Henneman, C. 2006. Habitat associations of red-shouldered hawks in central Minnesota landscapes. M.S. Thesis, University of Minnesota. 56 pp.
The red-shouldered hawk (Buteo lineatus) is a species of special conservation concern in much of the Great Lakes region, although little information exists about its habitat associations at the landscape scale. I conducted repeated call-broadcast surveys and nest searches in central Minnesota in 2004 and 2005 to assess habitat characteristics associated with red-shouldered hawk nest sites and occupancy. For call broadcast surveys, I estimated the probability of detection and occupancy, and assessed habitat associations at 2 spatial scales (100 and 314-ha circular plots), which were based on reported minimum and maximum red-shouldered hawk home-range size. To evaluate red-shouldered hawk habitat associations at nests, I used standard logistic regression methods to compare nests sites to random sites at 3 spatial scales (25-ha, 100-ha, and 314-ha circular plots). I estimated habitat amount, average patch size, patch density, edge density, and habitat diversity at all 3 spatial scales. I chose 4 study areas that represent a gradient of habitat conditions, from large, contiguous tracts of mature forest to small, isolated stands that have been fragmented and reduced in size, mostly due to timber harvest. In 2004, I conducted call broadcast surveys at 128 locations in 2 study areas, and in 2005, I surveyed 247 locations in 4 study areas. Estimates of probability of detection ranged from 0.1747 to 0.7500 and occupancy ranged from 0.5948 to 1.00 across years and study areas. I found a total of 68 red-shouldered hawk nests at 3 study areas in 2004 and 2005. For both nest sites and call-broadcast survey locations, I developed models relating habitat characteristics at multiple spatial scales to red-shouldered hawk nest site use and occupancy, and assessed support for these models using an Information-Theoretic framework. Overall, the amount of non-forest (grass,clear-cut area, forest <5 years old) and the amount of mature deciduous forest (>40 years old) had the strongest association with red-shouldered hawk occupancy and nest sites, but their importance varied across years, study areas, and survey techniques. The amount of non-forest was negatively correlated and amount of mature deciduous forest was positively correlated with red-shouldered hawk occupancy and nest-sites. Red-shouldered hawk nests in central Minnesota were associated with the amount of mature deciduous forest in combination with low levels of non-forest. With call broadcast surveys, red-shouldered hawk occupancy was either associated with amount of mature deciduous forest or limited amount of non-forest, rather than the combination of both, as observed for nest sites. Other metrics describing patterns of mature deciduous forest, such as the number of patches, mean patch size, and landscape diversity were retained in some best-supported models and may be important in red-shouldered hawk-habitat associations. Based on circular plots surrounding nests, the lower limit of mature forest (including mature deciduous and mature coniferous) at red-shouldered hawk nests was approximately 30% and did not vary across spatial scale. Most nests and call broadcast sites with red-shouldered hawk responses were associated with ≥40% and averaged approximately 50% mature deciduous forest. My findings suggest that red-shouldered hawks are associated with a high proportion of mature forest and a small proportion of open, non-forested areas across a range of spatial scales.