Abstract
Over the past three decades, large tracts of tropical forests have been
converted to crop and pasturelands across southern Amazonia, largely to
meet the increasing worldwide demand for protein. As the world’s
population continue to grow and consume more protein per capita, forest
conversion to grow more crops could be a potential solution to meet such
demand. However, widespread deforestation is expected to negatively
affect crop productivity via multiple pathways (e.g., thermal
regulation, rainfall, local moisture, pest control, among others). To
quantify how deforestation affects crop productivity, we modeled the
relationship between forest amount and enhanced vegetation index
(EVI—a proxy for crop productivity) during the soybean planting season
across southern Amazonia. Our hypothesis that forest amount causes
increased crop productivity received strong support. We found that the
maximum MODIS-based EVI in soybean fields increased as a function of
forest amount across three spatial-scales, 0.5 km, 1 km, 2 km, 5 km, 10
km, 15 km and 20 km. However, the strength of this relationship varied
across years and with precipitation, but only at the local scale (e.g.,
500 meters and 1 km radius). Our results highlight the importance of
considering forests to design sustainable landscapes.