Birds of a feather: Interpolating distribution patterns of urban birds

Geostatistical methods provide valuable approaches for analyzing spatial patterns of ecological systems. They allow for both the prediction and visualization of ecological phenomena, a combination that is essential for the conceptual development and testing of ecological theory. Yet, many ecologists remain unfamiliar with the application of these techniques. Here, we apply the methodology of geostatistics to an urban avian census in order to investigate and illustrate the utility of these tools. We derive habitat probability maps for three bird species known to differentially occupy the urban to rural gradient within the Phoenix metropolitan area and surrounding desert (Arizona, USA). We aggregated avian censuses conducted seasonally at 40 sites over two years and applied two processes process of interpolation, ordinary Kriging and indicator Kriging, and compared both methods. Ordinary Kriging interpolates values between measurements; however, it requires normally distributed data, which is commonly invalidated in ecological censuses. While indicator Kriging is not able to produce numerical predictions of measurements, it has the advantages of not requiring normally distributed data and requiring fewer statistical decisions. Each of the species exhibited strong deviations from normality due to many observations of zero. Given the skewness of the data, we anticipated that indicator Kriging would be a more appropriate method of interpolation. However, we found that both methods adequately captured spatial distribution of the three species and are sufficient for creating distribution maps of avian species. With additional census monitoring, Kriging can be used to detect long-term changes in population distribution of avian and other wildlife populations.