Table 2. Environmental covariates considered for generating species distribution models for the fisher in the Pacific States using Maxent.

The 6 covariates selected for Maxent modeling are shaded gray.

Category	Covariate	Reference	Description
Climatic	Winter snowfall (SNOW)	[46]	Average precipitation (mm) as snow during winter (Dec-Feb) from 1971 to 2000. Derived from 4 x 4-km PRISM data and downscaled to 250 x 250-m pixels.
Topographic	Topographic position index (TPI)	[47]	Estimate of terrain ruggedness, based on differences in elevation between a cell and the cells in the surrounding neighborhood. Negative values indicate valleys or canyon bottoms, and positive values indicate ridges or hilltops. Calculated from a 250 x 250-m digital elevation model.
Vegetative	Forest biomass (BIOMASS)	[40]	Aboveground live forest biomass (mg/ha). Used at the original resolution of 250 x 250-m pixels.
	Vegetation class (VEGCLASS)	[39]	Forest vegetation classes based on canopy cover, basal area of hardwoods, and quadratic mean diameter of dominant and codominant trees. Derived from 30 x 30-m raster data resampled to 250 x 250-m pixels using a majority algorithm.
	Stand age (AGE)	[39]	Basal area weighted stand age (year). Derived from 30 x 30-m raster data resampled to 250 x 250-m pixels using bilinear interpolation.
	Canopy cover (CANCOV)	[39]	Canopy cover of all live trees (%). Derived from 30 x 30-m raster data resampled to 250 x 250-m pixels using bilinear interpolation.
	Diameter diversity index (DDI)	[39]	Diameter diversity index of forest stands based on tree densities in different DBH classes. Derived from 30 x 30-m raster data resampled to 250 x 250-m pixels using bilinear interpolation.
	Quadratic mean diameter (QMD)	[39]	Quadratic mean diameter of all dominant and codominant trees (cm). Derived from 30 x 30-m raster data resampled to 250 x 250-m pixels using bilinear interpolation.