Global Biofuel Expansion and the Demand for Brazilian Land: Intensification versus Expansion

We use a spatially disaggregated model of Brazilian agriculture to assess the implications of global biofuel expansion on Brazilian land usage at the regional level. This Brazilian model is part of the FAPRI agricultural modeling system, a multimarket, multi-commodity international agricultural model, used to quantify the emergence of biofuels and to analyze the impact of biofuel expansion and policies on both Brazilian and world agriculture. We evaluate two scenarios in which we introduce a 25% exogenous increase in the global demand for ethanol and one scenario in which we increase global ethanol demand by 50%. We then analyze the impact of these increases in terms of land-use change and commodity price changes particularly in Brazil. In the first scenario, we assume that the enforcement of the land-use reserve in Brazil remains at historically observed levels, and that abundant additional land can be readily incorporated into production. The second scenario involves implementing the same exogenous biofuel demand shock but with a different responsiveness in area expansion to price signals in Brazil, reflecting varying plausible assumptions on land availability for agricultural expansion. The third scenario, which is similar to the first scenario but with a larger increase in global ethanol demand, is run to check whether increasing volume of ethanol requires the incorporation of additional quantities of land per unit of ethanol. We find that, within Brazil, the expansion occurs mostly in the Southeast region. Additionally, total sugarcane area expansion in Brazil is higher than the increase in overall area used for agriculture. This implies that part of the sugarcane expansion displaced other crops and pasture that is not replaced, which suggests some intensification in land use. The lower land expansion elasticities in the second scenario result in a smaller expansion of area used for agricultural activities. A higher proportion of the expansion in sugarcane area occurs at the expense of pasture area, which implied land intensification of beef production. This explains the small change in commodity prices observed between the first and second scenarios. These results suggest that reducing the overall responsiveness of Brazilian agriculture may limit the land-use changes brought about by biofuel expansion, which would in turn reduce its environmental impacts in terms of land expansion. Additionally, the impacts on food prices are limited because of the ability of local producers to increase the intensity of land use in both crop (by double cropping and raising yields) and livestock production (by increasing the number of heads of cattle per hectare of pasture or stocking rate) releases area that can be used for crops. In scenario three, we find that larger ethanol volumes did not require more land per unit of ethanol. Doubling the demand for ethanol does not change the results, which indicates that the limit for intensification is beyond the 50% expansion assumed in Scenario 3. In this range, the same amount of land is incorporated into production per additional unit of ethanol.