Leaellynasaura

Abstract: Early Cretaceous (late Aptian–early Albian) fossil localities in the south coastal region of Victoria, southeastern Australia, provide a snapshot of the rich terrestrial biota that lived on the floodplain within the extensional rift system that existed between the eastern Gondwanan landmasses of Australia and Antarctica. Amongst the terrestrial vertebrate assemblage, agile, bipedal, small-bodied (‘turkey’- to ‘ostrich’-sized) ornithopod dinosaurs have been considered abundant and arguably diverse. Although the Victorian ornithopods potentially hold significant information on Gondwanan dinosaur phylogeography, their taxonomy and phylogenetic systematics have been poorly understood.This research provides a sedimentological and taphonomic reassessment of the Leaellynasaura amicagraphica holotype locality at Dinosaur Cove in the Eumeralla Formation (Otway Group). The osteology and systematics of the three previously named Victorian ornithopods (Atlascopcosaurus loadsi, Leaellynasaura amicagraphica and Qantassaurus intrepidus) are revised, and potential new taxa recognised. The postulated diversity of the Victorian ornithopods is reassessed and their phylogenetic and phylogeographic relationships are investigated.Contrary to the original interpretation of a quiet-water depositional setting (such as a billabong), the holotype locality of L. amicagraphica is reinterpreted to be a point-bar deposit. Fossil accumulation in the heterolithic cross-strata of the point-bar is considered to be allochthonous. As a result, the referral of additional specimens to L. amicagraphica, other than materials directly attributable to the holotype, cannot be adequately supported. In addition, newly identified material of the L. amicagraphica holotypic cranial fragment (MV P185991) provides important new information on this taxon, and facilitates anatomical and taxonomic differentiation of the cranial table (MV P185990), originally considered to be from the holotype individual.The validity of L. amicagraphica (Eumeralla Formation) and Q. intrepidus from the Wonthaggi Formation (Strzelecki Group) is upheld. However, Atlascopcosaurus (Eumeralla Formation) is considered to be a junior subjective synonym of Leaellynasaura (L. loadsi nov. comb. [Rich & Rich, 1989]). A third species of Leaellynasaura (Leaellynasaura sp. nov.) extends the range of this genus to the Wonthaggi Formation. Two premaxillary, one maxillary (Leaellynasaura sp. nov.) and two dentary morphotypes (including Q. intrepidus) are recognised from the Wonthaggi Formation, whilst four maxillary morphotypes (including two species of Leaellynasaura), two dentary and four postcranial morphotypes are recognised from the Eumeralla Formation. Although taxonomic differentiation of these morphotypes has not been ascertained, it is possible that some pertain to the same taxon. The five maxillary morphotypes, provide an overall measure of the diversity of theVictorian ornithopod assemblage.Of the postcranial morphotypes, three partial postcranial skeletons are recognised from the Eumeralla Formation: two from Dinosaur Cove (Victorian Ornithopod Postcranial morphotypes [VOPC] I and II), both previously attributed to L. amicagraphica, and the third (VOPC III) from the locality of Eric the Red West. VOPC I and II are differentiated from each other on the basis of their haemal arch morphology. VOPC I differs from all other ornithischians on account of its extreme taillength (~13 times femoral length), while the hind limb proportions of VOPC II suggest a highly cursorial ornithischian. VOPC III appears to be a more robust morphotype than both VOPC I and II. The femur of the fourth postcranial morphotype, Victorian ‘Hypsilophodontid’ Femur Type 2 of Rich & Vickers-Rich (1989), differs morphologically from the femora of VOPC I and II, and represents a larger more robust body-type. The morphological differences evident in both the maxillary and the postcranial morphotypes suggest differing habitat preferences among the ornithopods from the region.Cladistic analysis using implied weighting methodology suggests that Leaellynasaura and the postcranial morphotypes VOPC I and II are most closely related to Australia’s Muttaburrasaurus and Patagonia’s Anabisetia and Gasparinisaura. It is hypothesized that these taxa form a Gondwanan non-iguanodontian ornithopod clade that diverged from a Pangaean progenitor of Iguanodontia through vicariance during Pangaean fragmentation in the Late Jurassic. Interestingly, the Laurentian Parksosaurus and Thescelosaurus may be more closely related to Muttaburrasaurus and Patagonia’s Talenkauen than to basal ornithopod taxa from Asia. The hypothesised relationship between these non-iguanodontian ornithopods from Laurentia and Gondwana may stem from the opening of an Antillean dispersal route between North America and South America during the Late Cretaceous. Although Qantassaurus was not included in the cladistic analysis, its dental morphology resembles that of Kangnasaurus from the Late Cretaceous of South Africa. In addition, the proportions of metatarsal II of VOPC II and III more closely resemble those of dryosaurids (Dryosaurus, Dysalotosaurus and Kangnasaurus), Anabisetia and Gasparinisaura. These similarities either suggest convergence between these taxa, or a closer relationship between dryosaurids and Gondwanan non-iguanodontian ornithopods than suggested by the results of the cladistic analysis.The Victorian ornithopods currently represent the oldest body-fossils of non-iguanodontian ornithopods from Gondwana. It is speculated that the high diversity of this key ornithopod fauna from Victoria is linked to the geologically dynamic, high-latitude floodplain palaeoenvironment of the Australian-Antarctic Rift System, and, tantalizingly, the early radiation of angiosperms in eastern Gondwana.

Abstract: The holotype individual of the small-bodied ornithopod dinosaur, Leaellynasaura amicagraphica from Dinosaur Cove in the Lower Cretaceous of Victoria, southeastern Australia, traditionally comprises the holotype, a left-side cheek fragment of a juvenile (MV P185991), and three other specimens: a cranial table (P185990) and a partial postcranium (P185992, P185993), discovered at the same site and at about the same time as the holotype. The latter three specimens have significantly contributed to the systematics of Leaellynasauria amicagraphica and anatomical arguments for its status as a “dinosaur of darkness,” pre-adapted to existence in the Antarctic polar circle. The original attribution of the scattered material (cranial table and partial postcranium) to the Leaellynasaura amicagraphica holotype was based on the assumption that the sizes of the specimens were comparable, and the interpretation of the facies in which these associated fossils accumulated as a quiet-water deposit, such as an oxbow lake, billabong or pond. The inferred low-energy depositional conditions were used to suggest that associated material, other than that attributable to the holotype, was unlikely to be present in the facies hosting the holotype individual. However, a detailed sedimentological study supporting the interpretation of a quiet-water deposit hosting the Dinosaur Cove material is lacking, and the presence of a larger second partial ornithopod postcranium (P186047) in the same deposit, seems contradictory to arguments that all of the scattered associated skeletal specimens from this site are attributable to the Leaellynasaura amicagraphica holotype. Our revised sedimentological investigation indicates that all vertebrate remains from the Leaellynasaura amicagraphica holotype locality were deposited under active hydraulic flow on a migrating point bar in a meandering river. We term the host deposit the “Tunnel Sandstone.” As a result of this new interpretation, we regard the total vertebrate fossil assemblage from this site as time-averaged, and interpret the associated ornithopod remains as an allochthonous accumulation of up to four separate individuals, some potentially with unknown taxonomic affinities. Without unequivocal anatomical evidence of skeletal association, we regard the traditional attribution of the scattered cranial table and partial postcranium to the Leaellynasaura amicagraphica holotype as inadequately supported. We consider the referral of any specimen to Leaellynasaura amicagraphica should contain features that are compliant with those features on the holotype cheek fragment or other conclusively referred specimens.