Modelling late Oligocene C4 grasses and climate

Evidence suggests that C-4 grasses, adapted to conditions of low CO,, high temperatures, and water-stressed environments, expanded relatively quickly, and in several geographically distinct regions, during the Late Miocene (approximately 9-7 Ma). Using a variety of modelling tools, we attempt to explain this expansion in terms of changes in global climate. We use a fully coupled atmosphere-ocean General Circulation Model (AOGCM, HadCM3L) to simulate global climate before and after the expansion (the Late Oligocene, 27 Ma, and pre-industrial). We then use the simulated climates to drive a process-based global vegetation model (LPJ). Contrary to indications from palaeo data, the vegetation model predicts a greater C4 distribution during the Late Oligocene than the pre-industrial. The principal reason for this is that the relatively high tropical temperatures which we predict for the Late Oligocene confer a competitive advantage on C4 grasses over C-3 grasses which outweighs the disadvantage imposed by the higher CO2 concentrations. We find that this result is robust to prescription of different CO, levels in the Late Oligocene AOGCM simulation, and to the vegetation model used. Of the several possible explanations proposed for the modeldata discrepancy, three seem most likely. First, that the vegetation models do not adequately represent certain processes under extreme climate conditions, such as fire disturbance, or key parameters related to the competition between C-3 and C-4 grasses. Second, that there existed an evolutionary niche for C-4 grasses prior to their expansion in the Late Miocene, which for some reason was not filled. Third, that we have not adequately modelled the Late Oligocene climate with the AOGCM. Future research should concentrate on the collection and analysis of additional data for the assessment of palaeo tropical land temperatures, and on further testing of the sensitivity of the AOGCM results. (c) 2007 Elsevier B.V. All rights reserved.