Spillover effects from temperate marine protected areas

Economic benefits of permanently closed areas can accrue to fisheries in two ways. Export of reproductive products can increase recruitment in open areas, while movement of harvestable-sized animals provides benefits in the form of "spillover" from the refuge, resulting in elevated catch rates near closed area boundaries. Here, we evaluate potential spillover effects from four large marine protected areas in temperate New England waters, closed beginning in 1994. True spillover, as differentiated from seasonal, ontogenetic, or environmentally driven movements, requires differential densities within and adjacent to the closed areas. Density-related spillover, thus, is typified by a biomass or abundance gradient beginning at the boundary and declining as a function of increasing distance. Moderate rates of dispersion are required to establish the density gradient, which may be enhanced by differential distribution of fishing intensity. We tested for density gradients as a function of distance from the closed areas using otter trawl tow-by-tow data collected by scientific observers aboard commercial fishing vessels. Data were adjusted for tow duration (e.g., kg/h towed). A total of 51 species-area combinations were evaluated for the presence of density gradients consistent with implied spillover effects. Of this total, five species-area combinations exhibited statistically significant declines in catch rates with distance. These significant combinations were generally consistent with research vessel surveying information showing year-round catches higher inside the closed area as compared to adjacent open areas after they were closed (e.g., "reserve effects"). Combined groundfish species catches did not show significant declining trends as functions of distances from the four closed areas, nor did the numbers of species caught per haul. Haddock Melanogrammus aeglefinus associated with Closed Area I demonstrated the most pronounced apparent spillover, but the pattern of relative fish density at the reserve boundary exhibited a seasonal cycle associated with spawning. We conclude that spillover effects are not a universal consequence of siting marine protected areas in temperate waters but are related to the specifics of the degree of random and directional movements, the fishing intensity field in the adjoining open areas, seasonal migration patterns, and optimal habitat preferences of individual species in relation to the placement of reserve boundaries, all of which may confound the interpretation of spillover.