marine ecosystems, and thus this level of reserve commitment does not meet current goals for marine reserves as a management tool. In terrestrial systems, slightly more than 5 percent of the earth's land area enjoys some sort of protected status (Groombridge 1992). Clearly, even that amount of protection, while helpful, is not enough to prevent continuing loss of biodiversity. Marine systems are much more open, with more geographic exchange, than most terrestrial systems and thus need an even greater area of protection than terrestrial systems when species with dispersive life stages are involved.
A second line of evidence independent of the first is that goals of fishery management often focus on the protection of a certain fraction of the spawning stock (Clark 1996). If only 2 percent of the standing stock of a species is allowed to spawn, each individual must produce 50 offspring for the population to maintain itself. Such high reproduction per individual is very sensitive to environmental conditions and can lead to the collapse of the standing stock. As a result, fishery managers often try to protect at least 20 percent of the population. Spawners must then produce at least five offspring each, but this value is likely to be more easily achieved on a long-term basis than the 50-offspring requirement discussed above. An example of this approach was given by Bohnsack (1994), who calculated that protecting 20 percent of the habitat of red snappers (Lutjanus campechanus) would increase the total productivity of the fish population. That increased productivity would more than compensate the fishers excluded from the closed areas. These arguments are supported by information that fishing drastically reduced the numbers of spawners in several species (Mace and Sissenwine 1993, Goodyear and Phares 1990).
A third line of reasoning is an application of the inverse-square law. One of the goals of protected areas is for them to export eggs, larvae, or juveniles to other areas. The sea is a powerful dilution agent, and eggs and larvae of even coastal species will decrease in density as they spread out from a center of production. Settlement away from a protected area will decline rapidly with distance, and unless the protected area is very large, or occurs as a string of protected areas along a coast, the export of larvae and juveniles will be limited. Although research on this topic is critically needed, the large dilution capacity of the oceans suggests that a substantial fraction of habitats need to be larval exporters for reproductive individuals within reserves to have an effect outside reserves.
The establishment of MPAs has the potential to affect many fishers, especially to the degree that they lack mobility and the MPA excludes them from traditional fishing areas. It also might appear to impose a heavy cost on industry with no offsetting benefits. Nonetheless, there are economic considerations that can actually favor MPAs. This report has emphasized the unavoidable uncertainties involved in resource management, great enough to lead to the collapse of