stakeholders in tests of the value of reserves relative to conventional management approaches.
Although there is a well-developed theory of terrestrial reserves (Higgs and Usher, 1980; Gilpin and Soule, 1986; Pressey et al., 1993), a corresponding theory for marine systems has yet to be fully developed (Simberloff, 2000). Marine and terrestrial reserve theories need to be substantially different because the taxa involved have such different life-history traits and because marine systems are usually expected to yield commercial quantities of food from fishing. Terrestrial models often reference island biogeography theory (Simberloff, 1988) and tend to focus on preserving species or habitat richness in reserves, with little or no emphasis on repopulating adjacent areas to support hunting. In contrast, marine reserve models typically focus on single species, with an emphasis on population dynamics under conditions of human exploitation.
Gerber et al. (in review) categorize existing marine reserve models primarily by (1) whether the model is for single or multiple species, (2) what the key life-cycle elements and larval redistribution mode are, (3) what the density-dependent recruitment mode is, and (4) whether adult migration, stochasticity, and rotating spatial harvest are included in the model. They summarize several key results that appear to be general to virtually all types of models. In particular, at a constant level of effort that would otherwise result in overfishing, models indicate that reserves would increase the yield of the fishery relative to conventional management. A primary value of reserves in these circumstances is the higher reproductive capacity of adults protected from fishing. Models suggest that larger reserves would be required for species with high rates of juvenile and adult movement.
A second condition for high efficiency of reserves that has emerged from various models is that the target species should have moderate rates of juvenile and adult movement (DeMartini, 1993). These modeling results appear to apply generally to fishery reserves, but they are based on a very limited set of environmental assumptions. For example, most models assume that all larvae come from a larval pool distributed equally to potential juvenile habitats. In addition, most models do not allow for a number of reserves, and none allow reserves to be of different sizes or embedded in realistic current regimes. Thus, these conclusions from the first simple models are not realistic enough to be used predictively in specific reserve situations.
Currently, the most common form of reserve modeling is to assume that a habitat is divided into a reserve portion where fishing is limited and a portion in which fishing continues. Eggs produced throughout the habitat develop into a