Spatial modelling of deforestation-related factors in the Brazilian semi-arid biome
Carlos Fabrício Assunção da Silva, Alex Mota dos Santos, Silas Nogueira de Melo, Anderson Paulo Rudke, Pedro Monteiro de Almeida
Abstract
The Caatinga biome is the largest and wettest semi-arid environment globally, occupying 10% of the Brazilian territory. Half of the original vegetation from the Caatinga has already been deforested. This research aims to identify the factors associated with Caatinga deforestation (2010–2016). Therefore, the researchers used Geographically Weighted Regression to analyse the drivers of deforestation at the municipal level. Our results showed that five factors are crucial for understanding the spatial variability of deforestation in the Caatinga biome: agricultural credit, cattle, goats, distance from roads, and mining activities. The results were interpreted carefully since the semi-arid Biome is marked by regional asymmetries, mixing subsistence agriculture with agribusiness (MATOPIBA and Vale do Rio São Francisco). The outcome maps allow us to discuss the implication of spatial arrangements and intervention policies towards sustainability.