Using palaeontological data to assess mammalian community structure: Potential aid in conservation planning

Abstract

Taxon-free metrics of biodiversity health are crucial for present and future conservation efforts in the face of current global change. We investigated the distribution of species in combined diet and body size functional groups over the past 16Ma in the Northern Great Plains to establish a pre-Holocene (before 11,000ya, when humans arrived in North America) and pre-industrial baseline (11,000–500ya) of this measure of community structure. Functional group distributions were compared on two scales to gauge the impact of time-averaging on patterns of community structure change: 1) North American Land Mammal Ages (NALMAs), and 2) individual diverse localities. Distributions were statistically compared using pairwise Fisher's exact tests with Monte Carlo P-value simulations and Holm P-value adjustment, and qualitatively assessed using correspondence analysis. When averaged over entire NALMAs, major changes in functional group distribution only take place in the Hemphillian, and at the start of the Holocene. Locality-level patterns also indicate long periods of stasis in the metric (Barstovian–Clarendonian and Hemphillian–Holocene). A threshold of global climate change is one possible explanation for the change that began in the Hemphillian, but further study is needed in that regard. Extinction of megaherbivores (>44kg) is the primary driver of apparent differences between the Holocene and previous time periods. Although the extent of time-averaging and other taphonomic biases affect the details of observable patterns per time period, overall, proportional diversity of functional groups is a promising metric for assessing mammalian community health because it is remarkably stable through time and changes only with major external perturbations to ecosystems.