Fossils and fossil climate: the case for equable continental interiors in the Eocene

Abstract

There are many methods for inferring terrestrial palaeoclimates from palaeontological data, including the size and species diversity of ectothermic vertebrates, the locomotor and dental adaptations of mammals, characteristics of leaf shape, size, and epidermis, wood anatomy, and the climatic preferences of nearest living relatives of fossil taxa. Estimates of palaeotemperature have also been based on stable oxygen isotope ratios in shells and bones. Interpretation of any of these data relies in some way on uniformitarian assumptions, although at different levels depending on the method. Most of these methods can be applied to a palaeoclimatic reconstruction for the interior of North America during the early Eocene, which is thought to be the warmest interval of global climate in the Cenozoic. Most of the data indicate warm equable climates with little frost. Rainfall was variable, but strong aridity was local or absent. The inferred palaeoclimate is very different from the present climate of the region and from model simulations for the Eocene. This suggests that models fail to incorporate forcing factors that were present at that time, that they treat the heat regime of continents unrealistically, and/or that model inputs such as sea surface temperature gradients or palaeotopography are incorrect.