Testing coevolutionary hypotheses over geological timescales: interactions between Cretaceous dinosaurs and plants

Abstract

Testing coevolutionary scenarios over extended geological timescales is fraught with difficulties. Most tests rely on comparisons of temporal variations in taxonomic diversity for the groups of interest: however, this approach typically excludes spatiotemporal data. Here, we apply a quantitative method that incorporates the spatiotemporal distributions of the proposed coevolving groups using a Geographical Information System. Distributional data for Cretaceous dinosaur and plant groups were mapped onto palaeogeographical reconstructions in a series of time-slices. Within each time-slice, palaeocontinental surfaces were divided into a series of grids, each of which was scored as present, absent or inapplicable (unsampled) for each group. Distributions were compared statistically to determine whether the putative coevolving groups co-occurred within grid squares more or less frequently than expected by chance. Pairwise comparisons were made between herbivorous dinosaur clades and major plant groups (e.g. cycads, angiosperms) on a global scale. Only three nonrepeated associations of marginal significance were recovered, demonstrating that, in general, current knowledge of the spatiotemporal distributions of these groups provides little support for coevolutionary hypotheses. The Geographical Information System methods used are readily applicable to many other questions whose answers are reliant on a detailed knowledge of organismal distributions in time and space. © 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 100, 1–15.