Multi-strata "camera columns": an effective approach to characterize non-volant mammal communities in tropical forests.

Understanding how arboreal mammals partition vertical space is essential to accurately characterize tropical forest communities, yet most camera trap surveys focus on the forest floor, or pair canopy and ground surveys. While canopy camera trapping is revealing new insights about arboreal mammals, sampling at a single height at each sampling point may obscure patterns in community structure. To address this gap, we developed a "camera column" technique, deploying 3 vertically aligned cameras at ground, understory-midstory, and canopy levels across 3 survey areas in Lomami National Park and its buffer zone, Democratic Republic of the Congo. We used this method to evaluate the species richness, community composition, and vertical habitat use of non-volant mammals. We identified 47 mammal taxa, including 8 outside their documented geographic range. Species accumulation curves showed that pooling detections across strata increased estimated richness relative to single-stratum sampling. However, pairwise pooling of strata to produce rarefaction curves revealed that combining ground and canopy cameras can sufficiently estimate species richness, suggesting redundancy in understory detections. Ordination analyses and height-distribution data revealed strong vertical structuring of mammal communities, with understory detections overlapping both terrestrial and canopy assemblages and capturing transitional habitat use. Notably, one-quarter of the arboreal taxa favored the understory-midstory level, indicating that reduced-strata surveys may mischaracterize vertical habitat use and bias analyses informed by presence/absence data. Our results demonstrate that while ground-canopy camera combinations may approximate species richness, full vertical sampling is necessary to accurately interpret community structure and ecological roles of species. Multi-strata camera columns provide a scalable approach for biodiversity assessment and conservation monitoring in structurally complex tropical forests.