Cost-effective conservation decisions are robust to uncertainty in the species-area relationship.

Guilhaumon et al. (1) use an extensive dataset to determine the most appropriate functional form for species–area relationships (SARs) among terrestrial plants and vertebrates. Their discovery that the traditional power-law relationship is not the most appropriate model in the majority of ecoregions raises compelling macroecological questions. However, their subsequent conservation analyses are based on outdated hotspot methods (2), and the ‘‘significant concerns’’ they raise about the use of power-law SARs in conservation prioritization may therefore be premature. In addition to biodiversity values, conservation resource prioritization depends critically on the heterogeneous costs of action (3), as well as the existing land use distribution (e.g., the amount of land reserved) in each priority region. The hotspot approach used by Guilhaumon et al. (1) considers neither of these factors. A more appropriate measure of conservation priority is the return on investment (ROI) that can be secured in each region (4). We investigated whether ROI priority rankings depend on the functional form of the SAR, using the three best models identified by Guilhaumon et al. (negative exponential, Monod, and power-law). We calculated ROI rankings for the globe’s 34 biodiversity hotspots (5) using endemic bird species richness as a proxy for biodiversity value. Once we include costs and existing reserves, the precise functional form of the SAR has little effect on the relative priority of the hotspots [Spearman rank correlation: r(P vs. N) 0.904; r(P vs. M) 0.903; r(N vs. M) 0.990]. From a conservation perspective, the different SARs all propose that ROI diminishes in a similar manner as the regional allocation increases.