Transgressing the limits of palaeoenvironmental data for Southeast Asian Pleistocene faunal assemblages: A critical review to go further.

For the Pleistocene of South-east Asia, the 'savannah corridor' hypothesis put forward by Lawrence Heaney in 1991 on the basis of a few pollen records rightly aroused a certain amount of enthusiasm and was the source of many attempts to confirm or refute a significant regression in forest cover, depending on the taxa considered, during the Last Glacial Maximum. The work carried out to verify this hypothesis used a number of tools, including palaeontology, the results of which seem to us to have been greatly embellished given the paucity of reliable data available. Indeed, it is clear that the hypothesis concerns a geographical area that is largely inaccessible today, located at a depth of 120 meters in the South China Sea, and that the faunal assemblages taken into account remain imperfectly defined to allow such a hypothesis to be tested. We provide a summary of the main work carried out to date and a critical assessment of the articles published over the last few decades. It emerges that the fossil series considered (essentially mammals) are neither well chronologically set nor well defined from a taphonomic point of view, notably because they are based on unsatisfactory references, which are themselves poorly dated or based on mixtures of different faunas. These shortcomings also call into question the renewed and attractive hypothesis of the flexibility of the diet and change of habitat (ecological niche) of certain taxa, based on isotopic analyses, as recently proposed. We advocate the implementation of systematic fine excavations that take into account the taphonomy of the sites, but also show that progress has been made to go further. [ABSTRACT FROM AUTHOR]