Breeding Bird Communities Across an Upland Disturbance Gradient in the Western Lake Superior Region

Abstract

Upland breeding bird communities were sampled from 225 points in 15 survey routes in the coastal region of western Lake Superior to examine relationships to human land use. Eighty-four species were detected and 50 were abundant enough to be included in data analysis. Monotonic quadratic regression models were constructed for these 50 species by using species counts as the dependent variable and the proportion of human conversion of the landscape (residential, agriculture, and commercial/industrial land uses) within each study area as the independent variable. Twenty-seven bird species had significant regressions (P < 0.05), 18 of which generally avoided areas developed by humans and 9 of which were attracted to development. Detrended correspondence analysis using counts of these 27 bird species was used to investigate multivariate, community responses to development. The first DCA axis was interpreted as a gradient from urban avoiding to urban exploiting bird species and was strongly correlated with land cover variables related to human development. Our results advance the idea that breeding bird communities can be used as indicators of ecological condition and can diagnose potential causes for changes in these conditions. Further, our study points out the usefulness of bird monitoring data in regional planning efforts that incorporate goals for maintaining native biological diversity.