BULLETIN OF MARINE SCIENCE, 74(3): 727–747, 2004 727 Bulletin of Marine Science © 2004 Rosenstiel School of Marine and Atmospheric Science of the University of Miami TOWARD ECOSYSTEM-BASED EXTRACTION POLICIES FOR PRINCE WILLIAM SOUND, ALASKA: INTEGRATING CONFLICTING OBJECTIVES AND REBUILDING PINNIPEDS Thomas A. Okey and Bruce A. Wright ABSTRACT Trade-offs between the benefits and the costs of marine resource extraction become increasingly conspicuous as ecological limits are approached. Unfortunately, the his- torical lack of trade-off accounting has led to unexpectedly adverse effects. An integra- tive policy-search procedure in the modeling software Ecopath with Ecosim was used to shape fisheries extraction policies for Prince William Sound, Alaska, that explicitly account for trade-offs among economic, employment, and ecological objectives, in ad- dition to the thermodynamic constraints of the systemʼs food web. When economic and employment objectives were emphasized, the modeling routine reduced predators (Pacific halibut and pinnipeds) to maximize the production of prey groups whose mar- ket values and potential system biomasses promised maximum fisheries values. When ecological objectives were emphasized, the routine increased predators (orcas, halibut, porpoise, pinnipeds, lingcod, and seabirds), along with their salmon and herring prey, while decreasing gadoids (sablefish, Pacific cod, and pollock). Competition between fisheries and predators was indicated by increases in predators with decreases in fishing. A 20% increase in the biomass of pinnipeds was achieved with comparatively modest, but “smart,” fisheries reductions. Commercial and overall fisheries catches consistently decreased by the end of all 20-yr simulations, indicating that 1994–1996 commercial fishing levels were unsustainable. Oceanographic factors strongly influence the structure of marine ecosystems, and their influence is especially conspicuous at higher latitudes such as the subarctic region sur- rounding Prince William Sound (PWS), Alaska (Hollowed and Wooster, 1992; Polovina et al., 1995; Brodeur et al., 1996; Piatt and Anderson, 1996; Tyler and Kruse, 1996; Francis et al., 1998; Rosenkranz, 1999). The amount of energy flowing through a given system is, nevertheless, limited in the sense that one use of biota can preclude or limit other uses. Fisheries, for example, directly and indirectly affect nontarget biota and hu- man interests such as other fisheries, nonextractive activities, and other values. Various trade-offs thus exist at the interfaces of a fisheryʼs (or an ecosystemʼs) economic, social, ecological, and legal aspects. Failure to consider such trade-offs adequately can have un- expected, and often catastrophic, consequences (see, e.g., Myers et al., 1996; Walters and Maguire, 1996; Harris, 1998; Myers and Worm, 2003; Cox and Kitchell, this issue). Rapidly evolving computing tools and newly emerging approaches promise some help with the serious challenge of rigorously accounting for trade-offs in complex and dy- namic ecological systems (Walters et al., 1997, 2000, 2002; Christensen and Walters, this issue). Fishing policies can theoretically be designed to help shape marine commu- nities to promote particular objectives or combinations of objectives, even in the context of physical factors that also shape the system. Balancing disparate or contested human objectives for an ecosystem is a central, but profoundly difficult, task for managers and policy makers. The development of such integrated analytical tools will aid in this dif- ficult task. We used a new dynamic simulation tool (Walters et al., 2002) in the modeling soft- ware Ecopath with Ecosim to explore the potential effects of fishing-policy scenarios in