Fragmented Forests with Variable Human Disturbance Shaping Habitat Utilization and Foraging Behaviour of Nocturnal Mammals
DOI:
https://doi.org/10.70102/AEJ.2026.18.1.24Keywords:
Forest fragmentation, Human disturbance, Nocturnal mammals, Habitat utilization, Foraging behavior, Wildlife conservation, Landscape ecology.Abstract
Forest fragmentation and a growing human presence have become important challenges for wildlife conservation, especially for mammal species that are mainly nocturnal, and depend greatly on stable habitat conditions and secure foraging environments. The project aims to assess how the use of forest habitat and foraging patterns of nocturnal mammals change with fragmentation of habitat in forests of varying anthropogenic disturbance levels. The study was conducted in forest patches that were subdivided into three levels of habitat degradation (low, moderate, and high), habitat proximity (distance from human settlements) and human activities (low, moderate, and high). Camera trapping, GPS tracking of movement, vegetation data and behavioral data were gathered. Results showed that there were significant differences in habitat use and behavioral response on disturbance gradients. Species richness was found to be 12.4 ± 1.3 in the former forests and 5.2 ± 0.9 in the latter forests, and the occupancy probability was 0.82 and 0.34, respectively. The use of corridors was also significantly lower (21% compared to 78% in intact habitats), representing limited movement of wildlife. The behavioral analysis showed that nocturnal mammals in highly disturbed forests were more vigilant (58%) and had shorter activity durations (4.5 hours/night) and fewer feeding events (6 events/night) than in less disturbed forests. The feeding success decreased from 76% to 49% between intact and fragmented forest, and dependency on anthropogenic food sources has greatly increased. The human disturbance level was related to changes in habitat utilization and foraging behavior when analyzed statistically (ANOVA and generalized linear models;