PaleoBios 19(3):l-7, December 15, 1999 © 1999 University of California Museum of Paleontology Late Cretaceous sea turtles from the Chico Formation of California JAMES FORD PARKAM1-2 and THOMAS ALLEN STIDHAM23 Department of Integrative Biology and Museum of Paleontology, University of California, Berkeley, CA 94720; parham@socrates.berkeIey.edu; 3Museum of Vertebrate Zoology, furcula@socrates.berkeley.edu. New chelonioid sea turtle specimens from the Late Cretaceous (Campanian) Chico Formation of Northern California include taxa from the clades Cheloniidae and Dermochelyidae. The dermochelyid material represents at least one undescribed species. These specimens are the oldest Pacific occurrence of cheloniids and dermochelyids, and one of the oldest records of their association. No protostegids are known from the Chico Formation, highlighting the taxonomic differences between the pre and post-Campanian Pacific Ocean turtle faunas. The turtles of the Chico Formation exhibit bioueographic affinities with both Japanese and Eastern North American Late Cretaceous chelonioids. INTRODUCTION Turtles invaded the marine realm several times (Gaffney and Meylan 1988). The most diverse radiation, i:i terms of ecology, morphology, and number of species, involved a group of cryptodiran turtles known collectively as chelonioids (cheloniids, protostegids, and dermochelyids). Although the oldest chelonioid is over 110 million years old (Late Aptian or Early Albian, Hirayama 1998), the fossil record of chelonioids remains sparse until the Coniacian (-90 Ma). Most of our knowledge of fossil chelonioids comes from the Coniacian to the Campanian (90-74 Ma) (Zangerl 1953a, b; Hirayama 1997). The Cretaceous epicontinental seas of North America were conducive to chelonioid habi- tation, as well as their fossilization, and numerous speci- mens are known from the Niobrara Chalk of Kansas, Mooreville Chalk of Alabama, and Pierre Shale of South Dakota (Hay 1908, Zangerl 1953a, b, Hirayama 1997). The record of contemporaneous Pacific chelonioids is very poor. In this paper we describe Late Cretaceous (Campanian) fossil sea turtles from the Chico Formation of California. Geographically, these specimens help bridge the gap be- tween the previously described Late Cretaceous sea turtles of North America and the Western Pacific Ocean. By comparing their stratigraphic occurrences, we hope to shed light on the biogcography and faunal turnover of Late Cretaceous chelonioids in the Pacific. MATERIAL The turtle fossils recovered from the Chico Formation consist of fourteen fragments, seven from each of the two main fossil localities (Fig- 1). Of these fourteen, only four are diagnostic beyond Chelonioidea and warrant detailed description. They include two girdle fragments, one hu- merus, and one cranial element. This is the first turtle material formally described from the Chico Formation. 1 author for correspondence Abbreviations: UCMP, University of California Mu- seum of Paleontology, Berkeley, CA; SC, Sierra College Museum of Natural History, Rocklin, CA vertebrate local- ity; VRT, Sierra College Museum of Natural History fossil vertebrate specimen number. GEOLOGY The Chico Formation unconformabiy overlies the Or- egon City Formation and unconformabiy underlies the lone and "Dry Creek" Formations (Crccly 1965) in the area where the fossils were collected (Fig. 1). The Chico Formation reaches a thickness of 600m (Creely 1965), but it is much thinner in Placer and Butte counties where the fossils were recovered (Hilton and Antuzzi 1997; J. Parham personal observation). This formation is composed pre- dominantly of tan and gray sandstone, mudstonc, and shale units containing abundant organic material (mostly carbon- ized plant fragments) interbedded with carbonaceous, glau- conitic, mollusk shell turbidites. The turbidites arc up to two meters thick and several meters wide. Fossils are abun- dant in the turbidites and include gastropods, ammonites, other mollusks, and many vertebrate bones and teeth. In contrast to the northern turbidite localities, the southern turtle fossils were all collected from limestone concretion and siltstone units (Hilton and Antuzzi 1997). These fos- sils do not show any obvious alignment within the beds. Chico Formation fossils appear to have been deposited in a locally anoxic, shelf environment close to a rocky beach or the mouth of a delta. This is supported by the presence of fluvial deposits 2 km east of the Dry Creek localities (Creely 1965). Overall, the depositional fades appear to be similar within the region where the fossils have been found (Fig. 1). Localities Vertebrate fossils have been collected from two main areas in the Chico Formation termed Granite Bay and Dry Creek (Fig. 1). Hilton and Antuzzi (1997) give locality information for the Granite Bay sites, and detailed locality 2 PALEOBIOS, VOL. 19, NUMBER 3, 1999 Fig. 1. Map of California showing the two main vertebrate fossil areas in the Chico Formation plotted within in their respective counties (shaded dark gray); Dry Creek, Placer (lounty and Granite Hay, Butte County. information for the Dry Creek sites is available from the UCMP and Sierra College. Granite Bay (also known as Swan Lake, SC1571) was discovered and collected in 1995 during a construction project in Granite Bay, Placer County, California (Hilton and Antuzzi 1997). Details of the local stratigraphy and many of the fossils were repotted by Hilton and Antuzzi (1997). This area is no longer acces- sible since it is now covered by landfill and homes. The second area lies about 3 km northeast of Oroville in Butte County, California. This area is referred to as Dry Creek because most of the vertebrate localities in this area lie within the glauconitic turbidites that compose the banks of Dry Creek for much of its length. This area includes Gohre's Glauconite (UCMP V98117, equivalent in part to SC 1612) and Dry Creek South (UCMP 94002). This site was a single, approximately one half cubic meter block from a turbiditc horizon about 2 meters above the level of Dry Creek. This block was destroyed as it was broken apart in the search for vertebrate fossils. Important fossils such as a pterosaur bone, an Ichthyornis humerus (Hilton et al. 1999), and a turtle bone were collected from this block. For this reason, and because it appeared that the block might have been displaced by erosion, this collection of vertebrate and invertebrate fossils was catalogued as a sepa- rate site. Dry Creek South is a general locality that includes all other vertebrates collected along Dry ("reek. Gohre's Glau- conite lies within what is mapped as Dry Creek South. The Sierra College sites 1612, 1613, and 1614 correspond to the western, central, and eastern parts of Dry Creek South. These localities lie south of the Dry Creek invertebrate fossil site? and outcrops described by Creely (1965). All Dry Creek localities are within 100 m of each other along the course of Dry Creek and are within 2 meters stratigraphically. Other vertebrate remains from these sites include diverse elasmobranchs (Long et al. 1997), mosa- saurs (Hilton and Antuzzi 1997), pterosaurs, and birds (Hilton et al. 1999). Horizon md Age The Dry Creek sites yield abundant remains of the ammonite Baculites cbicoensis (Creely 1965, J. Parham per- sonal observation). This species of Baculites is known only from the latter part of the early Campanian (Ward 1978, Matsumoto 1984), approximately 82-80 Ma. No other species of Baculites is known from the Chico Formation (Creely 1965) though many have been collected. The Baculites from Granite Bay have not been identified to species, bait other ammonites (lloplitoplaccnticeras and Mctaplacentieeras) from Granite Bay (Hilton and Antuzzi 1997) are consistent with the age estimate from Dry Creek. Therefore, we find no reason to consider the two main fossil areas as significantly different in age, and treat them as a single time-rock unit. The correlation of the Chico For- mation with other formations based on the presence of B. cbicoensis indicates an equivalent age for the Canadian Pender Siltstone (Ward 1978) and bed Illd-e of the Japa- nese Upper Yezo Group (Matsumoto and Obata 1963). SYSTEMATIC PALEONTOLOGY TESTUDINES Linnaeus 1758 cryptodira Gray 1825 chelonioidea Bauer 1893 dermochelyidaf. Gray 1825 Description-UCMP 172070 (Figs. 2A-B), from Dry Creek South (UCMP V94002), is a large left scapula missing most of the scapular prong. Its maximum length is 19.8 cm from the glenoid to the preserved end of the acromion. In life both the scapular prong and acromion were over 20 cm long and likely belonged to an individual with a carapace over a meter in length. The preserved base of the scapular prong indicates that the angle formed between it and the acromion was well over 90 degrees (Fig. 215). The acromion is nearly complete, missing only the most distal portion. Proximally it is subcylindrical, but squarish in cross-section midway to the distal end, which is dorso-ventrally compressed. A distinct raised surface is present near the posterior base of the acromion (Fig. 2A). This tubercle is approximately 3 cm in diameter, and it increases the width of the shaft by 0.5 cm posteriorly. At its base, the scapular prong is antero-posteriorly compressed. Discussion—Among Campanian chelonioids, the protostegds and dermochelyids have a wide angle between PARHAM &STIDHAM-CKY.TACEOUS SEA TURTLES FROM CALIFORNIA B Fig. 2. A. UCMl' 172070 (V94002), dermochelyid left scapula in ventral view showing acromion tubercle (arrow). Scale bar = 5 cm. B. UCMl' 172070, posterior view. C. VRT 19 (SC 1571), dermochelyid left coracoid in dorsal view showing proximal portion of the coracoid ridge (arrow). Scale bar = 5 cm. the scapular prong and acromion, although it docs occur convergently in some Cenozoic chcloniids 'Hirayama 1994). An acromion tubercle docs not occur in protostegids (Wieland 1906; Zangerl and Sloan 1960), and for this reason we assign UCMP 172070 to the Dermochelyidae. Walker (1973) noted the presence of" this tubercle in the sole extant dermochelyid, Dermochelys coriacca Fitzingcr 1843. In D. coriacca the tubercle receives the lateral por- tion of a divided acromiocoracoid ligament (Walker 1973). In contrast, the other known Cretaceous dermochelyids, Corsocbefys balincbcs Zangerl 1960 and Mcsodcrmochclys unrlulatus Hirayama and Chitoku 1996, lack the tubercle. Quintan and Plisnicr-Ladame (1968) briefly described the limbs of the F.ocene dermochelyid Eospbarnis fligas Owen 1861, but did not note the presence or absence of an acromion tubercle. Nielsen (1964) described a crushed scapula of the other species of Eospbarjjis, E. breineri Nielsen 1959, but there was no mention of an acromion tubercle, possibly due to poor preservation of the specimen. The presence of an acromion tubercle in D. coriacca and UCMP 172070 suggests that UCMP 172070 may represent a species more closely allied to D. coriacca than to the other known Cretaceous dermochelyids. Description-VRT 19 (Fig. 2C), from Granite Bay (SC 1571), was reported but not described by Hilton and Antuzzi (1997). It is a left coracoid missing only the proximal articulation. It is 29.2 cm long, but was 2 or 3 cm longer when complete, and probably belonged to a turtle with a carapace length over 150 cm. Although elongate, VRT 19 is dorso-ventrally flattened. An important feature of VRT 19 is a distinct ridge occupying the medial half of the dorsal surface. In VRT 19 the ridge is 1.4 cm wide at its greatest width and 1.1 cm high. The ridge tapers distally as the rest of the element widens and assumes a dorso- ventrally flattened shape. Discussion—Coracoids of less-specialized sea turtles are distally expanded and extremely flattened dorso-ventrally. In clades that exhibit extreme pelagic specialization (i.e. 4 PALEOBIOS, VOL. 19, NUMBhR ,?, 1999 Fig. 3. A. UCMP 172071 (V94002), dermochelyid left humerus in ventral view. Scale bar= 1 cm. B. UCMP 172071 in dorsal view. Dermochelys and protostegids) the coracoid is elongate and rod shaped. VRT 19 represents an intermediate morphol- ogv. I'nfortunatclv, the eoracoids of earl\ dermochelyids are poorly known. In Corsochelys and E. breineri, only the proximal ends are preserved (Zangerl 1960, Nielsen 1964) and in both species the eoracoids are not as rod shaped as in Dermochelys. Quintart and Plisnier-Ladame (1968) figured a well-preserved specimen of E. jjijjcts, which shows the eoracoids of this species are more distally expanded than those of the living Dermochelys. Hirayama and Chitoku (1996) described in detail the eoracoids of Mesodermochelys and although VRT 19 is straighter than the eoracoids of Mesodermochelys (Hirayama and Chitoku 1996, fig. 10), it is similar in degree of elongation. According to Hirayama and Chitoku (1996) some speci- mens of Mesodermochelys have a modest ridge, but based on their figures it is considerably less developed than the ridge in VRT 19 and in Dermochelys (the condition in Eospharjjis has not been described). In extant dermochelyids, this ridge serves as a partial site of origin for a massive supracoracoideus muscle (Walker 1973, Manlius 1995). The supracoracoideus is the primary retractor and adductor of the humerus and its enlargement is probably correlated to increased power for swimming. Like the scapula (UCMP 172070), VRT 19 probably represents a dermochelyid more closely related to Dermochelys than to Mesodermochelys. It is likely that VRT 19 and UCMP 172070 represent a single undescribed species, bi t this cannot be confirmed without an associated pectoral girdle. In light of the morphology of VRT 19 and UCMP 1 "2070, additional description of the well-pre- served materia] of E.giflas figured in Quintart and Plisnier- Ladame (1968) is warranted. Description-UCMP 172071 (Fig. 3), from Dry Creek South (V94002), is a poorly preserved left humerus missing most of the proximal end. The specimen measures 8.8 cm in length, but may have been as long as 11 cm. The diaphvsis is dorsoventrally flattened, widens posteriorly, and is nearly straight. The lateral process is more distal than the caput and projects anteriorly approximately perpen- dicular to the axis of the diaphvsis. Discussion—The lateral process is more distal and ven tral relative to the condition seen in Cretaceous cheloniids (Zangerl 1953b). This character is found in pelagic-special- ized turtles (Hirayama 1992), such as protostegids and dermoche yids. A more refined identification is possible because, in contrast to protostegids, the lateral process of UCMP 172071 is not expanded along the length of the shaft, but is a prominent anterior protrusion. In this respect it is similar to the Cretaceous dermochelyid M. undulatns. Given the incompleteness of the specimen and the poor knowledge of Cretaceous dermochelyids, we withhold as- signment to that specific taxon. Still, there is enough evidence to suggest that UCMP 172071 represents a dermochelyid. If this specimen belongs to the same species as UCMP 172071 and/or VRT 19 then its relatively smaller size may indicate it is from a juvenile. cheix)nhdaf. Cray 1825 Description—VRT 32 (Fig. 4), from Granite Bay (SC 1571), is a 3.3 cm long fragment ol a supraoceipital from posterior to the foramen magnum. The distal part of the parietal si tures is visible. The crista supraoccipitalis flares out ventrally. In conjunction with the parietal suture this gives the supraoceipital an I-shaped cross-section posterior to the foramen magnum. The morphology of VRT 32 indicates i hat the skull had a relatively long parietal shelf reaching almost to the posterior end of the skull and that the area posterior to the parietal sutures was short ( -1 cm). Discussion—The height of the supraoceipital is low relative tc its length compared to modern cheloniids and dermochelyids. A low supraoceipital spine is consistent with a broad, flat skull, a character present in early North American cheloniids. Zangerl (1953b) placed these taxa into the family Toxochelyidae, but Fastovsky (1985) showed that this group was paraphyletic. We continue to use the n; me in quotes because "toxoehelyids" represent a significant grade or level of morphological evolution. In modern cheloniids and dermochelyids, and in Mesodermochelys (Hirayama and Chitoku 1996) the ventral edge of the supraoceipital spine is laterally compressed posterior to the foramen magnum. The low, ventrally PARHAM &¦ STIDHAM-CV£TACEOUS SEA TURTLES FROM CALIFORNIA 5 flattened supraoccipital spine can be used to identify VRT 32 as a "toxochelyid"-grade cheloniid. c:heloniidae:incertae sedis Ten additional specimens are identifiable as sea turtles, but cannot be referred to any subclade within the Chelonioidea. Hive of these are from Granite' Bay (SC 1571). VRT 25, 27, and 62 are costal fragments. VRT 26 is a poorly preserved limb fragment. VRT 23 is a partial scapula, but it is weathered and rounded so thi.t the pres- ence or absence of an acromion tubercle cannot be dis- cerned. The angle between the two prongs is estimated to be 90 degrees. This is less than the angle preserved in UCMP 172070, which may indicate that VRT 23 belongs to a different taxon, probably a cheloniid. Because of its small si/.e and poor state of preservation we do not refer it to a specific clade. The five remaining specimens are from the Dry ('reek area. UCMP 172072 is a costal fragment from V98117, while the remaining specimens are from V94002. UCMP 172074 is notable because of its small si/.e. Its preserved length is only 5 cm and in life was probably only 2 cm longer. Although we cannot assign UCMP 174074 to a subclade within Chelonioidea, this specimen indicates that, in addition to the large turtles represented by UCMP 172070 and VRT 19, the Chico fauna included extremely young individuals. UCMP 172075 is a maxilU. fragment. The relatively low tomial ridge is consistant with the mor- phology known from Cretaceous cheloniids, but since the maxillae of Cretaceous dermochelyids are unknown we do not assign it to either clade. UCMP 172073 is a damaged peripheral. UCMP 169162 is the base of a supraoccipital that lacks the noteworthy characters of VRT 32. DISCUSSION The Late Cretaceous Chico Formation of north-central California has yielded the oldest known cheloniid and dermochelyid remains from the F.astcrn Pacific region. Although they are commonly found together in Cenozoic sediments, the Chico material represents one of the oldest associations of these two taxa in the fossil record. The Chico Formation is contemporaneous with the early Campanian Mooreville Chalk of Alabama that yielded the sole specimen of the dermochelyid Corsochelys balinches Zangerl I960, as well as numerous cheloniids and protostegids (Zangerl 1953a, b, 1960). The record of Campanian chelonioids in the Pacific is poor. Nicholls (1992) referred a fragmentary specimen from the earliest Campanian Trent River Formation of Vancouver Island, British Columbia to the protostegid species Dcsmatocbclys lowi VVilliston 1894, but this assign- ment must be considered tentative. Hirayama and Hikida (1998) refer one specimen from the early Campanian Osoushinai Formation to Mcsodcrmocbclys undulatus. Fig. 4. VRT 32 (SC 1571), "toxochelyid" supraoccipital in right lateral view showing showing posterior edge of parietal-supraoc- cipital suture (arrow). Scale bar 1 cm. Maastrichtian turtles from the Pacific region are rela- tively better represented as fossils. For example, the cheloniid Ostcopyjjis Cope 1868 is known from the Moreno Formation of California (Foster 1980), as well as the Quinquina Formation of Chile (Gaspirini and Biro- Bagoczky 1986). Hirayama and Chitoku (1996) described M. undulatus from the lower Maastrichtian Hakobuchi Group in Japan. The description of Cretaceous dermochelyids from mul- tiple formations in the Western Pacific is interesting be- cause this group is rare in North America. In North America, Cretaceous dermochelyids are only represented by the type specimen of Corsochelys and the Chico material described here. More common in the Western Interior of North America, but absent in Japan, are cheloniids and protostegids. Hirayama and Chitoku (1994) are the first to note this taxonomic difference and refer to the Japanese chelonioid fauna as being dermochelyid dominated. Cheloniids are known from the Chico, but none of the Chico specimens can be identified as belonging to a protostegid turtle. Although Hilton and Antuz/.i (1997) referred VRT 19 to the Protostegidae based on a prelimi- nary identification by J.F. Parham (1997, personal commu- nication), this study shows it is more likely a dermochelyid. On the basis of the Chico fossils alone, it would be prema- ture to hypothesize that protostegids did not occur in the Eastern Pacific during the early Campanian. Nevertheless, the apparent absence of protostegids is consistant with an emerging pattern of taxonomic turnover in Pacific chelonioids. All Pacific chelonioid fossils that predate the Campanian are referable to primitive protostegids, non- protostegines in the sense of Hooks (1998). A specimen named Cratocbclone bcrncyi Ix»ngman 1915 from the late Albian of Australia compares well with protostegids (Gaffney 1981). Notocbeltmc costata Owen 1882, from the same horizon, is known from several specimens with skulls. Finally, Hirayama (1992) briefly mentioned a well-pre- served specimen of 1). lowi from the Middle Turonian of Japan. Hirayama (1997) and Hooks (1998) consider Notochelonc and Dcsmatocbclys to be basal protostegids. In contrast, none of the four Pacific localities dating from the Campanian and younger (Japan, Chile, and the 6 PALF.OBIOS, VOL. 19, NUMHLR .?, 1999 Chico and Moreno Formations of California) have yielded protostegid fossils of any kind. Temporally intermediate between the pre-Campanian protostegids and the Chico Formation is the aforementioned Vancouver Island speci- men referred to Desmatochelys (Nicholls 1992). The true taxonomic affinities of this specimen affect the timing of the faunal change. If dermochelyids and cheloniids replaced primitive protostegids in the Pacific, then it occurred some- time between the middle Turonian (the age of the Japanese Desmatochelys) and the early Campanian (the age of the Vancouver "Desmatochelys"). The history of dermochelyids prior to the early Campanian is unknown. Based on the oldest protostegid (Hirayama 1998), the dermochelyid lineage should extend back into the Late Aptian or Early Albian (110 Ma). The tact that dermochelyids appear simultaneously in the Farly Campanian (83-80 Ma) of Alabama, California, and Japan may indicate a rapid diversification from an undiscovered center of origin. The presence of cheloniid fossils in the Chico and Moreno Formations of California serves to differentiate the Eastern and Western Pacific chelonioid faunas. Cheloniids, well represented in North America in the Late Cretaceous, are absent from the Japanese Yezo Super Group (Hirayama and Chitoku 1994). Thus, the Chico Formation exhibits a unique mixture of North American ("toxochelyid") and Pacific (dermochelyid) chelonioid taxa, a biogeographic pattern that is mirrored in the Chico elasmobranch fauna Long et al. (1997). Jelctzky (1971) erected the Upper Cretaceous North Pacific Biotic Province in recognition of the marine faunal similarities between the Pacific coastal and marine faunas of North America and Asia. In particular, this similarity is seen in the ammonite fauna, with 12 of 13 Late Cretaceous Pacific baculitids present in both Califor- nia and Japan (Ward 1978), but absent in the Western Interior seaway. The occurrence of dermochelyids in Cali- fornia and Japan, as well as their absence from the Western Interior Seaway, may reflect this biogeographic provincial- ity. The presence of Corsochelys halinches in the Mooreville Chalk of Eastern North America suggests that primitive dermochelyids also inhabited the proto-Atlantic Ocean and that their absence from the Western Interior may represent a preference for habitats other than inland seas. Despite the fragmentary nature of the Chico dermochelyids it is possible to formulate a preliminary phylogenetic hypothesis. The presence of an acromion tubercle and a well-developed coracoid ridge suggest that the Chico species is/are more closely related to Cenozoic dermochelyids than to the other Cretaceous forms. The presence of at least three distinct lineages of dermochelyids in the early Campanian (Corsochelys, Mesodermochelys, and the Chico species) is in contrast to the single cosmopolitan species living today, but consistant with the multiple lin- eages of dermochelyids that persisted through much of the Cenozoic (Wood et al. 1996). Hirayama (1997) noted a similar pattern for all Cretaceous chelonioids and hypoth- esized tha: this was because they lacked the long distance migration patterns exhibited by some ot the living species. The limbs of Cretaceous and Paleogenc dermochelyids are- as specialized as modern cheloniids (Hirayama 1994), sup porting his idea that the provinciality of the fossil species is not the result of primitive locomotory capabilities. ACKNOWLEDGMENTS We are greatly indebted to Eric Gohre and Pat Antuzzi for their extra ordinary field effort and generous donation of vertebrate materials to the UCMP and Sierra College. Richard Hilton of Sierra College has and continues to play an important role in the study of Chico Formation paleon- tology. Without their work, these fossils would not have come to light, and this study would not have been possible. We thank Chris Bell and Howard Hutchison for providing comments that greatly improved the quality of this paper. We also tiank Pat Holroyd and Howard Hutchison for curatorial and collections assistance, and Jane Mason for skillfully preparing some of the material. This is UCMP Contribution No. 1703. LITERATURE CITED Creely, R.S. 1965. Geology of the Oroville quadrangle, Califor- nia. Bulletin of the California Division of Mines and Geology 184,86pp. Fastovsky, 'XE. 1985. A skull of"the Cretaceous chelonioid turtle Ostcopygis and the classification of the Osteopyginae. Investi- gations of the New Jersey State Museum 3, 28pp. Foster, O.K. 1980. Osteopygis sp., a marine tunic from the Late Cretaceous Moreno Formation of California. PaleoBios 34:1- 13. Gaffhey, E.S. 1981. A review of the fossil turtles of Australia. American Museum Novitatcs 2720:1-38. Gaflhey, E.S., and P.A. Meylan. 1988. A phylogeny of turtles, pp. 157-219 in M. J. Benton (ed.). The Systematics Associa- tion Special Volume No. 35A, The Phylogeny and Classification of Tetrapods, Vol. 1. Amphibians, Reptiles, and Birds. Clarendon Press, Oxford. Gaspirini, /'.., and I.. Biro-Bagoczky. 1986. Osteopygis sp. (Rep- tilia, Testudines, Toxochelyidae) tortuga fosil de la formacion Quirquina, Cretacico Superior, surde Chile. Revista Geohgica de Chile 27:85-90. May, O.P. 1908. The Fossil Turtles of North America. Carnegie Institute of Washington, Publication No. 75, 680 pp. Hilton, R.P., and P.J. Antuzzi. 1997. Chico Formation yields clues to Late Cretaceous paleocnvironmcnt in California. Cali- fornia Ccology 50:135-144. Hilton, R.P., E.S. Gohre, P.G Embree, and T.A. Sridham. 1999. California's first fossil evidence of Cretaceous winged vertebra es. California Geology 52:4-10. Hirayama, R. 1992. Desmatochelys lowi (Family Protostegidae) from Upper Cretaceous (Saku Fm., Middle Yezo Group; PARHAM & STJDHAM-CBETACEOVS SEA TURTLES FROM CALIFORNIA 7 Middle Turanian) of Yuubari, Hokkaido, North Japan. Ab- stracts of the 141th regular Meeting of the Palaeontological Society of Japan, p. 12. Hirayama, R. 1994. Phylogenetic systcmatics of chelonioid sea turtles. The Island Arc 3:270-284. Hirayama, R. 1997. Distribution and diversity of Cretaceous chclonioids. pp. 225-41 in J. M. Callaway and E. L. Nicholls (eds.). Ancient Marine Reptiles. Academic Press, New York. Hirayama, R. 1998. Oldest known sea turtle. Nature 392:705- 708. Hirayama, R. and T. Chitoku. 1994. Fossil turtles of Japanese Cretaceous marine strata, pp. 17-24 in I. Kobayashi (ed.). Evolution and Adaptation of Marine Vertebrate's, Vol. 2. Monograph of the Association for Geological Collaboration in Japan. Hirayama, R., and T. Chitoku. 1996. Family Dermochelyidae (Superfamily Chelonioidea) from the Upper Cretaceous of North Japan. Transactions and Proceedings of the Palaeontological Society of Japan 184:597-622. Hirayama, R., and Y. Hikida. 1998. Mesodermochelys (Testudincs; Chelonioidea; Dermochelyidae) from the Late Cretaceous of Nakagawa-cho, Hokkaido, North Japan. Bulletin of the Nakagawa Museum of Natural History 1:69-76. Hooks, G.E. 1998. Systematic revision of the Protostegidac, with a redescription of Calcarichelysgemma Zangerl, 1953. Journal of Vertebrate Paleontology 18(l):85-98. Jelctzky, J.A. 1971. Marine Cretaceous biotic provinces and paleogeography of Western and Arctic Canada. Geological Survey of Canada Paper 70-22, 92pp. Long, D., D. Lander, and E.S. Gohre. 1997. A new and diverse fossil elasmobranch assemblage from the Cretaceous Chico Formation of California. PaleoBios 18(supplcmcnt to l):6-7. Manlius, N. 1995. Description de la musculature anterieure de la tortue luth nouveau-nee, Dermochclys coriacea Vardelli (Che- Ionia, Dermochelyidae,). Societe Linneenne de Lyon 64(7):313- 345. Matsumoto, T. 1984. Some ammonites from the Campanian (Upper Cretaceous) of Northern Hokkaido. Palaeontological Society of Japan Special Paper 27', 93pp. Matsumoto, T., and I. Obata. 1963. A monogn.ph of the Baculitidae of Japan. Memoirs of the Faculty of Science, Kyushu University, Series D Geology 13:1-116. Nicholls, E.L. 1992. Note on the occurrence of the marine turtle Desmatochelys (Reptilia: Chelonioidea) from the Upper Creta- ceous of Vancouver Island. Canadian Journal of Earth Sci- ences 29:377-380. Nielsen, E. 1964. On the post-cranial skeleton of Eosphargis breineri Nielsen. Meddelelser fra Dansk Geologisk Forening, K0banhavn I5(3):281-313. Quintart, A., and F. Plisnier-I^dame. 1968. Eosphargis gigas (Owen, 1861). A propos d'unc tortue geantc recemment exposee dans las salles de l'Institut Royal des Sciences Naturelles de Belgique. Les Naturalistes Beiges 49(7):406- 413. Walker, W.F. 1973. The locomotor apparatus of testudines. pp. 1-100 in Carl Gans (ed.). Biology of the Reptilia, Volume 4, Morphology D. Academic Press, New York. Ward, P.D. 1978. Baculitids from the Santonian-Maestrichtian Nanaimo Group, British Columbia, Canada and Washington State, USA. Journal of Paleontology 52:1143-1154. Wicland, G.R. 1906. The osteology of Protostega. Memoirs of the Carnegie Museum 2(7):279-305. Wood, R.C., J. Johnson-Gove, E.S. Gaffney, and K.F. Maley. 1996. Evolution and phylogeny of leatherback turtles (Dermochelyidae), with descriptions of new fossil taxa. Chelo- nian Conservation and Biology 2(2):266-286. Zangerl, R. 1953a. The vertebrate fauna of the Selma Formation of Alabama. Part III: The turtles of the family Protostegidac. Fieldiana: Geology Memoirs 3(3):60-133. Zangerl, R. 1953b. The vertebrate fauna of the Selma Formation of Alabama. Part IV: The turtles of the family Toxochclyidae. Fieldiana: Geology Memoirs 3(4): 137-277. Zangerl, R. 1960. The vertebrate fauna of the Selma Formation of Alabama Part V: An Advanced cheloniid sea turtle. Fieldiana: Geology Memoirs 3(5):283-312. Zangcrl, R. and R.E. Sloan. 1960. A new specimen of Desmatochelys lowi Williston; A primitive cheloniid sea turtle from the Cretaceous of South Dakota. Fieldiana Geology 14(2):7-42.