Selecting marine reserve locations: Optimality versus opportunism

Location of most marine reserves has depended more on social criteria and opportunism than on scientific study. Nevertheless, numerous studies from many habitats and places around the world indicate that fully protected reserves (areas closed to all fishing) have shown clear benefits, at least within their boundaries. This evidence suggests that reserves will work in most areas they are placed. Many people are uncomfortable with this haphazard approach and worry that we should be looking to science to help optimize placement. Here I examine some of the key factors affecting reserve performance and ways they might influence our approach to locating reserves. Habitat quality, intensity of exploitation around reserves, area and proximity of other reserves, protected species' life history and dispersal characteristics, and boundary porosity all affect how reserves perform. Any reserve we create will involve trade-offs among different objectives. For relatively sedentary species, precise reserve placement appears relatively unimportant to performance, but for migratory species, much more precise placement will be necessary to encompass migration bottleneck and nursery areas. Two nonbiological factors are of overriding importance to performance. Fully protected reserves will achieve much more than those that allow limited take, and well-enforced reserves will be much more effective than poorly enforced ones. Rather than seeking to optimize placement of individual reserves, we should construct networks of interacting reserves as a bet-hedging strategy against variability and uncertainty in the marine ecological processes, and resource-management policies, that affect reserve performance. Opportunism, informed by science, can achieve a great deal. The risk is much greater that we will fail to achieve our management objectives if we delay in order to embark upon lengthy studies than if we begin establishing reserves today on the basis of what we already know.