As more emphasis is put on biodiversity conservation, how best to select a system of protected areas for wildlife conservation is an issue of great importance. There is a rich economics literature on the reserve site selection problem. However, most economic studies assume the independence of risks that affect wildlife species, leaving the issue of spatially correlated risk largely unexplored.

This study contributes to the literature in twoaspects. First, this study incorporates spatially correlated risk, into a reserve site selection model. And second, this study incorporates heterogeneous spatial risk, in the context of land development risk in Virginia, both with and without a budget constraint.

To evaluate the significance of spatially correlated risk in conservation design, I apply the reserve site selection model to a Virginia landscape. In a basic setting, a hazard is introduced which is allowed to spread to adjacent land parcels, where I investigate the impact of spatially correlated risk at three spatial scales: one-county, four-county, and state-level. Optimal reserve designs are characterized by similar spatial patterns indicating that spatially correlated risk plays an important role in the selection of parcels for reserve. Specifically, as spatially correlated risk increases, I find that, in general, reserve connectivity decreases. I also examine a setting with heterogeneous risk and observe similar patterns in the optimal reserve design. I find that the reserve becomes more dispersed in higher risk areas primarily. Finally, I explore the tradeoffs between species protection and budget constraints in the presence of heterogeneous spatial risk. All comparative statics indicate that spatial correlated risk plays an important role in conservation reserve design.