Land management with ecological and economic objectives : developing a production possibility set of wildlife species persistence and timber harvest value using simulated annealing

Abstract

An integrated model combining a wildlife population simulation model, and timber harvest and growth models was developed to explore the tradeoffs between the likelihood of persistence of a wildlife species and timber harvest values on a landscape in the Central Oregon Cascades. Simulated annealing, a heuristic optimization technique, was used to identify harvest schedules that maximized the net present value of timber harvest relative to a given wildlife habitat constraint over a 100 year planning period. This habitat constraint was identified by a logit regression model that estimated the likelihood of species persistence from a series of weighted habitat scores developed from extensive simulation work within the wildlife population model. By solving this problem for a range of different species persistence constraints a production possibility frontier is developed that shows the relative tradeoffs between timber harvest value and likelihood of species persistence on this landscape. Although the results are specific to the wildlife species and the landscapes analyzed, the approach is general and may provide a structure for future models that will allow land managers and forest planners to become more informed about the tradeoffs among competing resources.