Using the fossil record to define natural biodiversity baselines in mammals

As global climate change and other human impacts accelerate ecological changes in the coming century, it is becoming more and more necessary to use the fossil record to define ecological baselines such that the true magnitude of future changes comes into focus. Here we examine the natural baseline of continental-level species richness in fossil mammals of North America, using data assembled from the MIOMAP and FAUNMAP databases of fossil mammal occurrences and associated geological and chronological information. We use rarefaction and other sample standardization techniques in combination with the Berkeley Mapper software to collate occurrence data into species-area curves for a variety of time-slices that vary in age from Miocene through Holocene. The fossil species-area curves are then compared with species-area curves derived from modern mammals, as obtained from zoology museum databases and subsampled to simulate fossil occurrences. Preliminary results demonstrate that for some geographic areas, notably the Great Plains, the Holocene species richness is depressed, suggesting a region-wide drop in species diversity as humans became more dominant in the ecosystem. We also identify differences between modern and pre-Holocene curves, although presently we are still trying to ascertain whether those differences are due to analytical artifacts or represent biological reality. By using the combination of fossil species area curves and modern ones we identify a range of variability that should bracket the extremes of species richness patterns that characterize landscapes little impacted by humans. Present or future species-area relationships that fall outside those limits indicate the extent to which human activity has perturbed natural systems.