Deterministic and stochastic forces in community ecology: integrating competing paradigms in theory and observation.

Abstract

An important goal in community ecology is to understand the interactions between multiple mechanisms - species differences and niche differentiation, stochasticity, environmental heterogeneity, and spatial processes - and their consequences to community structure. In this dissertation, I address three distinct issues within this general program. First, using a spatially explicit model, I compare assembly, structure, and invasibility of communities with varying levels of neutrality versus niche-differentiation. Communities' responses to invasions are determined by the extent of functional variation in the local species pool, predicting varying responses to inter-biome exchange and evolutionary diversification along the niche-neutral gradient. Second, I demonstrate statistical bias in the standard test for monoculture overyielding and develop a bootstrap correction algorithm. Correcting this bias is important to evaluating the relative importance of selection and complementarity effects to community processes. Finally, I analyze the extent to which species differences, dispersal, and stochasticity influence metacommunity dynamics in a long-term nitrogen addition experiment. I find that all three mechanisms are active in the study system, necessitating further development of metacommunity models.