An evaluation of alternative strategies for building networks of nature reserves

Abstract The challenge to maintain biodiversity with limited funding and personnel puts a premium on building reserve networks that maximize species number per area. Here I compare two alternative strategies suggested by the SLOSS debate (accumulating nature reserves from small-to-large (STL) or large-to-small (LTS) to the best possible network of reserves. The best network was determined by a computer program (BEST) which considered all possible combinations of reserves and then selected the subset that satisfied a specified selection criterion. Here, the selection criterion was that the subset included all target species in the smallest total area. In addition, I compared alternative strategies to a null model of randomly accumulated reserves. For ten data sets including a diversity of biotas (nonvolant mammals, bats, birds, reptiles, amphibians, ants, isopods and plants), I found STL and LTS strategies to be equivalent in maximizing species number per area. To include all target species, however, both strategies required much more area than the best possible network and, in most cases, more area than random accumulations of reserves. These results argue strongly against generic prescriptions for conserving biodiversity. Given empirical studies on the biota in question, analytical solutions such as presented here will serve as valuable tools for preserving biodiversity.