Habitat fragmentation and the species–area relationship: a focus on total species richness obscures the impact of habitat loss on habitat specialists

Aim The species–area relationship (SAR) is widely used in conservation science to predict the number of species likely to go extinct as a result of habitat loss. Often, studies employing the SAR use total species richness as the dependent variable. However, this overlooks the fact that habitat specialists and generalists differ in their susceptibility to habitat loss. We undertook a synthetic review of 23 habitat island datasets for birds to determine the impact of habitat generalists on the SAR. Location Global. Methods We sourced 19 habitat island datasets from the literature and combined these data with four of our own empirically gathered datasets. For each dataset, we classified all bird species as either forest habitat specialists or generalists. We then fitted the power SAR model (log–log and nonlinear forms) to the specialists, generalists and all species for each dataset and compared the resulting model parameters. We compared differences in the rate of change in richness with area between specialists and generalists using the first derivative of a multimodel SAR. Results We found that the slope of the power model was steeper for habitat specialists in the majority of datasets, and this difference was significant in 15 and 16 of the 23 datasets, for the nonlinear and log–log forms of the power model, respectively. Comparison of the multimodel SAR curve derivatives revealed further differences in the rate of change in species richness with area between subsets. Main conclusions The z values of both forms of the power model of the specialists’ SARs were generally larger, often considerably so, than the values used in most SAR studies predicting extinctions from habitat loss. Thus, studies that have used z values derived from SAR studies using total richness may be underestimating the impact of habitat loss on