Large-scale expansion of agriculture in Amazonia may be a no-win scenario

Using simplified climate and land-use models, we evaluated primary forests' carbon storage and soybean and pasture productivity in the Brazilian Legal Amazon under several scenarios of deforestation and increased CO2. The four scenarios for the year 2050 that we analyzed consider (1) radiative effects of increased CO2, (2) radiative and physiological effects of increased CO2, (3) effects of land-use changes on the regional climate and (4) radiative and physiological effects of increased CO2 plus land-use climate feedbacks. Under current conditions, means for aboveground forest live biomass (AGB), soybean yield and pasture yield are 179 Mg-C ha−1, 2.7 Mg-grains ha−1 and 16.2 Mg-dry mass ha−1 yr−1, respectively. Our results indicate that expansion of agriculture in Amazonia may be a no-win scenario: in addition to reductions in carbon storage due to deforestation, total agriculture output may either increase much less than proportionally to the potential expansion in agricultural area, or even decrease, as a consequence of climate feedbacks from changes in land use. These climate feedbacks, usually ignored in previous studies, impose a reduction in precipitation that would lead agriculture expansion in Amazonia to become self-defeating: the more agriculture expands, the less productive it becomes. (letter)