Subsidized predators and imperiled prey: The impact of avian predators on salmonids

Background/Question/Methods

Predators can exert strong top-down control on their prey with impacts that can be magnified when predators are subsidized by flow of nutrients and energy across ecosystem boundaries. Subsidized predation is increasingly realized to be an important issue in conservation; for example when subsidies increase predation on already imperiled prey. Recent work in coastal California watersheds has produced observations of seabirds (primarily Western gulls Larus occidentalis) congregating in estuaries and feeding on outmigrating juvenile salmonids. Western gulls are heavily subsidized by anthropogenic food resources and have experienced significant population increases as a result. Extensive salmonid population monitoring in five central California watersheds began in 2002 and since then over 20,000 juvenile salmonids have been injected with Passive Integrated Transponder (PIT) tags. For this study, we investigated the degree of predation by seabirds on endangered steelhead (Oncorhynchus mykiss) and coho salmon (Oncorhynchus kisutch) that persist in small populations in central California coast streams. In an effort to quantify seabird predation, we scanned roosting and nesting areas at nearby central California seabird breeding colonies for salmonid PIT tags that were regurgitated by seabirds. This methodology allows quantification of minimum predation rates and insight into individual level factors that influence predation through comparisons of characteristics of tagged fish that were eaten relative to the population as a whole.

Results/Conclusions

The recovery of PIT tagged salmonids on seabird breeding colonies accounted for 0.2-2.5% of the tagged population of salmonids in each watershed from 2002 through 2008. Coastal distance from stream river mouth to the seabird colony was inversely related to predation rates--a higher proportion of tags were recovered from streams closer to the seabird breeding colony than more distant streams. In Scott Creek, where the most extensive tagging has occurred, there was similar predation mortality between wild steelhead and wild coho salmon (0.7% and 0.6% of the PIT tagged population, respectively), while mortality between hatchery steelhead and hatchery coho salmon was more variable (1.0% and 0.4% of the PIT tagged population, respectively). These data reveal a previously overlooked source of mortality for endangered populations of steelhead and coho salmon, representing minimum estimates that may be substantially higher. It is possible that subsidized predators are contributing to salmonid declines.