Effects of fish predation on benthic communities in the San Juan Archipelago

Thesis (Ph.D.)--University of Washington, 2015-12Predation is a strong driver of community assembly, particularly in marine systems. Rockfish and other large fishes are the dominant predators in the rocky subtidal habitats of the San Juan Archipelago in NW Washington State. Here I examine the consumptive effects of these predatory fishes, beginning with a study of rockfish diet, and following with tests of the direct influence of predation on prey species and the indirect influence on other community members. In the first chapter I conducted a study of the diet of copper rockfish. Food web models benefit from recent and local data, and in this study I compared my findings with historic diet data from the Salish Sea and other localities along the US West Coast. Additionally, non-lethal methods of diet sampling are necessary to protect depleted rockfish populations, and I successfully used gastric lavage to sample these fish. Copper rockfish from this study fed primarily on shrimp and other demersal crustaceans, and teleosts made up a very small portion of their diet. Compared to previous studies, I found much higher consumption of shrimp and much lower consumption of teleosts, a difference that is likely due in part to geographic or temporal differences in prey availability. Given that copper rockfish diet was so dominated by shrimp, in the second chapter I used field experiments and surveys to determine the top-down effect of rockfish and other large demersal fishes on shrimp and other prey species. In three years of predator and prey surveys I found that shrimp abundance was negatively correlated with pooled predator biomass, but not abundance. Small fish and crab abundance were not correlated with predators. In two rounds of experimental exclusion of predatory fishes I found elevated abundance of both shrimp and small fishes in areas protected from predators. Despite this direct effect of predators on their prey, I did not find evidence of an indirect predator influence on the encrusting assemblage in the exclusion experiment. Trophic cascades are common in temperate marine ecosystems, often mediated by predators consuming urchins and urchins grazing on kelp. The San Juan Archipelago is notable for its lack of both urchin predators and strong grazing pressure from urchins. In the final chapter I looked for evidence of trophic cascades structuring the benthic community in this system. I surveyed the mobile invertebrates and sessile epibenthos at 12 sites within San Juan Channel, and compared these assemblages to predatory fishes to test for co-variance between the groups. Despite some limited evidence of co-variance between the predatory fishes and the other groups, the species involved did not suggest trophic relationships as the causal agent. Instead, predatory fishes may be responding to the biotic habitat provided by benthic organisms. Co-variation between the mobile invertebrates and sessile epibenthos provides supporting evidence of a three-species interaction between urchins, chitons, and social ascidians, and evidence of urchins reducing kelp cover. Finally, I compared these three assemblages to current flow and found strong evidence of current influencing assemblage composition. Although this benthic community does not appear to be controlled by a fish-initiated trophic cascade, consumptive interactions at lower trophic levels and dependence on abiotic factors play important structuring roles