THE IMPACT OF VARIABLE FOSSILIZATION ON RECONSTRUCTIONS OF PALEOCOMMUNITIES AND COMMUNITY DYNAMICS: THE SAN FRANCISCO BAY COMMUNITY

Variable fossilization potential hinders accurate paleocommunity reconstruction, affecting not only measures of diversity, but also the interpretation of community dynamics and biotic interactions. In order to assess the impact of biased fossilization, San Francisco Bay, an estuarine ecosystem, is analyzed both as a modern community and paleocommunity. In both cases, trophic networks are reconstructed probabilistically to reflect uncertainty in community composition, biotic interactions and fossilization. The modern network comprises 1241 invertebrate and 125 vertebrate species. To construct the networks, species within a community are parameterized ecologically and partitioned into guilds, where members of a guild share similar ecological characteristics. To construct paleocommunities, each species is assigned a preservation probability, drawn from a preservation distribution based on the ecological characteristics of that species. Knowing the preservation potential of each species allows us to stochastically assemble potential fossil communities of the Bay.

Given that network structure varies with species composition, we use simulations to assess the sensitivity of the networks to disturbance. Fossil communities will be subjected to the same perturbations as the modern community and the results compared. Because the extent to which extinctions propagate through networks is a function of the magnitude of the initial disturbance, as well as taxonomic and ecological diversity, we predict that the modern and fossil networks will respond differently to the same perturbation.