Richness, Population Structure, and Habitat-Specific Sampling Patterns of Yellowstone National Park's Large-Mammal Death Assemblage: Paleoecological and Neontological Implications

Paleoecological interpretations of fossil deposits rely on a detailed understanding of the taphonomic biases associated with skeletal assemblages – but the processes of surficial terrestrial skeletal accumulation, the characteristic differences in ecological sampling of various depositional environments, and the quality of ecological data available in bone assemblages remain largely unexplored. Here, I use the well-studied animal community of Yellowstone National Park, WY (YNP) to test the fidelity with which modern surficial bone accumulations mirror the population structure and habitat utilization of a living community in a temperate environment. Forty transects were used to survey the death assemblage of four habitats (grasslands, forests, river margins, and lake margins). Species representative of all body-size classes and nearly all vertebrate groups found in YNP are documented in bone remains – of the known YNP vertebrate diversity, only bats and reptiles are missing from the death assemblage. For ungulates (the group with the largest sample size and highest statistical power), the death assemblage accurately reflects the living community; all native ungulates are present and comparisons between the living community's rank-order abundances and their representation in the dead remains are strongly positive (Spearman rho ~0.7-0.9). The sampled habitats show quantitative differences in live-dead agreement – habitats with higher fossilization potentials (lakes and rivers) show higher landscape-scale ecological fidelity than environments less likely to yield fossil deposits (forests). Calibrating bone-weathering rates finds time-averaging durations to be at least decadal. Finally, the fidelity with which modern surficial death assemblages reflect large-mammal communities not only speaks to the paleocological potential of fossil assemblages, but suggests that multigenerational death assemblages may be used to compliment traditional biological surveys and provide historical contexts to better interpret the stability of modern populations and track their changes over ecological time.