studia Geologica Salmanticensia, Volumen especiale 3:65–184. Tong, H., and R. Hirayama. 2002. A new species of Tasbacka (Testudines: Cryptodira: Cheloniidae) from the Paleocene of Ouled Abdoun phosphate basin, Morocco. Neues Jahrbuch für Geologie und Paläontologie, Monatshefte 2002(5):277–294. Weems, R. E. 1988. Paleocene turtles from the Aquia and Brightseat formations, with a discussion of their bearing on sea turtle evolution and phylogeny. Proceedings of the Biological Society of Washington 101:109–145. Weems, R. E. 1999. Reptile remains from the Fisher/Sullivan Site; pp. 101–121 in R. E. Weems and G. J. Grimsley (eds.), Early Eocene Vertebrates and Plants from the Fisher/Sullivan Site (Nanjemoy Formation) Stafford County, Virginia. Virginia Division of Mineral Resources Publication, Report 152. Zangerl, R. 1948a. The vertebrate fauna of the Selma Formation of Alabama. Part 1. Introduction. Fieldiana Geology Memoirs 3:1–16. Zangerl, R. 1948b. The vertebrate fauna of the Selma Formation of Alabama. Part II. The pleurodiran turtles. Fieldiana Geology Memoirs 3:17–56. Zangerl, R. 1953. The vertebrate fauna of the Selma Formation of Alabama. Part IV. The turtles of the family Toxochelyidae. Fieldiana Geology Memoirs 3:137–277. Zangerl, R. 1969. The turtle shell; pp. 311–339, in C. Gans, A. Bellairs, and T. S. Parsons (eds.), Biology of the Reptilia I. Academic Press, New York. Submitted 20 December 2004; accepted 26 April 2005.