Scaling and habitat proportions in relation to bird diversity in managed boreal forests

Abstract

The study of the relevant scales for use in landscape biodiversity management involves several difficulties, like small sample sizes at large scales and varying trends between species. I studied richness in a pool of 25 bird species in relation to the amounts and coverage proportions of four forest habitats over eight scales based on data from forty-two 10 km transects. By adding habitat amounts from the transects, the accumulated species pool flattened out when about 400 ha of older forest was included. Transect pooling gave plot sizes that reflected plots ranging in size 10–80 km 2, and the correlation’s between bird species richness and the proportions of two good and two poor habitats were calculated at eight scales. For three of the habitat proportions, respectively, species richness was significantly influenced only at the smallest scale (10 km 2), while for old forest also at 20 km 2. These regressions changed slopes (positive or negative) from the smaller to the larger scales for three of the habitats. However, in areas of 10 km 2 bird diversity was significantly predicted (positive or negative) by the proportion of any single habitat (suitable or poor), while in areas of 20–80 km 2 it was not, which may suggest that to reach comparably high levels of species richness in large areas, a mixture of several habitats is decisive. Thus, management of forest bird diversity may not be efficient over broad areas if the proportions of only one or a few habitats are used.