ABSTRACT Bénoué National Park in Northern Cameroon represents one of the last remaining strongholds for large savannah mammals in the Sudano Sahelian zone of West‐Central Africa, yet intensifying anthropogenic pressures threaten its ecological integrity. Despite its conservation importance, spatial drivers of mammal assemblages in the park remain poorly quantified. We conducted dry‐season guided reconnaissance surveys from 20 systematically distributed sampling points covering 43% of the park. A total of 528 detections (58% direct sightings, 42% indirect signs) were recorded, representing 20 medium and large size mammal species across four taxonomic orders. Species diversity, calculated using presence‐based metrics, was highest in peripheral zones where habitat heterogeneity was greater but where human activities were also more prevalent. Using variation partitioning based on principal coordinates of neighbour matrices (PCNM), we quantified the relative contributions of environmental and spatial predictors to mammal community structure. Habitat characteristics explained 64.4% of variation, spatial predictors 17.9% and their shared fraction 17.1%, demonstrating a strong interaction between environmental gradients and spatial configuration. Vegetation structure, NDVI and distance from settlements emerged as key determinants of species distribution. Vulnerable species, including Giant eland ( Taurotragus derbianus ) and Giraffe ( Giraffa camelopardalis ), were predominantly associated with structurally complex habitats distant from human settlements, whereas several widespread species occurred more frequently in open and human‐influenced areas. Our findings reveal a spatial overlap between biodiversity concentrations and anthropogenic pressure in park peripheries, highlighting the urgent need for targeted protection of structurally complex habitats and reinforced law enforcement in edge zones. These results provide an empirical basis for adaptive, spatially explicit conservation planning in Sudano–Sahelian protected areas.
Majiteu et al. (2026) studied this question.