Summary courtesy of Kieran Suckling, Center for Biological Diversity, http://www.biologicaldiversity.org/swcbd/species/goshawk/gentilis.html

--
Spencer, J.A. 1995. Goshawk nesting habitat on the Prescott
National Forest, AZ. M.A. Thesis. Prescott College, Prescott, AZ.

Nine nest stands, ten nest sites, eight nest trees, and eight nests were studied on the Prescott National Forest in central Arizona. [Editors note: this forest is small, relatively isolated, and was heavily logged in prior decades. The dominant forest cover is ponderosa pine, pine/oak and pine/juniper, but little of this is mature. Some patches of mature mixed-conifer remain in drainages. A larger proportion of the mixed conifer stands are mature than the ponderosa pine types. Relatively little logging has occurred on the forest in the last decade].

Nest stand size: Nest stands (delineated based on uniformity of stand structure, composition and topography) were larger than random stands. The mean stand size was 359 ha vs. 197 ha at random stands (range = 235 - 525 ha). The author believes that identification based solely on structure and composition, omitting topography, would result in larger stands. “It appears that goshawks are using the best available stands with respect to canopy cover and stand size.”

Nest stand characteristics: 40% were mixed-conifer, 25% were ponderosa pine/oak, 18% were pine/juniper, and 17% were pure ponderosa pine. Canopy cover averaged 57% overall, but 65% in mixed conifer stands, mean basal area was 29.0 m2/ha, prevailing aspects were south and east. Nest stands did not differ from random plots in basal area, slope, canopy layering, snag density, distance to clearings/developments, or aspect. Goshawks selected nests stands of mixed-conifer more often than were available. Understories were more open than on random stands, had significantly less total trees, grass/forb cover and shrub/small tree cover; significantly more dead and down wood, and were significantly closer to water. Canopy cover was significantly greater, but tree height was significantly smaller.

The mean dbh (27.1 cm) was uniform across nest stands, smaller than random stands, and smaller than generally recorded in goshawk literature. The author believes this may indicate that 27.1 cm may be a minimum dbh requirement on the Prescott National Forest.

Nest sites (0.08 ha): 50% were ponderosa pine/gamble oak, 25 % were mixed-conifer, 13% were pure ponderosa pine, and 13% were pine/juniper. Eighty-eight percent were single-storied. Canopy cover averaged 61%. Mean basal area was 40 m2/ha. Nine of the nest sites had a southern exposure. Nest sites were flatter (11%) than nest stands (23%) and random stands (19%) and had taller trees.

Nests. All nest were in ponderosa pines. Eighty-eight percent faced southeast or southwest, 13% faced northeast. No preference was shown for slope position. Anecdotal evidence suggests that nests in exposed positions are more prone to wind damage and predation.