Agricultural expansion (both in the form of cropland and pastureland) is a persistent pressure on Amazonia’s tropical forest resources, with that land-use change having associated environmental impacts on one of the world’s most biodiverse and carbon-rich tracts of tropical forest. As population size and incomes increase globally, demand for calories and protein will increase as well. Amazonian agricultural systems can meet that increased demand via intensification of current systems, agricultural extensification, or – most realistically – some interrelated combination of the two. This study aims to determine how the ratio between intensification and extensification of key agricultural commodity systems will impact the delivery of a suite of ecosystem services. Our analysis uses a spatially-explicit model that takes climate and management into account to quantify the potential agricultural yield of several of the Amazon’s most important agricultural commodities driving land-use change – sugarcane, soybeans, maize and oil palm. We use different relative intensification/extensification scenarios to first satisfy a future calorie demand and subsequently assess changes in carbon storage, landscape ecology metrics, fertilizer addition and irrigation intensity across the Amazon basin.
Results/Conclusions
Preliminary results indicate that soybean and sugarcane yields are more controlled by climate and less by management, which in turn indicates that intensification strategies will have relatively limited influence on soybean and sugarcane yields over the extent of the basin. In contrast, oil palm and maize yields are relatively more sensitive to intensive management irrigation and fertilizer strategies. In terms of trade offs with ecosystem services, of the four key commodities, oil palm has the highest potential yields (in agricultural systems modeled for high-intensity management) in areas of the basin that have high carbon storage, including the relatively wet northwestern portion of the Amazon. Where intensification alone cannot meet a given simulation’s future demand, we use an extensification model based on an agricultural “suitability” datalayer. That extensification model contributes more to landscape patchiness metrics than meeting commodity demands via intensification strategies alone, indicating that aside from a gross loss of habitat, extensification could have implications for the landscape ecology of the Amazon. Finally, in our continued analysis, we hope to determine which ecoregion classes (based on ecoregion classification information in combination with abiotic metrics such as soil characteristics and precipitation regime) have been altered the most by agriculture in the Amazon and which classes stand to be altered the most in the future under different extensification scenarios.