Mixed integer linear programming approach for spatially optimizing wildlife and timber in managed forest ecosystems

Abstract

This paper presents mixed integer linear programming formulations for land allocation that optimize spatial layout for a single time period and that have the property that the number of integer variables is a linear function of the level of spatial resolution. The formulations focus on timber, edge-dependent wildlife, and area-dependent wildlife. They account for the amount of edge, the fragmentation of habitat area, and a habitat area threshold for minimum viable population size. Habitat area connectivity is modeled as a probabilistic condition. A case example demonstrates the approaches and includes landscapes characteristics such as terrain, hiking trails, campgrounds, and rivers, which result in different productivities over the landscape and obstacles to wildlife habitat area connectivity.