Optimal restoration of wildlife habitat in landscapes fragmented by resource extraction: a network flow modeling approach
Non-renewable resource extraction contributes greatly to degradation of wildlife habitats in boreal landscapes. In western Canada, oil and gas exploration and extraction have left a dense network of linear disturbances (seismic lines) and abandoned well pads that have fragmented boreal forest. Among multiple ecological effects, these disturbances have increased predator access to the preferred habitat of some wildlife taxa, most notably boreal woodland caribou, resulting in population declines. Restoration of seismic lines and abandoned well pads is a critical activity to improve the recovery of woodland caribou populations. We present a linear programming model that optimally allocates restoration efforts to maximize the access of caribou to nearby undisturbed habitat in a fragmented landscape. We applied the model to examine restoration scenarios in the Cold Lake First Nations area in northeastern Alberta, Canada, which includes caribou habitat but also areas of active oil and gas extraction. The model depicts the landscape as a network of interconnected habitat patches and combines three network flow sub-problems. The first sub-problem enforces the spatial connectivity of the remaining network of unrestored sites. The second sub-problem maximizes access to suitable habitat from the restored locations and the third sub-problem ensures the allocation of restoration activities in as few spatially contiguous restoration projects as possible. The approach is generalizable and applicable to assist restoration planning in other resource extraction regions and for other taxa.