Are wetland regulations cost effective for species protection? A case study of amphibian metapopulations

Recent declines in amphibian populations have raised concern among conservation biologists, with habitat loss and degradation due to human activities among the leading causes. The most common policies used to protect the habitat of pond‐breeding amphibians are wetland regulations that safeguard the wetland itself. However, many amphibians spend much of their adult lives foraging and over‐wintering in upland habitats and exist as metapopulations with dispersal among ponds. With no consideration of lands in the dispersal matrix, wetland policies may be ineffective at protecting amphibians or other wetland species that disperse across the landscape. This paper examined the adequacy and cost effectiveness of alternative conservation policies and their corresponding land use patterns on the long‐term persistence of pond‐breeding amphibians in exurban landscapes. We used computer simulations to compare outcomes of wetland buffer policies and broader landscape‐wide conservation policies across a variety of landscape scenarios, and we conducted sensitivity analyses on the model's species parameters in order to generalize our results to other wetland species. Results showed that, in the majority of human‐dominated landscapes, some amount of dispersal matrix protection is necessary for long‐term species persistence. However, in landscapes with extremely low‐intensity land use (e.g., low‐density residential housing) and high pond density, wetland buffer policies may be all that is required. It is not always more cost effective to protect core habitat over the dispersal matrix, a common conservation practice. Conservation costs that result from forgone residential, commercial, or agricultural activities can vary substantially but increase in a nonlinear manner regardless of land use zoning. There appears to be a threshold around an average habitat patch occupancy level of 80%, after which opportunity costs rise dramatically.