Finding Efficient Harvest Schedules under Three Conflicting Objectives

Abstract

Abstract Public forests have many conflicting uses. Designing forest management schemes that provide the public with an optimal bundle of benefits is therefore a major challenge. Although a capability to quantify and visualize the tradeoffs between the competing objectives can be very useful for decisionmakers, developing this capability presents unique difficulties if three or more conflicting objectives are present and the solution alternatives are discrete. This study extends four multiobjective programming methods to generate spatially explicit forest management alternatives that are efficient (nondominated) with respect to three or more competing objectives. The algorithms were applied to a hypothetical forest planning problem with three timber- and wildlife-related objectives. Whereas the ε-Constraining and the proposed Alpha-Delta methods found a larger number of efficient alternatives, the Modified Weighted Objective Function and the Tchebycheff methods provided better overall estimation of the timber and nontimber tradeoffs associated with the test problem. In addition, the former two methods allowed a greater degree of user control and are easier to generalize to n-objective problems.