Landscape level raptor habitat associations in northwest Connecticut

Abstract

<p>I measured habitat characteristics at 21 barred owl, 30 northern goshawk. and 17 redshouldered hawk nest areas, and at 49 random areas within a 780 km2 area in northwest Connecticut. Measurements were taken at a macro ground level and from 1:12,000 and 1:40,000 black and white and 1:58,000 and 1:126,000 color infrared aerial photography. Among-species nest area analyses at the ground level indicated that higher beech and red maple densities were associated with red-shouldered hawk nest areas than with nest areas of barred owls or northern goshawks. Analyses between species and random areas indicated that barred owl nest areas had higher sugar maple density and basal area than random areas. Denser canopy, greater tree density and basal area, a greater percentage of forested land, and greater hemlock density were associated with nest areas of both northern goshawks and red-sl1ouldered hawks. Northern goshawk nest areas had higher red maple and conifer densities and were found farther from human activity and forest openings than random (available) habitat. Red-shouldered hawk nest areas had higher beech and deciduous tree densities and basal areas than expected based on available habitat. Six of 21 habitat characteristics measured from 4 scales of aerial photography were significantly different either among species nest areas or between species and random areas. Northern goshawks nested in areas with little open water. Red-shouldered hawks had higher percentages of open water near nest sites and were significantly closer to water than available habitat. Four variables related to contiguity of the forest were Significant in the analyses. The greatest number of cover type patches generally was associated with random areas followed by barred owl then redshouldered hawk nest areas. Red-shouldered hawks and northern goshawk nest areas were in relatively contiguous tracts of forested land and farthest from forest openings. The highest percentages of agricultural and non-forested land were found at random areas and barred owl nest areas, with lowest percentages at red-shouldered hawk and northern goshawk nest areas. The percentage of open water and swamp were most easily identified and measured from 1:58,000 color infrared aerial photography. Distance to water was most consistently measured from 1:58,000 CIR or 1:40,000 B&W aerial photography. The distance to opening was measured most consistently from the 1:12,000 scale. The 1:40,000 and 1:12,000 scales of aerial photography provided the most consistent measurements of the percentages of agricultural and non-forested land.</p>