ABSTRACT
Ocean fishing seasons for Pacific salmon Oncorhynchus spp. are set to limit fishing impacts on certain naturally reproducing stocks. Fishing-related mortalities are estimated using harvest models based on historic salmon catch distributions in fisheries. Current models assume that, in a fishery, all stocks are caught at the same rate. Each stock is assumed to be a single, uniformly mixed population. New fisheries have been proposed in which most hatchery fish would receive an external mark and only marked fish would be retained. These “selective” fisheries would allow fishing access to hatchery runs while reducing mortalities on naturally reproducing stocks. In selective fisheries, stocks would not be equally vulnerable to the fishing gear, and would suffer several forms of gear-related mortality. A new family of harvest models is needed to permit managers to assess differential impacts of selective fisheries on marked and unmarked salmon. We describe such a model and use it to evaluate effects of a proposed selective fishery using 126 stocks of coho salmon O. kisutch in 183 fisheries from California to Alaska. Our results suggest that selective fisheries will have lower total exploitation rates on unmarked stocks compared with marked stocks. Modeled stocks that were caught in a mixture of traditional mixed-stock and selective fisheries showed less benefit than those caught primarily in selective fisheries. In the fisheries we modeled, catches were lower because many wild fish that would have been landed in a traditional fishery were released. Mass marking and selective fishing may provide fishing opportunity and access to hatchery fish while reducing mortalities of wild fish. However, non-catch mortalities of unmarked fish in selective fisheries may be substantial, and must be accurately accounted for if conservation goals are to be achieved.
