BEFORE AGRICULTURE, MOST OF THE PLANET WAS COVERED WITH PLANTS THAT LIVED YEAR after year. These perennials were gradually replaced by food crops that have to be replanted every year. Now scientists are contemplating reversing this shift by creating perennial versions of familiar crops such as corn and wheat. If they are successful, yields on farmland in some of the world's most desperately poor places could soar. The plants might also soak up some of the excess carbon in the earth's atmosphere.

Agricultural scientists have dreamed of replacing annuals with equivalent perennials for decades, but the genetic technology needed to make it happen has appeared only in the past 10 or 15 years, says agroecologist Jerry Glover. Perennials have numerous advantages over crops that must be replanted every year: their deep roots prevent erosion, which helps soil hold onto critical minerals such as phosphorus, and they require less fertilizer and water than annuals do. Whereas conventionally grown monocrops are a source of atmospheric carbon, land planted with perennials does not require tilling, turning it into a carbon sink.

Farmers in Malawi are already getting radically higher yields by planting rows of perennial pigeon peas between rows of their usual staple, corn. The peas are a much needed source of protein for subsistence farmers, but the legumes also increase soil water retention and double soil carbon and nitrogen content without reducing the yield of the primary crop on a given plot of land.

Taking perennials to the next level—adopting them on the scale of conventional crops—will require a significant scientific effort, however. Ed Buckler, a plant geneticist at Cornell University who plans to develop a perennial version of corn, thinks it will take five years to identify the genes responsible for the trait and another decade to breed a viable strain. “Even using the highest-technology approaches available, you're talking almost certainly 20 years from now for perennial maize,” Glover says.

Scientists have been accelerating the development of perennials by using advanced genotyping technology. They can now quickly analyze the genomes of plants with desirable traits to search for associations between genes and those traits. When a first generation of plants produces seeds, researchers sequence young plants directly to find the handful out of thousands that retain those traits (rather than waiting for them to grow to adulthood, which can take years).

Once perennial alternatives to annual crops are available, rolling them out could have a big impact on carbon emissions. The key is their root systems, which would sequester, in each cubic meter of topsoil, an amount of carbon equivalent to 1 percent of the mass of that dirt. Douglas Kell, chief executive of the U.K.'s Biotechnology and Biological Sciences Research Council, has calculated that replacing 2 percent of the world's annual crops with perennials each year could remove enough carbon to halt the increase in atmospheric carbon dioxide. Converting all of the planet's farmland to perennials would sequester the equivalent of 118 parts per million of carbon dioxide—enough, in other words, to pull the concentration of atmospheric greenhouse gases back to preindustrial levels.