Species-area relationships (SARs), the change in species numbers with increasing area, are an essential tool used to estimate broad biodiversity patterns, to compare species richness when regions differ in area, and to predict extinction rates following habitat loss. In the main, applications of SARs have assumed the classical form of a power function, S=cA^z, where S is species richness, A is area, and c and z are constants. If this assumption of a universal power SAR is released and if the underlying form of SARs actually differ markedly between biomes and between major taxonomic groups this would be of particular concern for ecology. Here we used data on the species richness of vascular plants and vertebrates across the world’s terrestrial ecoregions to conduct an analysis of global scale SARs. First, we calculated the probabilities of eight different models to best describe the SAR and determine whether those probabilities vary systematically across biomes and taxa. Second, we conducted a global identification of hotspots of richness incorporating the uncertainty about the best fit SAR model, and compare these results with those obtained when it is assumed that the SAR is described by a power model.

Results/Conclusions

Several issues with the assumption of a power SAR are highlighted by our results. First, this assumption overlooks the fact that there are situations in which no simple model may adequately describe the SAR. Second, all of the different shapes of SARs represented by the set of models used were selected as the best at least once for the different datasets. This suggests that none of a wide range of potential SAR models can a priori be ignored and that a universal model does not emerge. Third, there was substantial uncertainty as to which of SAR models was the best. This highlights the importance of considering multiple models when making inferences about SARs. Finally, assuming a power SAR, a rather different set of hotspots would be identified than is the case when alternative models are considered and there will tend to be systematic biases in these hotspots. In conclusion, we recommend that, in both empirical and theoretical analyses involving SARs, the relative fit of different models is examined, and uncertainty in this fit is accounted for. Failing to do so may well lead to conclusions and conservation prioritizations that are at odds with real patterns of spatial variation in species richness.