Winter Ecology of Cooper’s Hawks on Ames Plantation, Tennessee

Cooper’s hawks (Accipiter cooperii) were listed by Ganier (1933) as fairly common permanent residents of western Tennessee during the early 1900’s, but populations declined during the 20th Century presumably because of habitat changes, shooting, and effects of pesticides on reproduction (Newton 1979). Breeding bird survey data for Tennessee suggest that Cooper’s hawk populations are increasing (+4.6% increase/year, 1966-2000), although sample sizes are too small for significant trends (P = 0.39, Sauer et al. 2001). Nicholson (1997), in the Tennessee Breeding Bird Atlas completed during the early 1990\u27s, still described Cooper’s hawks as uncommon permanent residents. At the same time Cooper’s hawk populations appeared to be increasing, northern bobwhite (Colinus virginianus) populations were declining precipitously (-3.7 % decrease/year in Tennessee, 1966-2000; Sauer et al. 2001). Many quail biologists have suggested that the decline in quail populations may in part be linked to increasing avian mortality from increasing raptor populations (Hurst et al. 1996, De Maso et al. 1997, Rollins and Carroll 2001). The objectives of the hawk project on Ames Plantation were to describe Cooper’s hawk winter ecology, to describe seasonal patterns of raptor abundance and to gain breeding season data for Cooper’s hawks. The study area was Ames Plantation, Tennessee located in southwestern Tennessee. Ames Plantation is comprised of a variety of covertypes including hardwood forests, mixed forests, pine forests, crop fields, old fields, native warm season grasslands, hardwood conversion areas (areas consisting of savannahs where hardwood forests had been recently harvested, cleared of treetops and trunks, with only a few live trees remaining/ha), and fencerows. We trapped Cooper’s hawks with bal chatri traps baited with house sparrows(Passer domesticus) and fitted hawks with radio transmitters. Using radio telemetry, Global Positioning Systems (GPS), and Geographical Information System (GIS) procedures, diurnal and roost locations were obtained for 5 Cooper’s hawks. To account for telemetry locational errors, diurnal habitat use was determined by delineating covertypes within 50-m-radius circles around each daytime location. The composition of available habitat was also described within 50- m circles delineated around a systematic grid of random points. Habitat use was examined with compositional analysis (Aebischer et al. 1993) and chi-squared analysis (Neu et al. 1974). Logistic regression models were developed to identify the key habitat features that discriminated between roost locations and random sites. A 40-km raptor survey was conducted weekly during winters 1999/2000 and 2000/2001 to determine relative raptor abundance and species composition on Ames Plantation. All species seen or heard at each point and between points were recorded. Nonparametric statistics (Kruskal-Wallace H Tests) were used to compare the winters of 1999/2000 and 2000/2001, leaf-on and leaf-off surveys, and surveys on and off the field trial area. Nest observations and measurements were taken at 4 Cooper’s hawk nests, 4 redtailed hawk (Buteo jamaicensis) nests, and 1 red-shouldered hawk (Buteo lineatus) nest over two breeding seasons (2000 and 2001). A nest camera was set up at 1 red-tailed hawk nest for 2 days and 1 Cooper’s hawk nest for 2 days. We trapped for 533 trap hours on 34 days in winter 1999/2000 and 1,729 trap hours on 53 trap days in winter 2000/2001. We captured 35 raptors, including Cooper’s hawks, American kestrels (Falco sparverius), barred owls (Strix varia), northern harriers (Circus cyaneus), red-tailed hawks, red-shouldered hawks, and sharp-shinned hawks (Accipiter striatus). Nine Cooper’s hawks were trapped in winter 1999/2000 (59 h/capture) and 2 Cooper’s hawks in winter 2000/2001 (865 h/capture). Based on capture success, Cooper’s hawk abundance appeared to be much lower in winter 2000/2001 than in winter 1999/2000. The home range of the 1 male Cooper’s hawk tracked in the winter was 331 ha (95% minimum convex polygon) and the mean size of female home ranges was 836 ha and ranged from 8 ha to 2,529 ha. Diurnal winter habitats used by Cooper’s hawks were ranked in order of most preferred to least preferred as follows: Forests \u3e Edge \u3e Field \u3e Other, based on the compositional analysis (l=0.0722, F=12.84, P=0.0322; MANOVA). Cooper’s hawks used forests more than expected and fields less than expected compared to their availability (N=458, c2=68.76, df=8, P\u3c0.0001; Neu et al. 1974). Twenty-two roost sites were located in 1999/2000 and 34 roost sites were located in 2000/2001. Some of the roost sites were used repeatedly. Vertical cover (P=0.0332), canopy cover (P=0.0030), and canopy cover variance (P=0.0353) were greater in roost sites than random sites, whereas the height of the overstory (P=0.0674) was (marginally) lower in roosts than in random sites. Edges, fencerows, and pines were used more than expected for roosting, while hardwood stands and hardwood conversion areas were used less than expected (c2=30.27, P=0.0001). Radio-tagged Cooper’s hawks emerged from the roost site on average 38 min before sunrise and changed roost sites during the night 12.5% of the time. Cooper’s hawks roosted in dense honeysuckle (Lonicera japonica), briar (Rubus alleghaniensis), and cedar thickets (Juniperus virginiana) near the ground under canopies that were highly variable. A total of 1,671 individual raptors were recorded during 47 surveys (31 winter surveys). Red-tailed hawks (on average 4 - 14.5 birds were detected per survey by month) were the most abundant species while American kestrels (0 - 2.3), northern harriers (0 - 1.5), and red-shouldered hawks (0 - 4.3) were detected less. Cooper’s hawk (0 - 1) and sharp-shinned hawks (0 - 0.5) were detected the least. Total raptor abundance (all species) and red-tailed hawk abundance were greater in the second year (winter 2000/2001) than the first year (winter 1999/2000). Greater red-tail abundance the second winter likely occurred in response to the colder winter temperatures. Total raptor abundance, northern harriers, and red-tailed hawks were detected more on the field trial course, while American kestrels and red-shouldered hawks were detected more off the field trial course. The detection rate of Cooper’s hawks and sharp-shinned hawks averaged about 0.5 hawks per survey- too low to detect differences between years or on/off the field trial area. Visible areas along the 40-km survey route totaled 258 ha. Out of 4 Cooper’s hawk nests, 2 were confirmed to have failed while the other 2 have an unknown fate, but activity at late dates in the breeding season suggest a successful nesting effort. Out of 4 red-tailed hawk nests, 1 was confirmed successful, 1 was confirmed to have failed, 1 was probably successful, and 1 had an unknown fate. The success of the red-shouldered hawk nest is also unknown. The averages for the 4 Cooper’s hawk nests are as follows; nest height, 16.55 m; nest tree diameter, 53.08 cm; vertical cover, 6.06%; canopy cover, 48.8%; and basal area 30.58 m2/ha. The averages for the 4 red-tailed hawk nests are as follows; nest height, 18.7 m; nest tree diameter, 49.83 cm; vertical cover, 11.94 %; canopy cover, 62.73 %; and basal area 19.06 m2/ha. The measurements for the red-shouldered nest are as follows; nest height, 15 m; nest tree diameter, 53.75 cm; vertical cover, 12.5 %; canopy cover, 49.22 %; and basal area 34.39 m2/ha. Cooper’s hawks were found to primarily use forests in the winter at Ames Plantation. However, Cooper’s hawks and northern bobwhites occupied similar habitats (old fields and forest edges with a dense understory of honeysuckle and briars) a significant portion of each day. Prey remains (n = 19) from Cooper’s hawks were primarily passerines (n = 10), morning doves (Zenaida macroura) (n = 4), and northern bobwhite (n = 4). In both winters, radio-tagged Cooper’s hawks moved from Ames to a nearby plantation to prey on pen-reared quail that were released for weekend hunts