Where the Whip-poor-will will: Third and fourth order habitat selection of a migratory aerial insectivore

Abstract

Habitat selection is a hierarchical process operating across multiple spatial scales, with important consequences for survival and reproduction. The Eastern Whip-poor-will (Antrostomus vociferus) is a declining aerial insectivore that relies heavily on crypsis, making habitat use difficult to define because of low detectability and inconspicuous behaviour. As a result, fine-scale habitat selection in this species remains poorly understood. I defined third-order habitat selection of breeding Eastern Whip-poor-will by manually tracking radio-tagged birds across the landscape in Norfolk County, Ontario. Specifically, I characterized (i) whip-poor-will breeding home and core range size, (ii) percent overlap of breeding pairs and neighbouring males, and (iii) defined habitat composition within home ranges to compare with available habitat. Whip-poor-will home and core range size did not differ between sexes, with breeding pairs showing extensive overlap, possibly a result of bi-parental care of the chicks. Though female sample size was small, limiting my ability to robustly assess trends, my study contributes to a limited body of research representing females. Male whip-poor-will exhibited substantial home range overlap with neighbouring males, but little to no overlap at the core range, with density estimates that suggest the population may still be growing instead of steady state. Whip-poor-will home and core ranges were associated with higher proportions of restored open habitat than expected, based on availability. To examine the most refined scale of habitat selection (fourth order) in whip-poor-will, I conducted vegetation surveys to characterize microhabitat of nest sites. Vegetation characteristics at nests were compared with random available sites, and data was collected at three spatial scales: 10 m scale (nest plot plus two random 10 m plots, 30 m from the nest), 1 m random scale (1 m nest plot plus 1 m plot in two random 10 m plots), and 1 m nest scale (1 m nest plot and random 1 m plot 2.5 m away from the nest). Whip-poor-will selected for higher canopy cover at the 10 m scale, and more leaf litter and less grassy substrate at the 1 m random scale, with no selection preferences for the 1 m nest scale. This pattern of preference suggests that whip-poor-will may initially target areas with high canopy cover, then select nest sites based on suitable leaf litter of an area within approximately 2.5 m. Deciduous leaf litter allows for camouflage during incubation and may provide thermal buffering. Leaf litter was mainly comprised of oak leaves, a species that retains senesced leaves into the late winter and early spring (marcescence). Oak leaves also contain high lignin content, which increases rigidity and slows decomposition rates. In combination, marcescence and high lignin result in deciduous leaf litter that persists for longer on the landscape and remains present upon whip-poor-will return to breeding grounds. By characterizing habitat selection across multiple scales, these findings provide ecological insight for understanding how whip-poor-will use restored habitats. As wildlife populations decline, research that informs effective conservation strategies and guides habitat restoration is increasingly important.