A COMPARISON OF RESERVE SELECTION ALGORITHMS USING DATA ON TERRESTRIAL VERTEBRATES IN OREGON

We compare the number of species represented and the spatial pattern o f reserve networks derived using five types of reserve selection algor ithms on a set of vertebrate distribution data for the Stare of Oregon (USA). The algorithms compared are: richness-based heuristic algorith ms (four variations), weighted rarity-based heuristic algorithms (two variations), progressive rarity-based heuristic algorithms (11 variati ons), simulated annealing, and a linear programming-based branch-and-b ound algorithm. The linear programming algorithm provided optimal solu tions to the reserve selection problem, finding either the maximum num ber of species for a given number of sites or the minimum number of si tes needed to represent all species. Where practical, we recommend the use of linear programming algorithms for reserve network selection. H owever, several simple heuristic algorithms provided near-optimal solu tions for these data. The near-optimality, speed and simplicity of heu ristic algorithms suggests that they are acceptable alternatives for m any reserve selection problems, especially when dealing with large dat a sets or complicated analyses. (C) 1997 Published by Elsevier Science Ltd