Donate Sign up for e-network
CENTER for BIOLOGICAL DIVERSITY Because life is good

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player


An Occasional Review of Goshawk Research and Conservation Issues

#4: Summary of Good (1998) by Noah Greenwald and Kierán Suckling

Good, R.E. 1998. Factors affecting the relative use of northern goshawk (Accipiter gentilis) kill areas in south-central Wyoming. MS thesis, University of Wyoming, Laramie.

Eight male goshawks fitted with transmitters and tipswitches were monitored during the breeding seasons of 1996-1997 in the Sierra Madre and Medicine Bow Mountains of south-central Wyoming. The study area is dominated by lodgepole pine and aspen. Ponderosa pine was historically more widespread in the area, but because of logging currently occupies a reduced distribution. Goshawks were monitored from noon to noon (during daylight hours) for 14-day intervals using three towers placed on tall knobs. Use of tipswitches, which produce a different signal strength depending on whether the bird is perching or flying, allowed for determination of when goshawks were foraging. Locations where goshawks killed prey (kill sites) were identified by locating foraging goshawks just prior to delivery of prey to the nest, which was documented by observers in a blind.

Study objectives included: (1) determining the importance of prey abundance, habitat characteristics, and landscape structure as factors in the relative use of goshawk kill areas within 300 m and 1 km of kill sites; and (2) determining whether there are significant differences in the landscape characteristics of kill sites compared to random sites within goshawk home ranges.

Only territories where three or more kill sites were identified were used for analysis, including 18 kill sites in five territories.

Core Area Use Within Home Range. Goshawks foraged within their home ranges selectively (p=0.02), returning to kill sites (n=30) an average of 4.12 times more often than to random sites. In addition, goshawks reused at least one kill site twice as often as others identified within their home ranges.

Selection Between Kill Sites. On average, habitat characteristics had a greater influence on goshawk use of kill sites than prey abundance. Goshawks returned more often to kill sites with greater densities of large trees (23 centimeters to 37.5 centimeters d.b.h.), greater density of large conifer trees (23 centimeters to 37.5 centimeters d.b.h.), gentler slopes (mean = 24 percent), and less shrub ground cover (mean = 17 percent).

The author did not provide data on habitat conditions at random sites within the home ranges, but noted that lodgepole pines of 23 - 37.5 cm are larger than average on the landscape, while open understories are common. The percent of ground covered by shrubs was not correlated with density of conifers, so were likely independent correlations. The percent of slope was not correlated with either shrub cover or large tree density, but was highly correlated with distance from the nest. Goshawks, therefore, may have been selecting for nest-stand proximity rather than for slope. The slope correlation may also be confounded by the biased selection of goshawk study territories for topography that allowed efficient radio tracking.

Average distance to ecotone edge was 40 meters and was not correlated with use of kill sites. Average distance to natural opening was 46 meters. The telemetry resolution was not sufficient to determine if goshawks actually hunted in edges.

Average canopy cover of kills sites was 53 percent. Use of kill sites was not correlated with differences in canopy cover. The author described the average kill site canopy cover as moderately high and high, but did not provide data on canopy cover at random sites within the home ranges.

On average, goshawks did not return more often to kill sites with higher prey abundances. Two of the birds, however, did return most often to sites with very high abundances of prey. One bird returned 12 times to a site with relatively high numbers of red squirrels, American robins, and least chipmunks versus six returns to each of two other kills sites in his home range. According to the author, however, it is unclear whether the goshawk was using this site because of high prey abundance or because it had more large trees (23 centimeters to 37.5 centimeters d.b.h.) than the less often used sites. Another bird returned 13 times to a site dominated by aspen with relatively high numbers of least chipmunks and medium sized birds versus returning a total of twice to three other kill sites. This was the only bird whose core use area was negatively correlated with large trees (23 centimeters to 37.5 centimeters d.b.h.).

Four red squirrels, one lagomorph, one least chipmunk, one unidentified mammal (not a red squirrel), one unidentified small mammal (vole sized), two unidentified birds, and nine unidentified prey items were killed by goshawks. Red squirrels, American robins, least chipmunks, and deer mice were the most common species at kill sites. Golden-mantled ground squirrels, northern flickers, and hairy woodpeckers were not found at many sites.

Landscape Analysis of Kill Sites and Kill Sites Versus Random Sites. Goshawks significantly used kill sites more often than random sites in their home range at both the 300 meter (412 percent) and 1 kilometer (868 percent) scales.

Proximity to Nest. Kill sites were reused more often if they were closer to the nest with most kill sites within 2,500 meters of the nest (range = 162 meters to 4,546 meters, mean = 1,885 meters).

Forest and Conifer Patches. Kill sites were reused more if they had greater coverages and larger patches of forest and conifer forest and the density of conifer patches was higher on the most used kill sites at both the 300 m and 1 km scale. Kill sites, however, did not have greater coverages of conifers than random sites.

Natural Openings. Percent cover of natural openings was lower and the openings were smaller on the most heavily used sites. Similarly, kill sites had smaller and fewer natural openings than random areas. Goshawks, however, revisited areas more often with greater density of smaller natural opening patches at the 1 km scale and conifer and natural opening patches were closer in kill areas than random sites, suggesting openings may have been clustered in goshawk kill sites.

Clearcuts. On average, goshawks re-used kill sites more often with greater coverage and larger patch sizes of clearcuts at the 300 m and 1 km scale. The author, however, attributed this use to the presence of mature forest adjacent to clearcuts. In addition, kill sites had smaller patches of clearcuts than random sites at 300 m and 1 km scales and on average only 2% of goshawk kill sites were covered by clearcuts.

General patch characteristics. Goshawks re-used kill sites more often with greater total density of patches and more patch diversity at the 300 m and 1 km scale. Total patch density and diversity, however, was lower in kill sites than random sites.

Other features. The more used sites had smaller aspen patches. Goshawks revisited areas most often with less coverage of agricultural lands at 1 km. Goshawks revisited areas more often at higher elevations at the 300 m scale, but the average elevation did not differ between random and kill sites. Kills sites had less road acreage than random sites.

Photo © Robin Silver