Planning of production units for native forest management areas in the Amazon

Abstract

The objectives of this study were to present a model of integer linear programming (IP) to develop the creation of production units (PUs) regulated at the level of exploitable trees and to implement a heuristic method to obtain suboptimal feasible solutions. The studied area is in the municipality of Bujari, AC, Brazil. Using a sustainable forest management (SFM) area of 1,057.41 ha, 4,237 trees were selected for exploration through the census inventory. We applied the p-median model with volume and income restrictions; however, the adaptive heuristics development (AHD) is a random, greedy, and adaptive procedure that forms the PUs optimally. Four scenarios were formulated with a variation of ±10% to ±20% in volume and income values; the simulation was repeated 10 times with 10,000 iterations in each scenario. As expected, the major variations were in the scenarios with the most flexible restrictions of ±20%. Overall, 400,000 iterations were performed with average processing time of 523.74 s for each scenario. The methodology fostered the creation of PUs efficiently by grouping the trees into an optimized set, and at the same time, respecting the constraints of the problem.