Palaeaspis conybearii ( Owen, 1849 ) Gray, 1870

GAFFNEY, EUGENE S, TONG, HAIYAN & MEYLAN, PETER A, 2006, EVOLUTION OF THE SIDE-NECKED TURTLES: THE FAMILIES BOTHREMYDIDAE, EURAXEMYDIDAE, AND ARARIPEMYDIDAE, Bulletin of the American Museum of Natural History 300 (300), pp. 1-698 : 109-113

Palaeaspis conybearii ( Owen, 1849) Gray, 1870

Palaeaspis bowerbanki ( Owen, 1842)

Gray, 1870

TYPE SPECIMEN: There may be more than one taxon involved in this series of names, as suggested by Broin (1977: 49), although Williams (1954a) synonymized them all as Palaeaspis conybearii . The oldest of the species is Platemys bowerbanki Owen, 1842 , with a type that is apparently lost ( Williams, 1954a), but the genus Palaeaspis was named by Gray (1870) for Emys conybearii Owen, 1849 (in Owen and Bell, 1849) with BMNH R39449 as a type. Both are partial shells.

DISCUSSION: The Early Eocene London Clay of England has yielded a series of shells ( Clouter et al., 2000) identifiable as Pelomedusoides, which were reviewed by Williams (1954a), who concluded that there was one species, correctly named Palaeaspis conybearii (Owen) . Broin (1977: 48–49) commented on the material and compared it with Taphrosphys , noting that the surface texture and position of the pectoral/abdominal sulcus were similar. However, she concluded that the shell material was poorly described and probably included more than one taxon. A partial skull, BMNH 38953 ( Owen, 1850: pl. 29, figs. 1, 2) from Sheppy, Kent, figured and described by Owen (1850) as Platemys bowerbanki , was suggested by Broin (1977: 49) as possibly Erymnochelys . However, Ren Hirayama has recently noted (personal commun.; we are also grateful to Sandra Chapman for assistance) that the skull has a processus trochlearis oticum and appears to be a carretochelyid. Broin (1988: 138) placed ‘‘ Palaeaspis bowerbanki ( Owen, 1842) ’’ as possibly being part of Neochelys Bergounioux, 1954 without comment. Lapparent de Broin (2001: 169) placed Palaeaspis Gray, 1970 in the Bothremydidae , also without comment, and we cannot find a previous reference explaining either of these attributions. On the basis of the described shell material (and the BMNH material that we have seen), it is not possible to assign any of the separate (or synonymized) species of the London Clay pleurodire/pleurodires to a family.

CURRENT STATUS: Pelomedusoides incertae sedis.

‘‘ Podocnemis ’’ somaliensis Walker, 1966

TYPE SPECIMEN: Partial shell, Sedgwick Museum, Cambridge, C 54.276 ( Walker, 1966).

DISCUSSION: This nearly complete shell from the Eocene of Somalia has the sutural pattern preserved, but apparently lacks sulci impressions. Entering it in our dataset produces a whopping 87 % missing data, but the miracle of our ability to analyze meager information resolves ‘‘ Podocnemis ’’ somaliensis with the Bothremydidae , a conclusion apparently consistent with that of Lapparent de Broin (2000a), who listed it as ‘‘ Bothremys somaliensis ( Walker, 1966) .’’ It is not possible to assign this shell to a genus, however, and its recognition as a bothremydid must be considered tenuous at best.

CURRENT STATUS: Bothremydidae incertae sedis.

‘‘ Podocnemis ’’ parva Haas, 1978a

TYPE SPECIMEN: A nearly complete shell, Hebrew University of Jerusalem, Israel, Department of Zoology collection, HUJP-Testudinata-3 ( Haas, 1978a).

DISCUSSION: See ‘‘ Podocnemis ’’ judea.

‘‘ Podocnemis ’’ judea Haas, 1978b

TYPE SPECIMEN: A nearly complete shell, Hebrew University of Jerusalem, Israel, Department of Zoology collection, HUJP 3664.

DISCUSSION: Haas (1978a, 1978b) described a series of small shells from the Cenomanian of Israel. The criteria differentiating two species among these shells seem to be within the variation seen in some recent pleurodire species. Apparently Broin (1988) and Lapparent de Broin and Werner (1998) came to the same conclusion, as they only list the older species as valid. They also place parva in Bothremys . When parva/judea is entered in the dataset (83 % missing data) and analyzed in the larger dataset, it resolves as the sister taxon to Foxemys + Polysternon . This is a little surprising considering the large amount of missing data, but the few characters available produce this single cladogram. However, this material has not been reexamined in recent years, and a new assessment of the character distributions is needed to clarify some of the characters that are inconsistent in the figures. It is necessary to leave parva/judea without a generic assignment and place it incertae sedis within the Bothremydini until it is better known.

CURRENT STATUS: Bothremydini incertae sedis.

‘‘ Taphrosphys ’’ olssoni ( Schmidt, 1931)

TYPE SPECIMEN: FMNH P14172, partial shell (fig. 267).

DISCUSSION: This partial shell was described by Schmidt (1931) from the Eocene of Peru as Podocnemis olssoni . Zangerl recognized its similarities to Taphrosphys and changed it to Taphrosphys olssoni . This was accepted by Gaffney (1975a), who provided comparative diagnoses for T. olssoni and T. sulcatus (table 24). When entered into the dataset, this taxon comes out in a multichotomy with all members of the Taphrosphyini . We remove the species olssoni from Taphrosphys .

CURRENT STATUS: Taphrosphyini incertae sedis.

‘‘ Taphrosphys ’’ ambiguus ( Gaudry, 1890)

TYPE SPECIMEN: MNHN-MTA1, a plastron (fig. 267).

DISCUSSION: A plastron from the Paleocene of France was named Tretosternum ambiguum by Gaudry (1890:) 251, fig. 355 and reassigned to Taphrosphys by Broin (1977), who redescribed and figured it ( Broin, 1977: fig. 4, pl. 4, figs. 9, 10). ‘‘ Taphrosphys ’’ ambiguus has enough characters to enter into the dataset, which, when analyzed, shows it in a multichotomy with all Taphrosphyini . The available material is insufficient to place in a genus (table 24).

CURRENT STATUS: Taphrosphyini incertae sedis.

‘‘ Taphrosphys ’’ miocenica Collins and Lynn, 1936

TYPE SPECIMEN: USNM 13784, anterior lobe of a plastron.

DISCUSSION: This partial plastron from the Miocene Calvert Formation of Camp Roosevelt, Maryland, was named Taphrosphys miocenica ( Collins and Lynn, 1936: pl. 1). The assignment to Taphrosphys was rejected by Gaffney and Zangerl (1968: 208) because Taphrosphys is characterized by a large intergular scale separating the gular scales, humeral scales, and part of the pectoral scales. In ‘‘ Taphrosphys ’’ miocenica the intergular separates only the gulars and part of the humerals, as in Bothremys . Gaffney and Zangerl therefore assigned ‘‘ T.’’ miocenica to Bothremys . However, Gaffney (1975a) rejected this generic assignment, because the Bothremys scale pattern also occurs in Podocnemis and other taxa and is inadequate for a generic determination. ‘‘ Taphrosphys ’’ miocenica was therefore made a nomen dubium. The Bothremys scale pattern, in fact, occurs widely in Pelomedusoides, within Bothremydidae as well as Podocnemididae . It is notable that Bairdemys (fig. 275; Gaffney and Wood, 2002; Wood and Díaz de Gamero, 1971) has an anterior plastral lobe nearly identical in size, shape, and scale arrangement to that in ‘‘ Taphrosphys ’’ miocenica . Bairdemys has a short, rounded, anterior plastral lobe (fig. 275), as in many bothremydids. Bairdemys occurs in the Miocene of Venezuela and Puerto Rico. However, the group that it belongs to within the Podocnemididae , the Shweboemys Group, is known throughout the Caribbean ( Domning and Clark, 1993 [this record contains the posterior end of a lower jaw that is a podocnemidid]; Domning et al., 1997; Sánchez-Villagra et al., 2000; Gaffney and Wood, 2002). Furthermore, an undescribed lower jaw in the South Carolina State Museum (SCSM SC90.16.24) from the Oligocene of South Carolina is very similar to the lower jaw of Bairdemys from Venezuela (we are very grateful to Dr. R. Weems for bringing this specimen to our attention). The fact that Bairdemys and its near relatives are now known to occur in the mid-Tertiary of the Atlantic coast as well as the Caribbean, and that the morphology of ‘‘ Taphrosphys / Bothremys ’’ miocenica is so similar to that group, suggests that this specimen could easily be a podocnemidid and should not be used as a mid-Tertiary record of the Bothremydidae . The anterior plastral lobe is too incomplete to enter into the dataset.

CURRENT STATUS: Pelomedusoides incertae sedis.

Taquetochelys decorata Broin, 1980

TYPE SPECIMEN: MNHN-GDF 847, a right hypoplastron ( Broin, 1980: pl. 3, fig. 10).

DISCUSSION: These shell fragments from the Aptian of Gadoufaoua, Niger ( Broin, 1980) , have a surface texture similar to that of Araripemys . The texture is not unique to Araripemys , however. The type alone is not diagnosable as a distinct taxon. It is unique only if locality and age are considered. The described fragments are inadequate to show a sister-group relationship to Araripemys . There are too many missing data to enter into our dataset. If new articulated specimens from the same locality become available, this taxon might become diagnosable.

CURRENT STATUS: Pleurodira incertae sedis.

TAXA NOMINA DUBIA Apertotemporalis baharijensis Stromer, 1934

TYPE SPECIMEN: BSP uncataloged, now lost ( Crumly, 1984).

DISCUSSION: This partial skull from the Late Cretaceous of Egypt was described and figured by Stromer (1934: plate 1, fig. 1a–c) but was apparently later destroyed in World War II ( Crumly, 1984) with much of the Munich collection. The figures show a turtle skull, which Antunes and Broin (1988), Broin (1988), Lapparent de Broin and Werner (1998), and Lapparent de Broin (2000a) assigned to the Bothremydidae . Antunes and Broin (1988) based this on ‘‘Le dessin du reste de crâne montre une morphologie et les dimensions compatibles avec Bothremys et Rosasia …’’ ( Antunes and Broin, 1988: 179). In the absence of the original, the only information on ‘‘ Apertotemporalis ’’ is the description and figures in Stromer (1934), which show a badly eroded and damaged partial skull of something that would probably be difficult or impossible to objectively diagnose even if the original were available. Even if new, complete material were to be found at the type locality, identifying it with the type figures would be very argumentative. The only apparent feature of the skull from the figures is a fully enclosed incisura columellae auris, consistent with the Bothremydidae , but also found in other groups. In fact, an Egyptian bothremydid skull, Arenila , has been found and it could very well be the same species as ‘‘ Apertotemporalis ,’’ but this is impossible to determine given the absence of a type specimen. The figures alone are inadequate to diagnose the taxon, and in the absence of a type specimen, the taxon should be ignored.

CURRENT STATUS: Nomen dubium.

Carteremys leithi ( Carter, 1852)

Williams, 1953

TYPE SPECIMEN: None designated in original description of species ( Carter, 1852); no specimens seen by subsequent authors; present whereabouts unknown. Williams (1953) based his diagnosis of a new genus on the figures of Carter (1852), which consist of two plates showing a reconstruction based on nine partial specimens. The material figured shows a carapace, plastron, partial skull, and partial lower jaws. If available, this material would presumably allow an adequate diagnosis, but there is no evidence that this material has been available since 1852.

LOCALITY AND HORIZON: Possibly Maastrichtian or Paleocene Intertrappean beds of Bombay ( Mumbai), India ( Wood, 1970).

DISCUSSION: ‘‘ Testudo ’’ leithii Carter was originally recognized as a pleurodire by its author, despite being referred to ‘‘ Testudo ’’. Subsequently, Gray (1871) assigned the species to the chelid genus, Hydraspis , which was questioned by Williams (1953), who suggested that the original figures showed a mesoplastron and that the skull looked like Stereogenys , a podocnemidid. However, the original description and Williams’ diagnosis of the genus are inadequate to diagnose this taxon. The original figures are reconstructions. No figure of the original material exists, so that there is no possibility of designating a lectotype on the basis of the figures, as the original figures are reconstructions. If at some future date Carter’s original, properly labeled material becomes available, then this taxon might be resurrected.

Wood (1970) added further arguments that C. leithii is a pelomedusid (sensu lato, now equals Pelomedusoides) and concurred in recognizing Carteremys as a diagnosable genus. Jain (1977) added another species, C. pisdurensis , an undoubted Pelomedusoides later reassigned to the genus Shweboemys ( Jain, 1986) . We concur that this species is a podocnemidid and not a bothremydid. Singh et al. (1998), in the initial announcement of the skull that was later named Sankuchemys , also considered Carteremys a nomen dubium.

Williams (1953) also figured a carapace from the Intertrappean beds of Worli Hill, Bombay ( Mumbai), that he identified as Carteremys leithii , and which was apparently found in the 1940s by Dr. R. N. Sukheswala. This carapace was figured and described by Singh et al. (1998), along with a plastron anterior lobe. Williams (1953) apparently identified this carapace as Carteremys on the basis of its locality, small size, and surface texture. The association of the plastron with the carapace may not be original, as Williams (1953) specifically stated that the plastron and skull were missing. However, the plastral lobe is of the right size and apparently is from the same locality. The plastron shows very small gular scales with a large pair of intergular scales completely separating the humerals, similar but not identical to the figure in Carter (1852). Therefore, it is quite possible that this shell material could be the same species that Carter (1852) described, and that this is the shell of the skull described as Sankuchemys . However, even if the new shell material were accepted as the shell of Sankuchemys , tenuous at best, it is even harder to argue that these shell specimens are the same species as those described by Carter in 1852.

CURRENT STATUS: Nomen dubium.

Crassachelys neurirregularis

( Bergounioux, 1952)

TYPE SPECIMEN: None designated by original author of Gafsachelys neurirregularis Bergounioux, 1952 . A neotype designated by Moody (1972) is a carapace fragment in plate 46, figure 2 of Bergounioux (1952).

DISCUSSION: In 1972, Moody used the neural number to recognize three genera in the Tunisian Eocene Pelomedusoides. Two of these, Gafsachelys and Eusarkia, had been named, so he added a third for Gafsachelys neurirregularis . In Moody’s scheme, Gafsachelys Stefano, 1903 is restricted to seven neural bones in contact ( Moody, 1972: pl. 16); Crassachelys Moody 1972 is characterized by smaller neurals, five in number, some separated by costals meeting on the midline ( Bergounioux, 1952: pl. 46, figs. 4, 5; 1956: figs. 22, 23, pl. 10, fig. 2, pls. 11–13; Moody, 1972: pl. 16, fig. 2, pl. 17); and Eusarkia Bergounioux, 1952 is characterized by an absence of neurals ( Bergounioux, 1952: pl. 46, figs. 1–3; 1956: fig. 20; Moody, 1972: fig. 3). Broin (1977) and Antunes and Broin (1988) have disputed the use of this character, ascribing it to individual variation and lumping all three genera into Taphrosphys . Broin (1988) and Lapparent de Broin (2000a) also synonymized Crassachelys with Taphrosphys . However, the neotype designated by Moody (1972) consists only of a central part of a carapace showing discontinuous neurals and a Taphrosphys - like texture. Given the variations in the neural pattern in North African Paleogene Pelomedusoides, this neotype must be considered inadequate for a diagnosable taxon.

When dealing with this large suite of Tunisian shells, the Moroccan specimens must be kept in mind. Broin (1988) and Lapparent de Broin (2000a) synonymized all these Tunisian taxa as Taphrosphys phosphaticus . Aside from the designation of phosphaticus as a nomen dubium, there is a problem using shell characters alone to diagnose Taphrosphys in light of the new discoveries in Morocco. The Moroccan fauna has yielded a diverse group of taxa related to Taphrosphys . The discovery of a skull-shell association for Ummulisani shows that many shell characters used by earlier authors to diagnose Taphrosphys have a wider distribution, perhaps wider than the tribe Taphrosphyini . Although there are a number of shells known from the Tunisian phosphates, most are steinkerns. Complete shells with extensive bone preserved are relatively rare, so analyzing the possible correlation of neural number with other characters is difficult. Therefore, the identification of alpha taxa among the Tunisian shells is problematic, especially in the absence of any skull-shell associations. It might be possible to recognize some of the best preserved specimens as tribe Taphrosphyini incertae sedis, but unfortunately this excludes most of the type specimens.

CURRENT STATUS: Nomen dubium.