The Antarctic Ice Sheet During the Late Neogene

Until recently the ice sheet in East Antarctica was believed to have changed little since first covering the continent in the late Middle Miocene. Recent discoveries of marine diatoms apparently of Pliocene age, and of plant remains, including wood of a Nothofagus species, associated with glacial sediments of the Sirius Formation at high elevations in the Transantarctic Mountains, have dramatically changed this picture. Some workers have inferred open-marine conditions in what are now subglacial basins on the East Antarctic craton throughout most of the Pliocene, ending when dropping temperatures allowed the ice sheet to occupy these basins, partly submerging the Transantarctic Mountains where it deposited the diatom-bearing Sirius Formation in late Pliocene to early Pleistocene time. Others point to the scarcity of evidence for prolonged warmth in the late Pliocene and favour restricting open-marine conditions in East Antarctica to the well-documented early Pliocene warm interval about 5-4 Ma, believing the species diachroneity of 1-2 million years accounts for the apparent age of the youngest diatom assemblage. However, evidence that long-distance ice-rafting, which implies the production of tabular bergs from a cold ice sheet, continued during the early Pliocene conflicts with the presence of open-marine embayments and the growth of vegetation, including tree species, at latitude 85° S, which imply that ice was temperate at sea level. Such conditions may have prevailed during a warm interval centred ~ 8 Ma, immediately before the welldocumented terminal Miocene refrigeration and glacial build-up and before the start of long-distance ice-rafting. This chronology is supported by the characteristics of the Nothofagus wood, which closely resembles that of one species in southern South America and one in Tasmania. Reasons are given for believing that the antarctic Nothofagus species was not introduced from South America during the Pliocene, but was an antarctic survival, probably related to the Tasmanian species and part of an endemic plant assemblage whose hardiest constituents were extinguished by the terminal Miocene refrigeration. A major obstacle for this hypothesis is that it requires a 4-5 22 million-year diachroneity of diatom assemblages.