Agricultural change in the Pastaza River Basin: A spatially explicit model of native Amazonian cultivation

We use a spatial perspective to analyze traditional land use systems in Western Amazonia, with emphasis on agriculture. This research examines the linkages between the social and spatial contexts of indigenous cultivators and agricultural change through the integration of geographic information systems, remote sensing, and logistic multiple regression. Results show that the probability of the presence of agriculture increases with population pressure in areas closer to existing cultivation fields. Areas that are close to service infrastructure and existing residential zones are less likely to be used for agriculture since these areas are recuperating as fallows. The model shows that the presence of agriculture is not affected by soil quality; however, extensive cultivation currently occurs in relatively steep slopes with poor soils. These findings suggest that there may be an overall scarcity of land resources for food production. Further research should involve time series data to model land use trajectories and determine when landscape changes could be expected.