Biodiversity loss in deforestation frontiers: Linking occupancy modelling and physiological stress indicators to understand local extinctions
Semper Pascual, Asunción; Decarre, Julieta; Baumann, Matthias; Busso, Juan Manuel; Camino, Micaela; et al.; Biodiversity loss in deforestation frontiers: Linking occupancy modelling and physiological stress indicators to understand local extinctions; Elsevier; Biological Conservation; 236; 8-2019; 281-288
Semper Pascual, Asunción
Busso, Juan Manuel
Gomez Valencia, Bibiana
Tropical deforestation is a main driver of the global biodiversity crisis. Impact assessments typically focus on species' presence, which means impacts are detected when local extinctions have occurred – and thus when it is too late. Here, we pioneer the combined use of two approaches that can detect deforestation impacts earlier, at the level of populations (using occupancy modelling) and at the level of individuals (using stress hormonal indicators). We tested this approach for the collared peccary (Pecari tajacu) in the Argentine Chaco, a global deforestation hotspot. We used camera-trap data to model peccary occupancy in relation to woodland cover and loss, and measured glucocorticoid metabolites in peccary feces to assess individuals' stress level in deforestation areas. We found that peccary occupancy was highest in remote areas with high woodland cover, but low otherwise. Peccaries were typically absent from areas where deforestation had been widespread recently. Where peccaries were present, physiological stress was correlated with the extent of edge between cropland and forest (a proxy for food availability), and not with deforestation. This, and the observation that peccaries disappear quickly as deforestation progresses, suggests that peccaries do not adapt well to the new conditions in deforestation frontiers. In terms of conservation management, our results underpin the importance of protecting large, contiguous woodland blocks to prevent large mammals from going extinct in deforestation frontiers. More broadly, we show how combining stress hormonal indicators and occupancy modelling can provide deep insights into processes underlying local extinctions in dynamic landscapes.