Aspects of the results of studies of the fossil-rich Cainozoic deposits of Riversleigh, northwestern Queensland, are reviewed. A summary of five selected Riversleigh faunas representing the prirnaryperiods of the region's Cainozoic histoy is provided. Faunal and environmental changes over the last 25 000 000 years in the Riversleigh region are identified and changes in Australia 's rainforest mammal communities over the same period are discussed. Evidence for the origin of Australia's modem mammal groups from ancestors now known to have lived in the Tertiary rainforests of northern Australia is reviewed. The geological record for Riversleigh's more than 100 local faunas is considered. At least three primary intervals of Oligo-Miocene deposition, one of Pliocene and many of Pleistocene and Holocene deposition are identified. An appendix is provided in which the principal faunal assemblages from Riversleigh are allocated to these depositional intervals. The evidence for correlating Riversleigh local faunas with faunal assemblages in the rest of Australia and the world is reviewed. Oligo-Miocene, Pliocene and Pleistocene marsupial and monotreme fossds correlate Riversleigh's local faunas with others kom central and eastern Australia; bats correlate them with faunal assemblages in Europe; rodents correlate them with Pliocene assemblages in eastern Australia.

Constrained seriation of a species-locality matrix of the Australian Cenozoic mammal record resolves a preliminary sixfold succession of land mammal ages apparently spanning the late Oligocene to the present. The applied conditions of local chronostratigraphic succession and inferences of relative stage-of-evolution biochronology lead to the expression of a continental geological timescale consisting of, from the base, the Etadunnan, Wipajirian, Camfieldian, Waitean, Tirarian, and Naracoortean land mammal ages. Approximately 99% of the 360 fossil assemblages analyzed are classifiable using this method. Each is characterized by a diagnostic suite of species. An interval of age magnitude may eventually be shown to lie between the Camfieldian and Waitean, but is currently insufficiently represented by fossils to diagnose. Development of a land mammal age framework marks a progressive step in Australian vertebrate biochronology, previously expressed only in terms of local faunas. Overall, however, the record remains poorly calibrated to the Standard Chronostratigraphic Scale. Codifying the empirical record as a land mammal age sequence provides an objective basis for expressing faunal succession without resort to standard chronostratigraphic terms with the attendant (and hitherto commonly taken) risks of miscorrelating poorly dated Australian events to well-dated global events.

Resolving faunal responses to Pleistocene climate change is vital for differentiating human impacts from other drivers of ecological change. While 90% of Australia's large mammals were extinct by ca. 45 ka, their responses to glacial-interglacial cycling have remained unknown, due to a lack of rigorous biostratigraphic studies and the rarity of terrestrial climatic records that can be related directly to faunal records. We present an analysis of faunal data from the Naracoorte Caves in southeastern Australia, which are unique not only because of the species richness and time-depth of the assemblages that they contain, but also because this faunal record is directly comparable with a 500 k.y. speleothem-based record of local effective moisture. Our data reveal that, despite signifi cant population fl uctuations driven by glacialinterglacial cycling, the species composition of the mammal fauna was essentially stable for 500 k.y. before the late Pleistocene extinctions. Larger species declined during a drier interval between 270 and 220 ka, likely refl ecting range contractions away from Naracoorte, but they then recovered locally, persisting well into the late Pleistocene. Because the speleothem record and prior faunal response imply that local conditions should have been favorable for megafauna until at least 30 ka, climate change is unlikely to have been the principal cause of the extinctions.

No abstract

The Late Oligocene Kangaroo Well Local Fauna from the Ulta Limestone (new name), northwestern Lake Eyre Basin correlates best with vertebrate assemblages from the Etadunna, Namba and Wipajiri Formations of the central Lake Eyre Basin, and from the Carl Creek Limestone (Karumba Basin) of northwestern Queensland. The biochronologically informative marsupials, Neohelos tirarensis (Diprotodontidae, Zygomaturinae), Marlu sp. cf. M. kutjamarpensis and Pildra sp. cf. P. magnus (Pseudocheiridae), and Ektopodon ulta sp. nov. (Ektopodontidae), indicate that the Kangaroo Well Local Fauna may be slightly older than the Kutjamarpu Local Fauna (Wipajiri Formation) and slightly younger than the Ngama Local Fauna (zone D of the Etadunna Formation) of Late Oligocene age. A new species of primitive ?Wynyardiidae, Ayekaye jaredi sp. nov., is described, and the nomenclature of two extinct gastropods, Glyptophysa rodingae (McMichael) and Cupedora lloydi (McMichael) (new combinations), the type localities of which are in the Ulta Limestone, is revised in line with current taxonomy. The Ulta Limestone, an alluvial calclithite composed primarily of caliche fabrics, and its correlatives were deposited during the Miocene oscillation climatic event. Palaeoclimatic modelling using sedimentological data, crocodilians and extant analogs of fossil terrestrial gastropods indicates that the average annual temperature at Kangaroo Well during the Late Oligocene was probably between 14 and 20°C, while mean annual rainfall was probably <600 mm. Similar associations from central parts of the Lake Eyre Basin, from Riversleigh, northwestern Queensland, and from Bullock Creek, north‐central Northern Territory, indicate that such conditions were widespread during depositional phases of the Miocene oscillation. Palaeoclimatic indicators do not support the presence of widespread closed forests in northwestern Queensland and across the inland of the Northern Territory and South Australia during the Miocene oscillation.