Road network and deforestation of indigenous lands in the Brazilian Amazon

• Published in: Transportation research. Part D, Transport and environment Vol. 119; p. 103735
• Main Authors: da Silva, Carlos Fabricio Assunção, de Andrade, Mauricio Oliveira, dos Santos, Alex Mota, de Melo, Silas Nogueira
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-06-2023
• Subjects: Deforestation; Gaussian Mixture Model; Indigenous Lands; Unofficial Roads; WLS Regression; Deforestation; Indigenous Lands; Gaussian Mixture Model; WLS Regression; Unofficial Roads
• ISSN: 1361-9209; 1879-2340
• DOI: 10.1016/j.trd.2023.103735

This research evaluates the relationship between the official and unofficial road network and deforestation in the Brazilian Amazon within and around 332 indigenous lands from 2008 to 2021. The study uses infrastructural and environmental data. Two analyses were performed: clustering by the Gaussian Mixture Model to find group similarities and regressions by weighted least squares (WLS) and the generalized linear models (GLM) to extract factors related to deforestation considering the complete set and the clusters. The results revealed six clusters. Some suffer direct impacts from roads, while others are better preserved. Regression models revealed active fires, illegal mining, and that unofficial roads within ILs to be the drivers for deforestation. The general results indicated that for every 1 km of an additional unofficial route, deforestation increases by 0.036 km2. However, when analyzing the regression models for clusters, we concluded that only 45% of the lands are currently affected by road infrastructure.