Cassava is a perennial plant grown as an annual crop. Farmers harvest the edible roots just before the rainy season, then take the stems and replant them for the next year. Cassava is drought resistant and particularly adapted to harsh, poor environments.

Because farmers replant the stems from year to year, the crop basically reproduces as a clone. To produce an improved hybrid through conventional methods could take hundreds of thousands of crosses over many decades, Ceballos said.

"We find good hybrids by chance, but we want to find them by design," he said.

Cultivated bananas and plantains (a starchy cousin of the banana that has to be cooked before eating) are practically sterile. Breeding new hybrids requires either using the occasional fertile seeds from domestic crops or seeds from wild bananas, which are a threatened resource, Lorenzen said.

Chan's technology could be used to make homozygous plants -- with identical sets of chromosomes -- that could be tested for new and useful traits, then used to breed new hybrids, Lorenzen said.

Propagating crop plants like cassava, banana and plantain as vegetative clones is a big disadvantage when it comes to storing and distributing the plants, Chan said. Seeds are much hardier and easier to store and ship than roots or stem cuttings. Seeds are also less susceptible to viruses, and not subject to the same quarantine rules as vegetable materials.

The ability to propagate these crops as seeds would be a "huge, huge advantage," Chan said.

The objective of the NSF-BREAD program, part of the NSF's Plant Genome Research Program, is to support innovative basic scientific research designed to address key constraints to agriculture in the developing world. The program is jointly funded by the Bill & Melinda Gates Foundation and the NSF, and grants are reviewed and awarded through the same processes as other NSF grants.