Application of Habitat and Occupancy Modeling to a Wood Duck Next Box Program

Programs to augment wood duck (Aix sponsa) nesting habitat by providing artificial nest boxes are commonly implemented. In northern Utah, where such programs are relatively new, I proposed a method to identify sites suitable for deployment of next boxes through a combination of habitat and occupancy modeling using site-specific, biotic, and abiotic, data collected from 105 next boxes over one nesting season in Cache County, Utah. An inductive habitat model was first developed which identified possibly suitable habitat (8.74% of county) based on proximity to hydrologic features. Next, based on comparing competing single-species, single-season, occupancy models using a modified Akaike\u27s Information Criterion (AICc), I found that the three variables most affecting wood duck nest box occupancy in the study area were 1) previous establishment of next boxes in the area; 2) the presence of the European starling (Sturnus vulgaris), an exotic and competitive passerine; and 3) the proximity of the nearest hydrologic feature. Using back-calculated estimates of wood duck occupancy of nest boxes, I found that estimated wood duck occupancy was maximized by deploying next boxes nearest to water if European starlings were present and between 40-160 m from water if European starlings were absent. Although the rigor of my occupancy estimates was limited by the study design used, these results can inform nest boxy deployment site selection base on spatial considerations (i.e., placement of nest boxed in relation to water features and historic concentrations of boxes), ecological considerations (i.e., competition with non-target wildlife species), and provide a starting point for future research involving this little-studied wood duck populatio