Predictive Model of Golden Eagle Distribution in Idaho and Montana

Because of growing concern in recent years about the status of golden eagle
populations in western North America and the need for a more complete
understanding of the preferences of this species, we developed a Predictive
Model of Golden Eagle Distribution in Idaho and Montana. This preliminary
model follows a Resource Selection Functions (RSF) approach for golden eagles
and is based on data collected via satellite telemetry from 6 golden eagles
tracked in the western U.S from 1996-1998. The initial golden eagle predictive
model is done at a relatively low resolution; the dataset includes the
predictive layer and was created from 5 different environmental data layers
that were originally compiled and published elsewhere. GIS overlays were
reformatted and merged from diverse sources (including state GAP programs,
USGS, Snake River Field Station, ESRI and other sources). Aspect, slope, river
and road distance were derived from the original data layers. All work was
completed in ArcView 3.3. For model building we used MARS and CART from
Salford Systems.