|Abstract or Summary
- Foresters in the Pacific Northwest are directing increasing effort into the management of young-growth forests. These foresters find themselves in a transition period in which they are beginning to develop the framework for managing young-growth forests while they continue to manage substantial reserves of old-growth timber. In this transition period, foresters have an excellent opportunity to experiment with alternative silvicultural practices and management systems. In order to help foresters cope with newly emerging problems of young-growth forest management, this study was designed for the purpose of developing and analyzing new methods of forest management planning, especially as applied to problems involved with planning thinning programs. In this study, these new methods of planning are based upon economic criteria and mathematical programming. It is desired to find out what difficulties are involved with applying these methods to a practical forest management situation and what benefits, if any, can be expected from their use. The planning problem assumed for this study involves a situation in which a manager must plan an annual thinning program involving the choice of a particular set of thinning opportunities from a larger available set. Complicating his choice are vague, often conflicting, administrative guidelines which place limitations on the conduct of field operations. The planning problem is formulated and solutions are obtained in terms of both linear and 0-1 integer programming. The two new planning systems are applied to a case study area in western Oregon where they are compared to current rule-of-thumb methods as well as being compared to each other. The linear programming approach (value type approach) incorporates a classification system which identifies thinning opportunities on the basis of variables thought to influence the value of these opportunities. The solution to the Linear program gives the number of acres of each type that should be thinned. Thinning opportunities are evaluated on the basis of their contribution to present net worth. The 0-1 integer programming approach (unit approach) views thinning opportunities as specific thinning harvest units. Each unit has definite boundaries, a specific geographical location, and is treated as a unit. The solution to the 0-1 integer program indicates which units should be thinned. Units are subject to a yes-no decision. A present net worth value is calculated for each unit based on the expected results of thinning. Results indicate that both systems have something different to contribute to planning. Linear programming has limited usefulness for planning thinning programs involving the specific location of harvesting units. Because of the lack of a definite tie to the ground, several alternatives and constraints cannot be adequately formulated, resulting in thinning programs that are not truly optimal. The value of the linear programming system seems to lie in more general problems, such as how thinning should be done with respect to certain key variables, such as age. Zero-one integer programming is quite useful for planning thinning programs when the location of specific harvesting units is involved. Solutions to this approach are quite definite as to where thinning should take place. Also, constraints related to specific locations are quite easy to formulate. However, in regard to more general problems, not specifically involving units, the 0-1 integer approach is of limited usefulness due to the large number of constraints imposed artificially by the units themselves. In addition to providing optimal thinning programs, the two systems provide an organization and documentation of information lacking in current rule-of-thumb methods. Moreover, they provide the means of rapidly obtaining optimal solutions under a wide range of possible assumptions leading to an enlarged potential for rational planning and increased returns.