9709559 Naeem The composition of biological communities varies enormously over most scales of investigation. Changes in community composition are especially prevalent in small ecosystems where local extinction, colonization, and biological invasion are commonplace, but large communities are arguably experiencing such changes as well. How sensitive are community properties to such variation in composition and can patterns be predictable? Although communities exhibit a tremendous range of responses to even small-scale changes in community composition, a diverse body of theory suggests that this broad range of responses is predictable. This body of theory is based on the compensatory abilities of species within trophic guilds and the distribution of interaction strengths and the number of trophic groups in a community. The large number of replicate communities necessary for testing this theory, however, has limited empirical explorations. This research will use small analogs of freshwater microbial communities reared under controlled environmental conditions with sufficient replication to test this body of theory. Composition is experimentally varied using both assembly and fixed-structure methods to produce communities that contain 2, 3, 5, or 5 trophic levels with 1 to 5 species per level. Autotrophic and heterotrophic biomass will be measured as community response variables. Three sets of experiments are designed to explicitly assess the roles of compensatory growth and interaction strengths in determining patterns in biomass distribution. This study will provide an in-depth exploration of the interface between a specific set of theories and its biological analogs. It will link experimental results with equivalency- and uniqueness-based theory to predict patterns in the association of biomass distribution with community composition.
