Selecting forest reserves with a multiobjective spatial algorithm

We applied a multiobjective spatial heuristic algorithm in a landscape area of 10 000 ha in eastern Finland, to select forest stands that would best complement the existing reserves in terms of achievement of conservation objectives while minimizing costs. Several non-spatial and spatial objectives for quality, area, and spatial distribution of forests were defined in cooperation with forest managers. Furthermore, existing reserves were pre-selected. The cost of each stand was calculated on the basis of the economic value of forest land and timber. The algorithm calculated cost-benefit ratios for the stands on the basis of how well each stand fulfilled the non-spatial and spatial objectives and at what cost. In every cycle, the algorithm selected the stand with the best cost-benefit ratio. A comparison between solutions, one without and three with differently weighted spatial objectives, showed that the network of selected stands was more clustered with than without spatial objectives. The use of the spatial objectives increased costs and the area slightly, and the non-spatial objectives were met almost equally in all solutions. The network of selected stands in one 2 100 ha sub-region was more clustered when stands were selected from the whole region rather than only from that sub-region. (C) 2003 Elsevier Science Ltd. All rights reserved.