Brazil’s House of Representatives last month approved profound changes to the Forest Code, a series of laws to regulate the country’s land use. The changes would jeopardize ecosystems and allow deforestation of about 20 million more hectares than the original Forest Code, according to estimates by Gerd Sparovek of the University of São Paulo (see go.nature.com/n8lwfu). The Brazilian Senate must seize its last opportunity to block the changes.

The proposed changes would substantially increase carbon emissions. Assuming that half of the newly affected forest (10 million hectares) will be given over to agriculture and that the average carbon biomass in the Amazon, Atlantic Forest and Cerrado is 50 megagrams per hectare (L. F. Alves et al. Forest Ecol. Manage. 260, 679–691; 2010), emissions would amount to some 500 teragrams of carbon — on top of the 100–300 Tg that arise annually from deforestation of the Amazon (J. P. H. B. Ometto et al. Oecologia 143, 483–500; 2005).

The government’s efforts to decrease Amazon deforestation would be wasted, as would investments made in the ethanol industry for a cleaner environment. Brazil’s commitment to the environment could lose credibility internationally, causing developed countries to raise import taxes on Brazilian agricultural products, making them less competitive. 

Such serious consequences would offset any short-term gain by the Brazilian agricultural sectors that might result from changes to the Forest Code. 

A team of researchers from Stanford University, the Carnegie Institution for Science, and Arizona State University has found that converting large swaths of land to bioenergy crops could have a wide range of effects on regional climate.

In an effort to help wean itself off fossil fuels, the U.S. has mandated significant increases in renewable fuels, with more than one-third of the domestic corn harvest to be used for conversion to ethanol by 2018. But concerns about effects of corn ethanol on food prices and deforestation had led to research suggesting that ethanol be derived from perennial crops, like the giant grasses Miscanthus and switchgrass. Nearly all of this research, though, has focused on the effects of ethanol on carbon dioxide emissions, which drive global warming.

"Almost all of the work performed to date has focused on the carbon effects," said Matei Georgescu, a climate modeler working in ASU's Center for Environmental Fluid Dynamics. "We've tried to expand our perspective to look at a more complete picture.  What we've shown is that it's not all about greenhouse gases, and that modifying the landscape can be just as important."

Georgescu and his colleagues report their findings in the current issue (Feb. 28, 2011) of the Proceedings of the National Academy of Sciences (see Direct Climate Effects of Perennial Bioenergy Crops in the United States). Co-authors are David Lobell of Stanford University's Program on Food Security and the Environment and Christopher B. Field of the Carnegie Institution for Science, also located in Stanford, California.

In their study, the researchers simulated an entire growing season with a state-of-the-art regional climate model. They ran two sets of experiments - one with an annual crop representation over the central U.S. and one with an extended growing season to represent perennial grasses. In the model, the perennial plants pumped more water from the soil to the atmosphere, leading to large local cooling. 

"We've shown that planting perennial bioenergy crops can lower surface temperatures by about a degree Celsius locally, averaged over the entire growing season. That's a pretty big effect, enough to dominate any effects of carbon savings on the regional climate." said Lobell.

The primary physical process at work is based on greater evapotranspiration (combination of evaporated water from the soil surface and plant canopy and transpired water from within the soil) for perennial crops compared to annual crops. 

"More study is needed to understand the long-term implication for regional water balance." Georgescu said. "This study focused on temperature, but the more general point is that simply assessing the impacts on carbon and greenhouse gases overlooks important features that we cannot ignore if we want a bioenergy path that is sustainable over the long haul."

We find your conclusion premature that there is no longer a direct correlation between food production in Brazil and deforestation in the Amazon (Nature 466, 554-556; 2010).

An increase in demand by international markets for export commodities such as soya beans and beef will mean more rainforest clearance. There is still potential for a huge increase in productivity, given that large producers of export goods are encouraged by government loans at favourable rates and fiscally exempt debt relief, which in turn attract investment in research and development.

Moreover, Brazil's Congress has proposed large structural changes to the Forest Code that could lead to further deforestation and threaten the preservation of the most important Brazilian biomes.

Brazil's own staple crops - rice, beans and cassava - account for very little deforestation. The small farmers producing these still suffer low credit and heavy debts, fragile land tenure, scant investment in crop research, and inferior storage conditions for their products.

A global farm should be socially fair as well as environmentally friendly. Although Brazilian agricultural policy is on the way to meeting these conditions, we are not yet there.

Global agricultural expansion cut a wide swath through tropical forests during the 1980s and 1990s. More than half a million square miles of new farmland - an area roughly the size of Alaska - was created in the developing world between 1980 and 2000, of which over 80 percent was carved out of tropical forests, according to Stanford researcher Holly Gibbs.

"This has huge implications for global warming, if we continue to expand our farmland into tropical forests at that rate," said Gibbs, a postdoctoral researcher in the Department of Environmental Earth System Science and in the Program on Food Security and the Environment, who led the study.

Gibbs and colleagues at several other universities analyzed Landsat satellite data and images from the United Nations to reach their conclusions. Theirs is the first study to map and quantify what types of land have been replaced by the immense area of new farmland developed across the tropical forest belt during the 1980s and 1990s.

While this huge increase was happening within the tropics, agricultural land in the non-tropical countries actually decreased in area.

The study was published this week in the online Early Edition of the Proceedings of the National Academy of Sciences.

The United Nations Food and Agriculture Organization estimates that to keep pace with increasing demand, global agricultural production will have to keep increasing, possibly even doubling by 2050. That would likely lead to millions of additional acres of tropical forest being felled over the next 40 years.

Direct impact on carbon released into atmosphere

"Every million acres of forest that is cut releases the same amount of carbon into the atmosphere as 40 million cars do in a year," Gibbs said.

Most of the carbon released comes from burning the forests, but even if the trees are simply cast aside, the bulk of the carbon from the plants makes its way into the atmosphere during decomposition, she said.

Gibbs and her colleagues found that about 55 percent of the tropical forests that had been cut between 1980 and 2000 were intact forests and another 28 percent were forests that had experienced some degradation, such as some small-scale farming, logging or gathering of wood and brush for cooking or heating fuel.

"The tropical forests store more than 340 billion tons of carbon, which is 40 times the total current worldwide annual fossil fuel emissions," Gibbs said. "If we continue cutting down these forests, there is a huge potential to further contribute to climate change."

The increasing demand for agricultural production stems in part from the ever-growing number of people on the planet, who all want to eat. Additionally, members of the growing middle class in emerging economies such as China and India are showing interest in eating more meat, which further intensifies demand. And incentives to grow crops for biofuel production have increased.

But Gibbs and her colleagues also observed some encouraging signs. The patterns of change in the locations they analyzed made it clear that during the 1990s, less of the deforestation was done by small family farms than was the case in the 1980s and more was done by large, corporate-run farms. Big agribusiness tends to be more responsive to global economic signals as well as pressure campaigns from advocacy organizations and consumer groups than individual small farmers.

In Brazil, where a pattern had developed of expanding soy production by direct forest clearing and by pushing cattle ranching off pastureland and into forested areas, a campaign by Greenpeace and others resulted in agreements by key companies to rein in their expansion. Instead, they worked to increase production on land already in agricultural use.

'Seeing positive changes'

"These farmers effectively increased the yield of soy on existing lands and they have also increased the head of cattle per acre by a factor of five or six," Gibbs said. "It is exciting that we are starting to see how responsive industry can be to consumer demands. We really are seeing positive changes in this area."

Along with wiser use of land already cleared, Gibbs said, improvements in technology and advances in yield intensification also could slow the expansion of farming into the forests.

Other studies that analyzed land use changes between 2000 and 2007 have shown that the pace of cutting down the tropical forests has begun to slow in some regions.

But as long as the human population on the planet continues to grow, the pressure to put food on the table, feed in the barnyard and fuel in the gas tank will continue to grow, too.

"It is critical that we focus our efforts on reducing rates of deforestation while at the same time restoring degraded lands and improving land management across the tropics," Gibbs said. "The good news is that pressure from consumer groups and nongovernmental organizations combined with international climate agreements could provide a real opportunity to shift the tide in favor of forest conservation rather than farmland expansion."

In addition to her position at the Department of Environmental Earth System Science and the Program on Food Security and the Environment, Gibbs is affiliated with Stanford's Woods Institute for the Environment. Jon Foley, a professor of ecology, evolution and behavior, and director of the Institute on the Environment at the University of Minnesota, was Gibbs' PhD adviser when the research was begun. He is a coauthor of the paper.

Initial funding for the project was provided by NASA. Gibbs is currently funded by a David H. Smith Conservation Research Fellowship.