Impacts of dispersal, ecological interactions, and fishing effort dynamics on efficacy of marine protected areas : How large should protected areas be ?

Management for sustainable fisheries requires effective tactics for limiting exploitation rates. Limitation based on annual stock assessments and total allowable catches calculated from these assessments can be very dangerous, and marine protected areas (MPAs) are one means by which to limit exploitation rate directly even when total stock size is highly uncertain. This application of MPAs would probably require much larger areas than are now envisioned for limited objectives related to protection of seed spawning stock and local biodiversity. It might in fact cause a basic shift in thinking-from regarding MPAs as exceptional areas to regarding fishing areas as the exception (as is now the practice in, e.g., salmon and herring fisheries). The present paper describes the use of ECOSPACE, a new modeling tool based on ecosystem simulations, for preliminary determination of how large MPAs need to be; ECOSPACE models suggest that dispersal, trophic responses (prey depletion, increased dispersal of predators in response to competition), and spatial fishing-effort responses (concentration of fishing near MPA boundaries) are all likely to reduce the effectiveness of small MPAs. The models suggest we should see not simple high-low density differences across MPA boundaries but rather spatial gradients from low density in exploited areas to high density near the centers of larger MPAs. Such spatial density gradients should be accompanied by spatially organized 'trophic cascade' patterns if trophic interactions are important determinants of abundance. MPA design can work with or against spatial variation in fishing effort caused by economic cost and risk factors; ECOSPACE can help to demonstrate ecological consequences of alternative design strategies, but the most important uncertainties are about socioeconomic responses (cooperation or competition) rather than ecological ones. Design of experimental policies and monitoring programs for evaluation of MPAs should proceed from careful modeling to define likely spatial, temporal, and trophic scales for both ecological and fishing responses.