In Conservation biology : the journal of the Society for Conservation Biology
Protected areas are a commonly-used strategy to confront forest conversion and biodiversity loss. While determining drivers of forest loss is central to conservation success, our understanding has been limited by conventional modeling assumptions. Here, we use random forest regression to evaluate a range of potential drivers of deforestation in protected areas in Mexico, while accounting for nonlinear relationships and higher-order interactions underlying deforestation processes. We find socioeconomic drivers (e.g., road density, human population density) and underlying biophysical conditions (e.g., precipitation, distance to water, elevation, slope) are stronger predictors of forest loss than protected area characteristics, such as age, type, and management effectiveness. Within protected area characteristics, we find variables reflecting collaborative and equitable management and protected area size to be the strongest predictors of forest loss, albeit with less explanatory power than socioeconomic and biophysical variables. Importantly, given that prior methods have typically assumed linear relationships, we show that the associations between most predictors and forest loss are nonlinear. Our findings can help inform decisions on the allocation of protected area resources by strengthening management in protected areas with the highest risk of deforestation and help preemptively protect key biodiversity areas that may be vulnerable to deforestation in the future. This article is protected by copyright. All rights reserved.
Powlen Kathryn A, Salerno Jonathan, Jones Kelly W, Gavin Michael C
2023-Jan-20
Mexico, conservation, deforestation, machine learning, random forest
