Abstract
Vertical stratification has long been recognized as a key dimension of biodiversity in structurally complex ecosystems, shaping animal movement and community structure. Camera traps provide a powerful means to extend biodiversity monitoring across forest strata through continuous, standardized observation. However, most camera trapping studies focus on terrestrial observations or a single arboreal layer, limiting inference about vertically structured communities. Here, we evaluate a standardized “camera column” approach for assessing mammal community occupancy in the Congo Basin. We deployed camera traps in three forest strata (canopy, understory‐midstory, and ground) at each sampling point in a standardized grid and used a multispecies, multi‐scale occupancy model to assess how incorporating observations across strata influences estimates of species occupancy. Our results demonstrate that incorporating vertical space can alter the inferred relationship between mammal communities and environmental gradients, with a significant positive effect of elevation on mean occupancy emerging only when observations from all forest strata were incorporated. These results suggest that even small elevation gradients in lowland tropical forests can shape mammal diversity, likely through soil‐mediated effects on habitat structure and resource availability. Furthermore, accounting for three‐dimensional habitat structure may be essential for accurately characterizing community–environment relationships in vertically structured systems.