A spatial habitat model for the marsh-breeding red-winged blackbird (Agelaius phoeniceus L.) in coastal Lake Erie wetlands

A spatial model was developed to predict the distribution of a marsh-nesting bird species, red-winged blackbird (Agelaius phoeniceus L.) in Lake Erie coastal wetlands. The study site was in two diked wetland basins in Ottawa County, Ohio, USA. Geographic Information System (GIS) was used for data handling, storage, analyzing and running the model. Nests of birds were located, and vegetation durability, vegetation height, vegetation density, water depth, open water and structural aspects of the study basins were quantified by low altitude aerial photography and extensive field work for ground truthing. Spatial variables such as distance to open water edge, distance to wetlands edge, and friction maps by spread over depth and stem density for simulating predator access to wetlands were generated using the GIS. Model variables and coefficients were determined using logistic regression. Vegetation durability, distance to open water and depth of water significantly (P < 0.01) affected the nesting outcome. Results show that the probability of a bird nesting in a location within a marsh increased with increasing vegetation durability, water depth, and with decreasing distance to open water edge. Model verification and validation indicated that the model did not show any interspecific interactions or clumping of nest sites defined by behavioral constraints. Results were in accordance with the conclusions of previous studies looking at breeding success of red-winged blackbird. This study validated the processes affecting nest site selection of red-winged blackbirds with extensive data and statistical analyses. The methodology and versatile character of the GIS-based spatial model enables adaptations to any other marsh-nesting bird species and geographic area. Data needed for the model are relatively easy to obtain and the model provides micro-habitat scale results that can be used by wetland conservationists and managers.