On paper, making biofuels from switchgrass and other perennials that need not be replanted seems like a no-brainer. Use the sun's energy to grow the crop, and then convert it to liquid fuels to power our cars without the need for gasoline. But so far, experiments with these "cellulosic" crop-based fuels have only been conducted on small scales, leaving open the question of how feasible the strategy is. Now, the first large-scale study shows that switchgrass yields more than five times the energy needed to grow, harvest, and transport the grass and convert it to ethanol. The results could propel efforts to sow millions of hectares of marginal farmland with biofuel crops.

Previous studies on switchgrass plots suggested that ethanol made from the plant would yield anywhere from 343% to 700% of the energy put into growing the crop and processing it into biofuel. But these studies were based on lab-scale plots of about 5 square meters. So 6 years ago, Kenneth Vogel, a geneticist with the U.S. Department of Agriculture in Lincoln, Nebraska, and colleagues set out to enlist farmers for a much larger evaluation. Farmers planted switchgrass on 10 farms, each of which was between 3 and 9 hectares. They then tracked the inputs they used--diesel for farm equipment and transporting the harvested grasses, for example--as well as the amount of grass they raised over a 5-year period. After crunching the numbers, Vogel and his colleagues found that ethanol produced from switchgrass yields 540% of the energy used to grow, harvest, and process it into ethanol. Equally important, the researchers found that the switchgrass is carbon neutral, as it absorbs essentially the same amount of greenhouse gases while it's growing as it emits when burned as fuel.

A final significant finding, Vogel says, is that yields on farms using fertilizer and other inputs, such as herbicides and diesel fuel for farm machinery, were as much as six times higher than yields on farms that used little or no fertilizer, herbicides, or other inputs to grow a mixture of native prairie grasses. That result contrasts sharply with a controversial study published just over a year ago in Science that suggested that a mixture of prairie grasses farmed with little fertilizer or other inputs would produce a higher net energy yield than ethanol produced from corn (Science, 8 December 2006, p. 1598). Instead, the current study shows that switchgrass farmed using conventional agricultural practices on less-than-prime cropland yields only slightly less ethanol per hectare on average than corn. "The bottom line is that low-input systems are not economically viable," Vogel says.

"This is a really important paper," says Christopher Somerville, who directs the Energy Biosciences Institute at the University of California, Berkeley. The impressive yield numbers, he adds, will likely serve as a baseline for future studies, because agricultural scientists are making rapid strides at creating new, higher-yielding switchgrass strains.