Biofuel expansion would send cattle into the rain forest
Biofuel production in the US has met with fairly mixed success, as the cost and fossil fuel use of corn-based ethanol has severely cut into the benefits provided by avoiding the use of fossil fuels. It's been a somewhat different story in Brazil, which has embraced ethanol derived from sugarcane and seen more promising results. The government has set aggressive targets for both ethanol and biodiesel production, but a study that will be published in the Proceedings of the National Academies of Science later this week urges caution: unless the goals are met through an integrated agricultural strategy, they'll drive deforestation that will offset most of the benefits.
The study looked at the expansion of the two crops that are expected to drive biofuels growth in Brazil: sugarcane for ethanol, and soy beans for biodiesel. To reach the country's 2020 goals, there will have to be a major increase in the production of both of those crops. Even assuming major increases in the efficiency of their production (the authors assume an increase at double the rate of the past 20 years), there's simply no way to get there without expanding the amount of land devoted to farming them, and there's no way to do that without secondary consequences.
To examine the consequences, the authors built a linked model that incorporated Brazilian land use, an economic model of its agricultural economy, and a global agricultural and vegetation productivity model. The model allowed them to determine both the direct land use changes caused by the focus on sugarcane and soy, as well as the indirect changes caused by the accompanying alterations in the Brazilian agricultural system. They were also able to estimate a carbon payback time, the number of years that it would take before the carbon footprint of the land use changes were equal to the reduced carbon emissions from the transportation sector.
To reach Brazil's 2020 targets, both crops would have to be radically expanded. Sugarcane would need an extra 57,000 square kilometers devoted to growing it, while soybeans would need nearly 110,000 new square kilometers. But those radical changes wouldn't require new land; projecting the trends of recent years into the future, almost 90 percent of this expanded cropland would be placed on areas currently used for cattle. Converting the rangeland to crops would have a minimal impact on the carbon sequestered there, so sugarcane would have a payback time of only four years. Soy, which can't be converted to biofuels as efficiently, would require 35 years. So far, not so bad.
Unfortunately, the cattle that were previously on that rangeland wouldn't simply disappear. Instead, the authors project that over 160,000 square kilometers of native habitat (over 120,000 of that forest) would be converted into rangeland. That process would liberate lots of carbon currently sequestered by those habitats. As a result, with these indirect land use changes incorporated, the payback times recede into the future. For sugarcane, the figure would be 44 years, while for soybeans it would be nearly 250 years. There seems to be a reasonable chance that we won't be relying on internal combustion engines in 250 years.
Given how problematic soybeans are, the authors consider alternative crops, and find that the oil palm provides some significant improvements to those numbers. Although the use of oil palms is associated with deforestation in many countries, it's possible to convert rangelands to oil palm plantations with a minimal impact, and the the high efficiency of these plants means that relatively little land would be required: just 4,200 square kilometers, only about four percent of the land needed for soybeans. The payback time for direct land use changes would only be seven years.
The researchers suggest that it might be possible to head off land use changes entirely, however. They argue that certain aspects of Brazilian agricultural law promote ranching at a low cattle population density, and it may be possible to both increase that density and manage the rangeland in a way that limits the carbon emissions. Under the oil palm scenario, the density of cattle would have to be increased by only 0.1 heads per hectare in order to eliminate indirect land use changes.
There are a lot of assumptions built in to this analysis, such as increased agricultural efficiency and continued growth of local agriculture, so the specifics should be interpreted cautiously. Nevertheless, the general outlines of the findings highlight important issues: it's easy for land-use changes to swamp the carbon impact of biofuel production, and any planning for biofuel production needs to consider them as part of the larger agricultural economy. Those are lessons that have been appearing pretty consistently in studies that consider a variety of crops and countries.
PNAS, 2010. DOI: 10.1073/pnas.0907318107