Land use change models that integrate quantitative and qualitative approaches better explain deforestation patterns in Amazonian protected areas

Agricultural frontiers in the Amazon Basin – many of which overlap with protected areas – experience rapid forest conversion to agriculture and pasture, threatening ecological health and globally significant ecosystem services. Effective responses to protected area deforestation require understanding the socio-environmental factors that increase the likelihood of forest conversion, which may be common region-wide or specific to each protected area. Crucially, standard, quantitative approaches to land use change modeling may not include some of these factors, constraining our understanding of and response to deforestation. Dominant discourses about deforestation – promoted by government actors and conservation organizations – also shape responses to deforestation. We integrate quantitative and qualitative analysis of deforestation dynamics into land use change models of three protected area complexes in the Amazon to understand region-wide and site-level factors related to deforestation and the ability of conservation discourses to explain deforestation patterns. Our integrative methodology yielded better model performance than standard land use change modeling for all sites. From 2008-2018, forests on steeper slopes with lower population densities were less likely to experience deforestation, while forests surrounded by non-forest pixels and located closer to agriculture and fire activity were more likely to convert. Finally, while dominant discourses sometimes aligned with the results of integrated models (e.g., fires were associated with increased deforestation probability in all sites), our models did not support some factors commonly cited in deforestation discourses (e.g., REDD+ concessions had no relationship with deforestation in Peru’s Tambopata National Reserve and Bahuaja-Sonene National Park). Our results can inform management responses to stem deforestation (in our study sites and across the Amazon) and emphasize the need for a balanced, integrative approach to operationalizing dominant discourses in conservation science and practice, as the framing of deforestation – through quantitative and qualitative approaches – shapes understandings of and responses to deforestation.