Key points are not available for this paper at this time.
Abstract: The SLOSS debate can only be resolved by considering the optimal spatial configuration of a nature‐reserve system. Only when target species have been identified by such considerations can the size of the constituent reserves be determined. We explored two null models for a spatial configuration of reserves in the Cape Floristic Region, South Africa, based on the distribution of species in the family Proteaceae (Angiospermae: Rosidae). In the first model, reserves (12 × 13 km grid squares) were assigned at random for comparison with existing, proposed, and theoretically ideal configurations. In the second model, the theoretically ideal reserve configuration was constructed using the first model as a preassigned configuration. Two methods were used to iteratively select a reserve from the grid system: richness of unprotected species (species‐richness algorithm), and the sum of the rarity indexes for each unprotected species (rarity algorithm). Both algorithms yielded configurations requiring fewer reserves than those contained in proposed or existing configurations, although neither algorithm was perfect. The existing reserve configuration performed no better than the random model, assuming that the principal goal is protection of all species. We confirm that the best approach to designing a reserve configuration that maximizes the protection of species richness is to identify areas of high endemism and richness within particular biogeographical regions.
Rebelo et al. (Mon,) studied this question.