Detecting the scales at which birds respond to structure in urban landscapes

The authors "explored how the variation in percent tree canopy cover, at four different scales, affected the abundance of bird species across various urban sites in North America. [UMN]

"Little empirical information exists about how birds respond to urban landscape structure across multiple scales. We explored how the variation in percent tree canopy cover, at four different scales, affected the abundance of bird species across various urban sites in North America. Bird counts were derived from previous studies, and tree patches were measured from aerial photographs that represented areas of 0.2 km2, 1.5 km2, 25.0 km2, and 85.0 km2. At each of the four areas, we conducted regressions between bird counts and percent cover of various tree patch sizes. From these analyses, we determined the area (called the best prediction area—BPA) and the patch size (called the best patch size—BPS) that accounted for a significant amount of the variation in bird counts, beyond the variation accounted for by these parameters measured at other scales. BPA and BPS were calculated primarily to take into account the high degree of collinearity that existed among the amount of tree canopy cover measured across the four scales.

We calculated BPA and BPS values for a variety of bird species and ascertained whether larger species had relatively larger BPS and BPA values. In the spring, middle-sized to large birds (16.6 g–184.0 g) had relatively larger BPS values than did smaller birds (3.2 g–16.5 g), but in the summer, the largest birds (61.7 g–576.0 g) had small BPS values. Spring BPA values showed a similar result but summer BPA values did not. A majority of birds of all sizes had summer BPA values at the finer scales of 0.2 km2 and 1.5 km2. Overall, body size was an approximate predictor of the area and patch size at which a species responds to trees in a landscape, but many exceptions did occur. These exceptions could be related to a variety of factors, one being the difficulty in relating human-biased measurements to avian measurements of a landscape. The method described in this study will help researchers design multi-scale studies to address the effect of landscape pattern on different animal species." [Abstract]