Species-area curves, homogenization and the loss of global diversity

Species–area relationships have been widely used to examine the latitudinal gradient in species diversity, the effect of habitat loss on species diversity and the impact of exotic species on species diversity. Using species–area curves, Rosenzweig concluded that the spread of exotic species will not affect long-term global diversity much. We disagree with this conclusion on several grounds. First, total terrestrial area is larger than any of the provinces that make it up. This makes extrapolation necessary to predict global diversity using species–area curves, and extrapolation is statistically unjustifiable. Second, z, the slope of the species–area relationship, does not necessarily equal 1, as Rosenzweig assumes. Third, the amount of variation in species diversity explained by area is small and the prediction intervals around the estimated global diversity are likely to be too large to be used for prediction. Fourth, complete homogenization of the world’s biota is unlikely to occur, which makes the prediction of global diversity in the steady state even more difficult. This also makes the predicted transitional regional diversity lower than under complete homogenization. Finally, and perhaps most importantly, we believe that focusing on species diversity is not the best way to think about the impact of homogenization. Instead, we need to focus on the impact of invaders on the abundance and distribution of native species and on the functioning of ecosystems.