Fig 1.
Study area for the annual late-summer survey of golden eagles across the western USA.
In every year during 2006–2013 we attempted to survey all primary transects, but alternates were flown when portions of primary transects could not be flown due to forest fires or weather events. Image background is from The National Map (United States Geological Survey).
Table 1.
Estimated probability of detection (SEs in parentheses) of golden eagle groups from each side of the aircraft based on surveys conducted 2006–2013 [2].
Three observers (back-left, and front-right and back-right) were present on surveys with exception of 68 transects in 2008, where only a back-left observer and a front-right observer were available.
Table 2.
Covariates considered for the model of late-summer habitat use by golden eagles.
Table 3.
Land cover covariates and their definitions considered for the model of late-summer habitat use by golden eagles.
The proportion comprised by each covariate was determined for every 2- × 10-km sampling unit. Modified from land cover classes and their definitions in the 2006 National Land Cover Database [32].
Fig 2.
Predicted intensity of use by golden eagles as a function of covariates in the final model.
Covariate values not represented in each graph were held constant at their median values, and final estimates were scaled so the maximum value was 1.0.
Fig 3.
Maps depicting the four covariates in the final resource selection function.
The final model contained covariates representing the median elevation (top left), median annual average solar radiation (top right), proportion of forested and developed land (bottom left), and median wind speed class (bottom right) in each sampled 2-km x 10-km rectangular cell within the study area. Image background is from The National Map (United States Geological Survey).
Table 4.
Estimates and upper and lower limits for 90% confidence intervals for coefficients in the final model of late-summer habitat use by golden eagles.
Fig 4.
Expected versus observed proportion of golden eagle observations for the 2012 (grey circles) and 2013 (black circles) surveys in 10 habitat use bins.
A perfect relationship between expected and observed use would occur along a line with an intercept of 0 and a slope of 1 (dashed line). The fitted relationship between observed and expected is shown as a solid line, and results of a Spearman’s rank correlation analyses (rs) are provided for each year.
Fig 5.
Predicted intensity of use by golden eagles in the western United States during late-summer.
Image background is from The National Map (United States Geological Survey).