GENTILIS

An Occasional Review of Goshawk Research and Conservation Issues

#6: Summary of Keane (1999) by Noah Greenwald

Keane, J.J. 1999. Ecology of the northern goshawk in the Sierra Nevada, California. Dissertation, University of California, Davis.

Keane studied three aspects of goshawk ecology in the Lake Tahoe, California region of the Sierra Nevada: the effects of climate and prey abundance on reproduction, size of breeding and nonbreeding home ranges, and nest-site habitat characteristics.

Reproduction Variables. Simple linear regression was used to compare variation in reproductive success (proportion of nests that fledged young) with the abundance of Douglas squirrel (a key prey species) and weather conditions (total precipitation, number of days with precipitation, mean temperature, etc.). Douglas squirrels were singled out for analysis due to their importance: They were the most frequently recorded prey species (23 percent of all remains) and contributed the most biomass to the goshawk’s diet (32.9 percent). Keane also hypothesized that as a non-hibernating, non- migrating species, the squirrel may provide especially important food when goshawks are building essential reserves for reproduction.

Spring Douglas squirrel abundance accounted for a high proportion of the between-year variation in the number of territories that successfully reproduced (adjusted R2 = .71), but the relationship was not significant at the 95-percent confidence level (P = .101). Mean low temperatures in February also explained variation in numbers of territories that successfully reproduced (adjusted R2 = .946, P = .018), with lower mean temperatures corresponding with years of lower reproductive success.

In both 1992 and 1995, high goshawk nest success corresponded with high Douglas squirrel abundance and increased frequency and biomass of the squirrel in goshawk diets. These years were preceded by years with high cone crop production, leading Keane to hypothesize a relationship between cone crop production, Douglas squirrel abundance and goshawk reproduction:

"In sum , I conclude that annual variation in northern goshawk reproduction
in this region of the Sierra Nevada is a result of an interaction between food
and weather such that reproduction is greatest in years following high cone
crop production, which affects Douglas squirrel abundance, and mild
late-winter/early springs with high temperatures and low total precipitation."

Keane suggested that management actions should be targeted towards maintaining both high cone crop production and Douglas squirrel abundance, which can be accomplished by sustaining mature forests:

"Factors related to cone crop production dynamics should be a management focus when considering management of habitat for Douglas squirrels. Cone crop production differs in both magnitude and frequency across tree size classes and between conifer species. Cone production is greater by mature conifer, in terms of both magnitude and frequency relative to younger, smaller conifers."

Home Range Size. Seven female and eleven male goshawks were radio-tagged and monitored for two to four years. Goshawks were found to be year-round residents. Home ranges were significantly larger in the non-breeding season than the breeding season. Non-breeding season home ranges were 9,379 ha. (sd = 6821) for males and 5,555 ha. (sd = 3289) for females. During the breeding, male home ranges were 2,698 ha. (sd = 1043) and female’s were 2,016 ha. (sd = 1690). Keane speculated that increased non-breeding period home range size may relate to reduced availability and abundance of prey and increased thermal stress, leading to a more energetically stressful period. In support of this conclusion, he noted that all five goshawk mortalities observed during the study occurred during winter. Because Douglas squirrels do not migrate or hibernate, Keane suggested that they are especially important winter prey that should be managed for:

"The evidence reported in this study and knowledge of the natural history
of important prey species suggests that Douglas squirrels may also be
particularly important prey during winter. These data suggest that while
management of northern goshawk habitat should consider the full spectrum
of prey species, emphasis should be placed on providing habitat that supports
widely distributed and abundant Douglas squirrel populations. A particularly
important component of Douglas squirrel habitat are large conifer trees and
mature/late seral forest stands that provide nest sites and also are the most
reliable producers of cones.”

Nesting Habitat. Habitat characteristics in 18 m circles around nest sites were compared to random points. Nest sites had significantly greater canopy cover, live trees between 60-100 cm dbh and >100 cm dbh, and dead trees between 60-100 cm dbh than random points. By contrast, random points had significantly greater herbaceous, rock, and shrub/sapling cover and live trees between 5-30 cm. Nest trees were larger than available trees and averaged 82 cm dbh (range = 39-155 cm). Based on stepwise logistic regression and repeated random samples of the random points, models (multiple runs) including canopy cover, live trees >100 cm dbh, and shrub/sapling cover best predicted goshawk nesting habitat. Keane concluded:

"Forest management practices such as selective timber harvest that
removes large diameter conifer trees and fire suppression, have likely
reduced the distribution and abundance of northern goshawk nest
habitat in the Sierra Nevada.”