Bootstrap Simulation and Response Surface Optimization of Management Regimes for Douglas-Fir/Western Hemlock Stands

Abstract

A method was proposed to simulate forest stand growth, timber prices, and interest rates by distribution-free bootstrapping, and then optimize management controls for economic and ecological objectives by response surface analysis. The method was applied to Douglas-fir/western hemlock stands to predict the effects on economic and ecological objectives of management alternatives defined by the cutting cycle, C, the residual stand basal area, B, the diameter of the largest tree, D, and the ratio, q, of the number of trees in adjacent diameter classes. The effects were described with response surfaces, which were used to determine the best combinations of B, q, and C for each management criterion. Adjusting B, q, and C could control for 97 to 99% of the variability in the expected value of species diversity, size diversity, percentage of peeler logs, and basal area, and for 80 to 90% of the variability in land expectation value and annual production. Economic and ecologic criteria were generally most sensitive to the q ratio, the residual basal area, and the cutting cycle. Annual production was negatively correlated with tree size diversity and wood quality. There was no apparent conflict between stand diversity and wood quality.