Scaling disturbance instead of richness to better understand anthropogenic impacts on biodiversity.

A primary impediment to understanding how species diversity and anthropogenic disturbance are related is that both diversity and disturbance can depend on the scales at which they are sampled. While the scale dependence of diversity estimation has received substantial attention, the scale dependence of disturbance estimation has been essentially overlooked. Here, we break from conventional examination of the diversity-disturbance relationship by holding the area over which species richness is estimated constant and instead manipulating the area over which human disturbance is measured. In the boreal forest ecoregion of Alberta, Canada, we test the dependence of species richness on disturbance scale, the scale-dependence of the intermediate disturbance hypothesis, and the consistency of these patterns in native versus exotic species and among human disturbance types. We related field observed species richness in 1 ha surveys of 372 boreal vascular plant communities to remotely sensed measures of human disturbance extent at two survey scales: local (1 ha) and landscape (18 km2). Supporting the intermediate disturbance hypothesis, species richness-disturbance relationships were quadratic at both local and landscape scales of disturbance measurement. This suggests the shape of richness-disturbance relationships is independent of the scale at which disturbance is assessed, despite that local diversity is influenced by disturbance at different scales by different mechanisms, such as direct removal of individuals (local) or indirect alteration of propagule supply (landscape). By contrast, predictions of species richness did depend on scale of disturbance measurement: with high local disturbance richness was double that under high landscape disturbance.