Ecosystem modeling of marine populations

Fisheries scientists have traditionally relied on single species models designed to estimate mortality from directed fishing, and to some degree from bycatch. Ecosystem modeling allows us to address the drivers of mortality that single species models neglect. In this thesis, I use several different types of ecosystem models to address issues relating to the conservation and management of leatherback turtles (Dermochelys coriacea, Chapter II), yellowfin tuna (Thunnus albacares, Chapters III and IV), and marlins and sharks (e.g. Makaira nigricans and Prionace glauca, Chapter V). To estimate the impact of coastal poaching, bycatch, and habitat disturbance on leatherback turtles I conducted a Bayesian risk assessment. This analysis revealed the importance of reducing coastal sources of mortality, as well as longline bycatch, if the populations are to avoid extinction. To estimate trends in yellowfin tuna recruitment from catch-effort data, I used a state space model that allowed for the estimation of both process error and observation error. I found a distinct increase in recruitment in the mid 1980's, most likely indicating a shift in climate or oceanographic conditions. To test the ability of multi-species models to provide tactical (annual) management advice for yellowfin tuna, I linked a food web simulation 'operating model' to an assessment model that could incorporate information about the abundance of up to five prey and five predator species. The results suggest that management performance is optimized by assessments that consist of a simple multispecies inverse model, which estimates the relationship between yellowfin mortality and the abundance of one predator and one prey. To evaluate the impacts of new longline gear (circle hooks) on marlin and shark populations, I used a food web simulation model of 25 species and groups in the Central North Pacific. The results suggest that because circle hooks have higher catch rates than the currently-used tuna-style hooks, switching to circle hooks depletes marlin biomass by 30--50% and shark biomass by 15--40% over 30 years. One indirect effect of these changes, which is illustrated by the food web model, is an increase in yellowfin tuna biomass, due to a release from predation by marlins.