A traits-based model of species diversity

Models which use species-specific trade-offs of life-history traits have been widely used to simulate stable, diverse communities of sessile organisms. However, the precise estimation of trade-off parameters required by many models cannot be easily realized for multiple species living within rapidly changing environments. We developed a spatially-explicit model, MCAP, that uses empirically estimated parameters to define species-specific life-history functions. This approach did not require trade-offs to be explicitly stated, but rather intrinsically represented trade-offs within the data structure defining each species life-history functions (e.g., age or size dependent changes in reproduction, dispersal, establishment, mortality and growth). The use of species-specific functions with empirically estimated parameters allowed each species to be uniquely characterized with subsequent estimation of model sensitivities due to changing functional forms or parameter errors. Empirical data was obtained from field observations and experiments for 16 species within tidal freshwater marsh communities of the Chesapeake Bay. Model results compared with these data showed that MCAP realistically represented processes structuring these tidal marsh communities. Model experiments varying weather and the range of habitat conditions accessible to each species showed that patterns of abundance and diversity were most affected by local habitat conditions and neighborhood interactions among competing species. The simulation approach used in MCAP was effective and efficient, providing a comprehensive tool for assessing the simultaneous effects of changing environmental drivers and species composition on the persistence and biodiversity of marsh communities.