This dataset consists of spatially anonymized movement data as well as environmental covariate data to estimate energy landscape step selection selections for migratory golden eagles that summer in Alaska.

1. Human modification of landscapes includes extensive addition of linear features, such as roads and transmission lines. These can alter animal movement and space use and affect the intensity of interactions among species, including predation and competition. Effects of linear features on animal movement have seen relatively little research in avian systems, despite ample evidence of their effects in mammalian systems and that some types of linear features, including both roads and transmission lines, are substantial sources of mortality.
2. Here, we used satellite telemetry combined with step selection functions designed to explicitly incorporate the energy landscape (el‐SSFs) to investigate the effects of linear features and habitat on movements and space use of a large soaring bird, the golden eagle Aquila chrysaetos, during migration. Our sample consisted of 32 adult eagles tracked for 45 spring and 39 fall migrations from 2014 to 2017.
3. Fitted el‐SSFs indicated eagles had a strong general preference for south‐facing slopes, where thermal uplift develops predictably, and that these areas are likely important aspects of migratory pathways. el‐SSFs also provided evidence that roads and railroads affected movement during both spring and fall migrations, but eagles selected areas near roads to a greater degree in spring compared to fall and at higher latitudes compared to lower latitudes. During spring, time spent near linear features often occurred during slower‐paced or stopover movements, perhaps in part to access carrion produced by vehicle collisions.
4. Regardless of the behavioural mechanism of selection, use of these features could expose eagles and other soaring species to elevated risk via collision with vehicles and/or transmission lines. Linear features have previously been documented to affect the ecology of terrestrial species (e.g. large mammals) by modifying individuals' movement patterns; our work shows that these effects on movement extend to avian taxa.

The variables X.1, X, and nothing are old indexes and should not be used for anything.  The variables k and tod were indicators previously used for subsetting the data.  The variable id is the individual identifier for each individual, and cluster identifies steps.  Use is a binary indicator of whether or not the point was used or available.  Gamma is the movement correlation parameter estimated from the CRW movement model.  Season is an indicator of fall and spring.  The majority of remaining columns are covariates used in the SSFs.  Any variables with raw in the header are the raw values, and any without are the centered and scaled quantities.