Land cover and NDVI are important predictors in habitat selection along migration for the Golden-crowned Sparrow, a temperate-zone migrating songbird

BACKGROUND: Migrating passerines in North America have shown sharp declines. Understanding habitat selection and threats along migration paths are critical research needs, but details about migrations have been limited due to the difficulty of tracking small birds. Recent technological advances of tiny GPS-tags provide new opportunities to delineate fine-scale movements in small passerines during a life stage that has previously been inherently difficult to study. METHODS: We investigated habitat selection along migration routes for a temperate-zone migratory passerine, the Golden-crowned Sparrow (Zonotrichia atricapilla), given GPS tags on California wintering grounds. We used a resource selection function combined with conditional logistic regression to compare matched sets of known stopover locations and available but unused locations to determine how land cover class, vegetation greenness and climate variables influence habitat selection during migration. We also provide general migration descriptions for this understudied species including migration distance, duration, and elevation, and repeated use of stopover areas. RESULTS: We acquired 22 tracks across 19 individuals, with a total of 541 valid spring and fall migration locations. Birds traveled to breeding grounds in Alaska and British Columbia along coastal routes, selecting for shrubland and higher vegetation greenness in both migration seasons as well as grasslands during fall migration. However, model interactions showed they selected sites with lower levels of greenness when in forest (both seasons) and shrubland (fall only), which may reflect their preference for more open habitats or represent a trade-off in selection between habitat type and productivity. Birds also selected for locations with higher daily maximum temperature during spring migration. Routes during spring migration were lower in elevation on average, shorter in duration, and had fewer long stopovers than in fall migration. For two birds, we found repeated use of the same stopover areas in spring and fall migration. CONCLUSIONS: Using miniaturized GPS, this study provides new insight into habitat selection along migration routes for a common temperate-zone migrating songbird, contributing to a better understanding of full annual cycle models, and informing conservation efforts. Golden-crowned Sparrows selected for specific habitats along migration routes, and we found previously unknown behaviors such as repeated use of the same stopover areas by individuals across different migratory seasons.