Abstract Land use change alters forest ecosystems by reducing habitat amount and often increasing fragmentation, but the relative importance of these drivers for community dynamics remains debated. It is also unclear whether functional traits consistently predict species' responses to forest change, particularly in tropical dry forests. Using 6 years of camera‐trap monitoring and remotely sensed forest metrics from the Chiquitano Dry Forest in Bolivia, we assessed how forest cover (habitat amount) and edge density (fragmentation) interact to shape mammal occupancy, and whether traits (body mass, home range size, activity cycle, diet breadth, and habitat breadth) mediate species' responses to forest change. Community occupancy tended to increase with forest cover, but this association weakened in areas with high edge density, indicating that fragmentation reduces the ecological benefits of greater habitat amount for mammal communities. Edge density added explanatory power for community‐level and several species' responses through its interaction with forest cover but showed little evidence for an independent effect. Forest cover showed strong positive associations with occupancy for several species, but the strength of individual species' responses was variable. Wide‐ranging, large‐bodied mammals consistently showed stronger dependence on high forest cover, while activity cycle, diet, and habitat breadth were poor predictors of species' responses. None of the examined traits was consistently associated with responses to edge density. Although trait models explained only modest variance in species' responses (~15%), our findings suggest that wide‐ranging, large‐bodied mammals may serve as useful monitoring targets in areas experiencing deforestation. At the same time, the interaction between habitat amount and fragmentation was most clearly supported at the community level, indicating that integrating species‐ and community‐level perspectives can provide a more complete basis for the conservation of intact mammal communities in human‐modified tropical forests.
Weiß et al. (Fri,) studied this question.