Habitat reduction and patterns of species loss

Abstract

Abstract This paper uses data of The Distribution Atlas of Polish Butterflies to simulate the effect of four different types of area loss (aggregated, fractal, random, and nested) on species diversity and species–area relationships (SARs). We found that power function and logarithmic SAR models overestimated species loss in the case of aggregated, fractal, and random patterns of area reduction. Instead, the modification of the power function by Plotkin et al. (Proc. Natl. Acad. Sci. USA 97 (2000b) 10850) ( S = S 0 A z e − kA ) with k being a shape-adjusting parameter worked better and gave sufficient predictions of species loss. The net effects of the aggregated, fractal, and random types of area loss on species diversity were very similar with an aggregated pattern of area loss leading to slightly higher rates of species loss than both other processes. We conclude that SARs might be useful tools for biodiversity forecasting if they are constructed in a case-specific manner. The use of standard models instead might be misleading.