Graduation date: 2008Predators are fundamentally important for regulating and driving prey population dynamics as well as structuring ecological communities. Over-exploitation of marine resources has caused dramatic depletions of predatory fish species and shifts in marine community composition, with attendant declines in productivity and diversity of marine ecosystems. Despite the important ramifications of these patterns for humankind, the specific ecological mechanisms and potential indirect effects underlying these trends have been largely unexplored. I combined small-scale experiments and broad-scale observations to investigate how fishing can affect interactions among predators and subsequently be transferred to interactions between predators and their prey. Previous observations had indicated that small, unfished predators increase in abundance on reefs where large, fished predators are removed. To test the effects of such shifts in relative abundance of different predatory fishes on community structure of lower trophic-level species, I manipulated the presence of fished and unfished predatory species on coral patch reefs in the Bahamas. This controlled field experiment demonstrated that different predatory fishes, which are often assumed to have similar effects on prey species, in fact had remarkably different effects on prey diversity and abundance. In a second field experiment, I found that increased abundance of a large, fished predatory species on coral reefs decreased the activity and growth of smaller, unfished predators. Although small, unfished predators had strong, negative effects on lower trophic-level prey on reefs with low abundances of larger, fished predators, these effects were reduced with increasing abundance of the fished species. Lastly, I used observational data from volunteer SCUBA-diver surveys to investigate how the structure of predatory fish communities on coral reefs across the greater Caribbean region has been affected by fishing. I used density of human populations as a proxy for fishing intensity. I found that, as human density increases, overall predator abundance and diversity decreases, and reef communities became dominated by only a few, small species. This research indicates that the effects of fishing on larger predatory species can permeate throughout ecological communities, well beyond simple reductions in the abundance of fished species. In light of these community-wide effects, ecosystem-based approaches are necessary for successful management of fisheries and conservation of coral reefs.