Oil palm (Elaeis guineensis), a rapidly expanding crop in northeastern Pará, first emerged in the 1970s as a crucial response to the global oil crisis. However, its swift expansion has subsequently generated significant socio-environmental conflicts, profoundly altering local socioecological dynamics. Therefore, we aimed to investigate land-use and land-cover changes within the headwater sub-basin of the Caeté River, focusing specifically on the municipality of Bonito, Pará. To achieve this, we employed remote sensing and geospatial analysis to accurately delineate the study area and perform supervised classifications. Specifically, we used the Random Forest algorithm to map five distinct periods: 1985, 1995, 2004, 2015, and 2023. In addition, we calculate an Anthropogenic Transformation Index (ATI) in order to observe the human influence in the landscape. Our classification models exhibited high accuracy, with overall accuracy values ranging from 0.63 to 0.87 and Kappa coefficients between 0.53 and 0.76, demonstrating consistent discrimination among LULC classes. The results revealed a marked transformation of the landscape, with oil palm monocultures progressively expanding at the expense of dense forest and human-modified vegetation. For instance, the ATI increased from 3.14 in 1985 to 5.56 in 2004, followed by a slight decline to 4.90 in 2023, suggesting a potential stabilisation—but not a reversal—of anthropogenic pressures. Nonetheless, the negative socioecological impacts of the oil palm monocultures in this Amazonian landscape remain severe, encompassing issues such as water pollution and ongoing socio-environmental conflicts. In conclusion, this research highlights the importance of understanding these dynamics to support sustainable management of the Caeté River basin. Furthermore, we underscore the urgent need for further research to rigorously evaluate effective mitigation strategies and foster genuinely sustainable development within the region.
Corrêa et al. (Mon,) studied this question.