Spatial Covariates
Several spatial covariates were used to delineate seasons and analyze resource selection. Landcover covariates were sourced from a 30 m resolution National Land Cover Database classification (USGS 2021) and reclassified into six landcover categories (water, exurban, grassland and scrub, forest, agriculture, and wetland) for seasonal delineation and four landcover categories (exurban, forest, agriculture, and other) for resource selection analyses. We also included a human modification covariate using a global layer which accounts for 13 anthropogenic global stressors at a 1 km resolution (Kennedy et al. 2019). We reprocessed a layer of Illinois streams and shorelines to create a Euclidean distance to water covariate at a 30 m resolution (Illinois State Geological Survey Prairie Research Institute 2015) and took the natural logarithm of the Euclidean distances to account for decreasing impact of a water source with increasing distance (Lehman et al. 2016). Similarly, we reprocessed a layer of Illinois paved roads to create a natural logarithm of the Euclidean distance to road covariate at a 30 m resolution (Illinois Technology Transfer Center 2020).