Ecologists have long sought to understand the relative importance of the wide variety of physical and biological factors that shape communities. Kelp forests on the Pacific coast of North America have served as a model system for examining the community level consequences of top-down trophic cascades and bottom-up effects of nutrient availability, recruitment, and physical stress. Here we use structural equation modeling (SEM) to explore a data set describing the distributions and abundances of multiple taxa and the corresponding oceanographic conditions across 16 sites around the Channel Islands National Park in southern California over more than a decade. We examine the effects of adding increasing biological detail into our models, and consider how SEM can be applied to long-term observational field data. We show that top down forces are important to the structure of kelp forest food webs, but often only when examined in a community context that incorporates variation in bottom-up and environmental factors. We also elucidate the important roles of recruitment, wave disturbance, and nutrient limitation on community structure.