Analysis of Bird Distribution Patterns in Forested and Agricultural Landscapes of North-Central Wisconsin
Andrew G. Cassini
This study uses a GIS approach for assessing bird-habitat relationships in two geographic areas of north-central Wisconsin. Field observers used standard point counts to survey birds during the breeding season and again during fall migration. Sites were randomly selected within a predominately forested northern region (MESA North) and the more agricultural region (MESA Central). Habitat classifications were based on a GIS database developed by the Wisconsin Initiative for Statewide Cooperation on Landscape Analysis and Data (WISCLAND). Land cover categories included: high intensity urban, low intensity urban, agriculture, grassland, coniferous forest, deciduous forest, mixed forest, emergent wetland, shrub wetland, forested wetland, bare soil and shrubland. GIS buffers of 100 m, 500 m and 1000 m were drawn around point counts to generate landscape variables for predicting bird occurrences. Logistic regression models were used to assess bird/land cover associations for all species appearing in at least 5 of the points.
During the study 137 species were recorded, 117 in MESA Central and 121 in MESA North. Relative abundance patterns followed a typical lognormal distribution. For many species, significant differences in abundance occurred between regions and among different landscape conditions. Models were highly successful in predicting occurrences of habitat specific birds such as Savannah Sparrow, Swamp Sparrow, Winter Wren, Alder Flycatcher, Red-breasted Nuthatch and Killdeer while performing marginally for more ubiquitous species such as American Crow or Blue Jay. This technique shows considerable promise for generating regional images of bird distributions and for assessing the impacts of landscape conditions on bird populations. The combination of field observations and land cover data can be used for broad-based management and conservation of birds, especially where comprehensive GIS information is available through resources such as WISCLAND.