ESTIMATING CHANGING RELATIVE ABUNDANCE IN FOSSIL COMMUNITIES

Temporal changes in the number of individuals are consequences of reproduction, recruitment and dispersal. Such changes in population size also have consequences for resource use in the said ecological community as well as the survival of the given population. However, it is not trivial to estimate changing population sizes among living communities, and even more difficult among organisms that have long died. Here, we present an expansion of occupancy models that have been used in statistical ecology to tease apart presence, true absence and detection probabilities. We use the occupancy modeling framework to design models that will allow us to estimate relative species abundances, using repeated sampling of sites within geological formations. We design simulations to demonstrate the strength of our approach, clearly showing that simply using detection ratios (the number of individuals of a specific species observed per number of total fossilized individuals found) as an estimate of relative abundance is problematic. We then apply our approach to an empirical system of cheilostome bryozoan species in the Pleistocene to the Recent of New Zealand. We end by discussing ways of improving on our current approach, both in terms of model modifications and the data we could accumulate for our own and other relevant empirical systems.