MODELING THE INFLUENCE OF SPECIES-DEPENDENT DIFFERENCES IN PRESERVABILITY ON RELATIVE ABUNDANCES IN DEATH ASSEMBLAGES

Paleoecologists use lists and counts of fossil species to infer the structure of ancient ecological communities and how they have changed through time. As has long been recognized, however, fossil assemblages are not strictly equivalent to living communities due to factors like time averaging of community variation, postmortem spatial mixing, and inherent differences in the population biology and preservability of different species. This study takes a modeling approach to explore the importance of differences in the inherent degree of preservability among species on their abundances in death assemblages. Local living communities are simulated by sampling individuals from a fixed master species pool with a log-series distribution of species abundances (determined by Fisher's α = Hubbell's θ). Local communities are of fixed size (J) and every individual member has an equal probability of dying in each time step (v). Dead individuals can be replaced either by recruitment from the master pool (probability m) or through reproduction of surviving living individuals in the local community (1-m). Rate and magnitude of change in the species composition of the local community depends on v, m, J, and θ. Dead individuals from each time step are added to an accumulating death assemblage; within the death assemblage, each individual has a probability (ρ) of being destroyed in each time step that depends on its species identity. Initial results indicate that the degree of alteration of dead species abundances relative to both the local living community and the master pool depends on the coefficient of variation of species' inherent preservabilities relative to the volatility of the living community and degree of time averaging in the death assemblage. The influence of differences in species preservability is most difficult to detect when temporal mixing is not of sufficient duration to average out living community volatility. Model results indicate that understanding how differences in preservation influence species abundances in death assemblages will require field-based, quantitative measurements of rates of postmortem loss (i.e., half-lives) from a variety of species and depositional settings.