Hammer-headed bat movement behavior and habitat selection

Abstract

Background: Animals with key ecological roles, such as seed-dispersing fruit bats, rely to varying degrees on habitat structure to indicate the locations of rewards and risks. Methods: To understand how variation in vegetation structure influences fruit bat selection, we related movement steps of hammer-headed bats (Hypsignathus monstrosus) to attributes of canopy height, vertical and horizontal structure, and habitat type in a mature rainforest of southern Cameroon. Vegetation structural metrics were measured with UAV-LiDAR at 10 m resolution for a 25 km2 study area. Because bats frequently moved outside the study area, we also characterized vegetation height and horizontal complexity over the full extent of bat movement trajectories by upscaling UAV-LiDAR measurements using spaceborne remote sensing, including GEDI LiDAR data. Results: At the site level, hammer-headed bats preferred areas of intermediate canopy height close to large canopy gaps (≥500 m). Individual bats varied in selection for vertical vegetation complexity, distance to smaller canopy gaps (≥50 m), and plant volume density of intermediate vegetation strata (10-20 m). Over the full extent of movement trajectories, hammer-headed bats consistently preferred intermediate canopy height, and areas closer to canopy gaps. At both spatial extents, bats moved the shortest distances in swamp habitats dominated by Raphia palms. These behaviors indicate use of forest types that vary structurally, with a preference for open airspace during foraging or moving among resources, and for dense swamp vegetation during roosting and foraging periods. In addition, most bats regularly made long flights of up to 17.7 km shortly after sunset and before sunrise and limited their movements to three or fewer destinations throughout the tracking period. Conclusions: These results highlight the importance of integrating remote sensing and animal tracking data to understand habitat selection of a species whose movements influence viral spillover risk and spatial patterns of seed dispersal.