Fig. 1

Movement trajectories of all bats and a depiction of the two spatial extents of habitat selection analyses. (A) Attributes of 3D vegetation structure measured at 10 m resolution were limited to the 25 km² site level within the UAV-LiDAR extent of Bouamir Research Site. Bat movement tracks are overlain on a map of canopy height (black = 0 m, white = 55 m). Note that canopy height (height of first LiDAR Return) was included in all models as a quadratic term (canopy height + canopy height²). Vertical complexity: total diversity of 3D point cloud distribution measured from ground to top-of-canopy; Distance to gap: straight-line distance to nearest area with no vegetation > 5 m; Plant Volume Density: leaf area per volume within a specified height bin (10–15 or 15–20 m). Swamp: habitat characterized by seasonal or permanent shallow water and characterized by dominance of Raphia palm species. (B) We used upscaled 3D vegetation structure metrics to quantify habitat selection at the landscape level, which encompassed the full scale of bat movement tracks, including the boundary of the Dja Faunal Reserve. Canopy height: predicted value of 95th percentile relative height (RH 95). Distance to gap: straight-line distance to nearest area with no vegetation > 15 m. Canopy heterogeneity: standard deviation of canopy height at a specified spatial resolution (100–1000 m). Swamp: same as in panel A. (C) The inset photo shows a male hammer-headed bat carrying a 15 g e-obs tag