Bronwyn Rayfield, Allan Brand, and Marie-Josee Fortin. University of Toronto
Long-term population viability in protected areas depends on both the supply of immigrants through dispersal and on the outcome of interspecific interactions such as predation. This latter consideration has been largely ignored in terrestrial protected area designs despite an increasing recognition that species’ distributions are dynamic in response to local ecological processes. Here, we compare the consequences of alternative protected area designs on predator-prey spatial dynamics in a simulated boreal forest landscape as it changes through time due to succession, fire, and logging. We contrasted protected areas selected for species-specific habitat quality alone with those selected for a combination of habitat quality, habitat contiguity, and habitat diversity. Using species’ spatial habitat requirements, we determined the density and spatial locations of American marten (Martes americana) and its two dominant prey species, the red-backed vole (Clethrionomys gapperi) and the deer mouse (Peromyscus maniculatus), over 500 years. We tracked changes in total habitat area for each species and spatial overlap of predator-prey habitats. Marten habitat declined across the landscape in all scenarios due to logging and fire, while all other species showed either no change or an increase in available habitat. Changes in the spatial overlap of marten and its prey differed between protected area scenarios, with the largest decreases occurring in protected areas selected for a single species without constraints on the spatial contiguity of habitats. Our results emphasize the potential short-comings of a single-species approach to protected area selection in dynamic forested landscapes which overlooks important interspecific interactions that may change due to dynamic habitat.