Testing the importance of patch scale on forest birds

The relationship between population density and habitat area is of central importance to conservation biology, particularly for species dependent on declining habitat. A recent study by Lee et al. in Oikos in 2002 found that the densities of three forest interior bird species (ovenbirds, wood thrushes, and red-eyed vireos) decline with increasing patch size, contradicting many other studies that demonstrate a positive correlation between area and density for these species. The authors' argument is based on a misapplied criticism of the prevailing approach based on density-and on an inappropriate statistical methodology, caused by incorrect specification of the null hypothesis after log–transformation. We use three different methods of testing area-dependence: a corrected log–abundance log–area regression; a similar abundance–area regression; and the prevailing density-area approach. For each species, all three methods agree broadly, suggesting the robustness of each approach. For ovenbirds and wood thrushes, we do not find that population density is negatively correlated to patch size. While our analyses of the red-eyed vireo data find a negative correlation between the two factors, the strength of the correlation is far weaker than that of Lee et al. and may derive from landscape factors unconsidered in the original data. Nevertheless, we do not find positive density–area relationships for any of these forest-interior species, further underscoring the site-specificity of the underlying mechanisms of area- sensitivity.