Thomson, S.A., White, A. & Georges, A. (1997) Re-Evaluation of Emydura lavarackorum: Identification of a Living Fossil. Memoirs of the Queensland Museum (1997),42(1):327-336

Post-cranial osteological characters can be used to diagnose Australian short-neck chelid turtles to genus. Morphological examination of the Pleistocene fossil Emydura lavarackorum, from Riversleigh, shows that it is aligned with the genus Elseya not Emydura and should be referred to as Elseya lavarackorum (White & Archer, 1994). Furthermore the fossil specimen is not distinguishable from an undescribed extant form of Elseya from the Nicholson drainage, with which it shares one unique feature so this name should apply also to this extant form, identified to date only from electrophoretic data. It is Australia's first living fossil turtle, an extant population of a Pleistocene taxon.

Key Words: Chelonia, Chelidae, Pleistocene, fossil, turtle.

1. Applied Ecology Research Group and CRC for Freshwater Ecology. University of Canberra, Canberra, ACT, 2601

2. School of Biological Sciences, University of New South Wales, Kensington, NSW, 2033.

    The taxonomy of Australian chelid turtles is poorly known and in dire need of review (Cogger et al., 1983). Recent electrophoretic surveys (Georges & Adams, 1992; 1996) have revealed that in some instances, currently accepted species boundaries are difficult to justify and in others, what are currently regarded as single species are in fact two or more. The detailed morphological analyses required to verify these findings have not been conducted (but see Thomson & Georges, 1996), and until recently it was not possible to distinguish even between extant short-necked genera on the basis of osteological characters (Gaffney, 1977). This paucity of osteological data suitable for distinguishing the extant genera makes the identification of fossil forms, most of which are incomplete specimens, difficult. In many instances, chelid fossils have been assigned to either Chelodina or Emydura, with little or no evidence presented to eliminate the possibility that the short-necked forms among them may be Elseya, Rheodytes or Elusor.
    In 1994 a partial carapace and associated plastron from Riversleigh was described as a new species, Emydura lavarackorum, by White and Archer (1994). The fossil specimen was from Terrace Site, a fluviatile site on the Gregory River. These authors interpreted the sediments as being Pleistocene in age because of the presence of remains of Diprotodon optatum (White & Archer, 1994). The holotype consists of the anterior half of the carapace with some anterior peripherals and an essentially complete plastron with some pelvic material present. The length of the plastron is 390 mm (White & Archer, 1994) which corresponds to a carapace length of approximately 420 mm. Two other plastra from the same site were also collected but not described.
    White & Archer (1994) assigned the specimen to Emydura on the mode of the insertion of the anterior bridge into the ventral surface of the carapace. They found that in the derived state, the anterior bridge is angled steeply backwards towards the rib/gomphosis (called transverse process in White & Archer, 1994), whereas in all other chelids the anterior bridge was found to form a continuous line with the rib/gomphosis.

     In this paper, we reassessthe generic assignment of the fossil by comparing the fossil material with post-cranial character states we have found useful in separating extant genera of Australian short-necked chelid turtles. We also propose that the fossil form is extant, a distinctive, undescribed form closely aligned with Elseya dentata.

     Specimens of each of the short-necked species identified using electrophoresis by Georges and Adams (1996) were obtained from museums, the Conservation Commission of the Northern Territory and the University of Canberra. Where forms have not been included in published keys or descriptions, the specimens were selected from those lodged as vouchers to accompany the electrophoretic data. The specimen collection was supplemented by limited field sampling.
     Each specimen was skeletonised by removing excess soft tissue and feeding the remaining carcase to domestid beetles. The skeletal material was bleached in 5% sodium hypochlorate solution, and the process stopped by immersion in 100% ethanol. Plastra were separated from carapaces by disarticulating the plastral-carapacial suture between the hyo- and hypoplastra of the plastron and the lateral peripherals of the carapace. This was done by the carefully heating the carapace until the sutures become mobile and the plastron was then gently prised off. This also required disarticulation of the pelvis from the carapace. Characters potentially diagnostic at the generic level were examined to establish their consistency across a range of specimens within the polytypic genera Elseya and Emydura, and across a range of specimens within each species.
     The fossil specimens of Emydura lavarackorum were examined to determine the presence of character states which are generically diagnostic in extant taxa. The fossil specimen was then assigned to genus.