Abstract
Documenting the role of past interactions in the assembly of present communities has proven problematic. Colonization is a key process in community assembly that is both potentially driven by past interactions and amenable to experimental approaches. Colonization and oviposition by an aquatic beetle (Tropisternus lateralis) was assayed in the presence and absence of both 'harmless' and tactilely/visually isolated predatory fish (Lepomis gibbosus and L. macrochirus). Beetles avoided each treatment with fish when compared to fish-free experimental pools. Activity levels after colonization also differed significantly between adults in fish and fish-free tanks. Predator effects on species composition are typically ascribed to contemporary predation events; the presence of a strong avoidance response demonstrates that past species interactions affect present distributions and may play an important role in the ongoing assembly of contemporary communities. Documentation of such avoidance behavior in a growing number of species fundamentally alters our view of the processes affecting species distributions and the process of community assembly.