Abstract:
A stand level model, PREscription generator under Multiple Objectives (PREMO) was built to generate prescriptions that address multiple objectives for management of the forests of the Applegate River Watershed. PREMO is a part of the landscape model of the Applegate River Watershed Forest Simulation Project and generates prescription choices for the landscape simulation. Possible goals at the stand level, in addition to present net value, include limiting fire hazard, limiting insect hazard, enhancing structural complexity, maintaining snags and down wood for wildlife habitat, and enhancing fish habitat. PREMO finds good but not necessarily optimal solutions using the RLS-PATH algorithm (Yoshimoto, 1990) with a multi-stage look-ahead. The goal programming objective function maximizes the present net worth minus the squared sum of deviations from the forest structure goals. Placing scalar multipliers on desired goals and choosing target levels for goal measurements creates a goal emphasis. At each stage, PREMO chooses the prescription with the highest PNV from among the prescriptions that maximize the objective function. Variables in PREMO are live trees, snags, and down woody debris by species and diameter class. Growth is simulated with relationships from the Forest Vegetation Simulator (Dixon et al., 1995). Several prescriptions were generated for a young pine stand using different goal emphases. Using PREMO to find prescriptions that come close to meeting all goals seems generally possible in this example, while at the same time producing a positive PNV. Meeting the fire behavior and effects targets proved to be the most difficult to meet in every period. A conflict when achieving multiple goals exists between creating large snags and maintaining large trees. A number of improvements in modeling could be made including recognizing species composition as a goal, developing a more sophisticated way of considering snags, and considering regeneration within the RLS-PATH function.
