The objective of this work was to analyze the implications of the model II, one of the strategies that have been used for generation of forest management alternatives, in the formulation and solution of problems of forest planning with covering constraint. The forest management problem was formulated via model I and model II, resulting in linear integer programming problems with 63 and 42 alternatives of management, respectively. As expected, there was no singularity constraint violation in the problem formulated via model I, while in the problem formulated via model II some management units were fractionated. This was already expected, since this formulation did not assure the integrity of the management units. To assure the integrity of the management units in the formulation via model II, it was necessary to reformulate the problem as a non-linear integer programming problem, which has a more complex solution than the linear integer programming problems. The efficient solutions of non-linear integer programming problem come up against the efficiency limitations of the main exact solution algorithms and in the lack of applications of approximative algorithms in the solution of this type of problem, such as simulated annealing metaheuristics, tabu search e genetic algorithms, becoming, therefore, attractive for this research field.
Forest regulation; non-linear integer programming; forest planning
