Download files
Abstract
A molecular phylogeny of marsupials based on a 20.1 kb nuclear and mitochondrial supermatrix, with associated divergence dates calculated using a Bayesian “relaxed” molecular clock approach, is presented. Monophyly of the currently recognised orders, Australidelphia and a clade comprising Dasyuromorphia, Notoryctes, and Peramelemorphia is strongly supported; the name Agreodontia is proposed for the latter clade. The molecular divergence dates indicate that the interordinal divergences within Marsupialia occurred during the Late Cretaceous and early Palaeogene, and that several groups within the Australian radiation radiated during periods of significant environmental change in the Miocene. The fossil record of didelphids and most Australian marsupial groups appears to be particularly incomplete.
Based on information provided by isolated petrosals and tarsals referred to this taxon, the early Eocene Australian metatherian Djarthia murgonensis appears to be the most plesiomorphic undoubted australidelphian and the oldest known crown-group marsupial from Australia. The presence of microbiotherians in South America may be the result of a back-dispersal from Australia or elsewhere in eastern Gondwana.
A qualitative functional analysis of the craniodental anatomy of the Oligo-Miocene Australian metatherian Yalkaparidon supports the hypothesis that this taxon was probably a “mammalian woodpecker”, similar to Daubentonia, Dactylopsila, and the extinct apatemyids. The craniodental anatomy of Yalkaparidon is described in detail, and isolated astragali and calcanea are tentatively referred to the genus. Yalkaparidon appears to be a member of Australidelphia but without close relationships to any other australidelphian order. The order Yalkaparidontia is maintained, and revised ordinal and species-level diagnoses are presented. A revised phylogenetic definition for Marsupialia is proposed.
Numbigilga ernielundeliusi, a bunodont metatherian from the early Pliocene of northeastern Queensland, is described, and the distribution of bunodont dental characters within Metatheria is reviewed.
An overview of Gondwanan metatherian biogeography is presented. It is argued that current evidence supports only limited dispersal of metatherians between South America and Australia during the Late Cretaceous or early Palaeogene, and that this was largely due to dispersal barriers (climatic and environmental) present within the Antarctic mainland. Future priorities for a better understanding of metatherian evolution and the origin of the modern Australian marsupial fauna are discussed.