The effects of fishing on coral reef fishes include both direct and indirect effects. Some of the former are obvious, such as reductions in overall abundance and body size by increasing overall population mortality. Additional obvious but indirect effects include impacts on reef habitat resulting from destructive fishing practices such as blast fishing. Other indirect effects are those most important to the maintenance of healthy reefs – effects on ecosystem functioning that result from disruption of the natural size and assemblage structure of reef fishes. Some evidence exists for reefal trophic cascades but the examples are most compelling when strongly interacting species are present in the system. Evidence for tri- and multi-trophic cascades is sparse and more data are needed to elucidate interactions among trophic levels, especially those involving apex predators. Most artisanal reef fisheries are unselective and multispecies, span the trophic spectrum, and include keystone predators and competitors and functionally dominant habitat engineers. Large-bodied fishes are disproportionately reduced by targeting and general overharvesting. The largest-bodied species and individuals within species, though, contribute disproportionately to population and community dynamics, and some like parrotfishes and other large herbivores further contribute disproportionately as habitat engineers to ecosystem functioning by bio-eroding reefs and controlling coral–algal dynamics. Progress in reducing fishing effects on ecosystem functioning will require careful balance between social safeguards and resource management that effectively applies new developments in conventional fishery science that either complement (or are complemented by) effective no-take MPAs.

Large-scale fishing of the world's oceans began only a few hundred years ago [1226, 1229, 2124, 2446]; as a relatively recent phenomenon, it has primarily been acting on ecological, not evolutionary, time scales. The genetic responses to extraction by fishes and other living resources [1349] of course are not independent of the phenotypic responses operating in ecological time [2522], and rates of these lagged genetic responses are dependent on the intensity of selection [2435].