You are here

Habitat quality for the northern goshawk

Status: 
Action
Dates: 
May, 2012

Management Context

A fledgling northern goshawk on the Kaibab National Forest.
A fledgling northern goshawk on the Kaibab National Forest.
The northern goshawk (Accipiter gentilis) is an apex predator in most forests in the United States and Canada. It is a 'species of concern' in all US Fish and Wildlife Service regions, a 'sensitive species' in all US Forest Service Regions, six Bureau of Land Management states, and numerous States, and is a 'Species at Risk' in the Canadian Province of British Columbia. The Forest Service has experienced three decades of litigation based on concerns that forest management negatively effects goshawk population viability by reducing its habitat quality.

This study is an opportunity to test the efficacy of forest management recommendations for creating and sustaining goshawk and their prey habitats in the western United States.

Approach

There is need to understand the relationship between goshawk demographic performance and the three-dimensional structure of its habitat. Towards addressing this need, goshawk demographic performance has been studied for 21 years (1991-2011) on 125 territories on the Kaibab Plateau, Arizona. Mark-recapture methods were used to estimate vital rates of goshawks, including territory occupancy, mate and territory fidelity, reproduction, recruitment, immigration, and juvenile and adult survival. These rates are necessary covariates for investigating habitat quality. The Kaibab goshawk study represents one of the longest and most extensive demographic data set on any population of this species in the world.

Map of the goshawk study area on the Kaibab Plateau.
Map of the goshawk study area on the Kaibab Plateau.
The study area is the Kaibab Plateau in Northern Arizona, a 700 mi2 area of ponderosa pine and mixed conifer forests, meadows, and forest openings (fire, wind-throw, timber harvest) that is managed by the Kaibab National Forest and the Grand Canyon National Park. In 2012, scientists acquired high resolution LiDAR (light detection and ranging) data to delineate horizontal and vertical (three-dimentional) vegetation structure for the entire Kaibab Plateau goshawk study area. LiDAR provides a mapping of the vertical and horizontal composition and structure of the vegetation within each territory, which then will allow an investigation of the relationship between goshawk demographic performance and identification of those vegetation elements that confer habitat quality.

The objectives are to determine the forest vegetation composition of the overstory and understory and 3-dimensional structure (e.g., tree height, canopy height, canopy base height, tree spacing, understory heights and densities, snags, logs, woody debris, patch size/juxtaposition) of forests in each of the 125 goshawk territories on the Kaibab Plateau. Researchers can then explore the relationship between the composition and structure of forests in territories and goshawk demographic performance using tools such as GIS, FRAGSTATS, CART, autologistic regression, proportional regression.

Forest composition and structure data will also allow researchers to determine the distribution of the habitats of important bird and mammal prey of goshawks, whose abundance is a critical factor limiting goshawk reproduction. This is an unequaled opportunity to investigate the relationship between a species' demography and the composition and structure of its habitat.

Future Direction

Recently, the research team started to put together a "consortium" of potential cooperators who might find the LiDAR data useful for their purposes and who might be willing to help finance its development. Examples of development include regression modeling of various forest metrics (tree species, tree diameters, down woody debris and snags), a process that requires sampling field plots. Regressing ground plot metrics on corresponding LiDAR-derived metrics provides an analytic means to extrapolate plot level relationships across the entire study area. Regression models developed in our study may also apply to other Southwest forests having similar forest species and structural compositions.

If you are interested in joining this project or helping develop these data please contact Richard Reynolds.

View looking south over a burned area at the fork of Moquitch and Snipe Canyons. Left image derived from LiDAR points colored by 2010 NAIP imagery. Right image derived from ground classified LiDAR points colored by elevation.
View looking south over a burned area at the fork of Moquitch and Snipe Canyons. Left image derived from LiDAR points colored by 2010 NAIP imagery. Right image derived from ground classified LiDAR points colored by elevation.

Publications

Reynolds, Richard T. ; Sanchez Meador, Andrew J. ; Youtz, James A. ; Nicolet, Tessa ; Matonis, Megan S. ; Jackson, Patrick L. ; DeLorenzo, Donald G. ; Graves, Andrew D. , 2013
Reynolds, Richard T. ; Graham, Russell T. ; Reiser, M. Hildegard , 1992


Principal Investigators:
Co-Investigators:
Collaborators:
Shelley Bayard - Rocky Mountain Research Station (previously)