The increase in soybean production as a source of protein and oil is being stimulated by the growing demand for livestock feed, food and numerous other applications. Significant greenhouse gas (GHG) emissions can result from land use change due to the expansion and cultivation of soybean. However, this is complex to assess and the results can vary widely. The main goal of this article is to investigate the life-cycle GHG balance for soybean produced in Latin America, assessing the implications of direct land use change emissions and different cultivation systems. A life-cycle model, including inventories for soybean produced in three different climate regions, was developed, addressing land use change, cultivation and transport to Europe. A comprehensive evaluation of alternative land use change scenarios (conversion of tropical forest, forest plantations, perennial crop plantations, savannah and grasslands), cultivation (tillage, reduced tillage and no-tillage) and soybean transportation systems was undertaken. The main results show the importance of land use change in soybean GHG emissions, but significant differences were observed for the alternative scenarios, namely 0.1-17.8 kg CO(2)eq kg(-1) soybean. The original land choice is a critical issue in ensuring the lowest soybean GHG balance and degraded grassland should preferably be used for soybean cultivation. The highest GHG emissions were calculated for tropical moist regions when rainforest is converted into soybean plantations (tillage system). When land use change is not considered, the GHG intensity varies from 0.3 to 0.6 kg CO(2)eq kg(-1) soybean. It was calculated that all tillage systems have higher GHG emissions than the corresponding no-tillage and reduced tillage systems. The results also show that N2O emissions play a major role in the GHG emissions from cultivation, although N2O emission calculations are very sensitive to the parameters and emission factors adopted.
