SAN DIEGO—In the next 50 years, humans will have to produce as much food as we have over the entire history of civilization. The planet’s ever-expanding population demands it. Yet productive farmland is scarce, and other resources such as water and fertilizer (which is made from fossil fuels) become more constrained by the day.

Such is the dilemma of the world’s agronomists, as described this weekend at the annual meeting of the American Association for the Advancement of Science. Fortunately, clever scientists are not without solutions. Jonathan Lynch of Pennsylvania State University argued that our present circumstances demand what amounts to a “second Green Revolution.” The first, he said, occurred in the 1960s. The widespread use of fertilizer in the developed world actually reduced the yields of crops such as corn—the extra nutrients made the plants grow so tall that they would simply fall over. The solution involved the development of dwarf varieties of corn and rice that could (quite literally) stand up to added fertilization.

What we need now, he says, is a system that expands yields in poor soil without the use of fertilizer, which is still prohibitively expensive in much of the world. The need is dire: every year Africa loses 95 percent of its potential corn yield to infertile soil, and more than seven million children under five die from hunger.

The primary constraint on many soils, he said, is low phosphorus content. The phosphorous that does exist is concentrated in the top layers of soil. He described a program in the highlands of Latin America that was able to extract more of the phosphorus from the soil by breeding varieties of the common bean—a critical food source for over one billion people—that have root structures that spread out horizontally beneath the soil. The new roots capture more of the phosphorous. They also have the added benefit of protecting the soil against erosion in the mountainous regions where beans are grown. He is currently extending his research to corn, which has the converse problem: it needs deep roots to capture nitrogen, which sinks deep beneath the surface.

In addition, Robert Fraley, the chief technology officer of Monsanto, told me in a private conversation about the power of genetic engineering to help build yields in corn and soybeans. He said that this is the first year that people will have access to the full genomic sequence of corn and soy, a development that he said was like turning on a light when you’re trying to find your keys in a dark room. “Breeding has been based on characteristics, not genes,” he said, “but modern plant breeding looks more like CSI.” Whereas biotech crops have in the past been used to create insecticide-tolerant crops, the next generation of biotech will attempt to increase yield directly. “We’re looking at a whole series of genes that can affect yields,” he said, “genes that can tweak the efficiency of photosynthesis, or that can be used for increased drought and temperature resistance.”

He said that for corn, he expects a doubling of yield from 2000 levels in the next 20 years. Ten percent of improvements will come from agronomic improvements such as GPS-equipped tractors that can plant seeds in finely optimized locations. Thirty to forty percent will come from breeding gains. The remaining half will come from biotech. And all these gains and yield can happen while reducing inputs such as water and fertilizer and pesticides by a third. “The way that agriculture is poised, it’s like computers were in the 1960s,” he said. “Now you’ve got fundamental breakthroughs in genome sequencing to add to the breeding. On the biotech end you’ve got the modes and actions of specific genes. And then you can integrate it together to plant the right seeds on the right fields at the right plant density at the right time.”