When Motoyuki 'Moto' Ashikari comes into his laboratory in the evenings, staff at Nagoya University in Japan often mistake him for a student. Who else would be working late wearing shorts and muddy sandals? But if you want to hunt genes that can make rice a hardier crop, Ashikari says, the best place to be is under the Sun in a swampy rice paddy.

In this issue, Ashikari and his team explain how certain varieties of rice manage to keep their leaves above water in flooded fields. When a gas released by the plant builds up in the hollow, submerged stems of a 'deepwater' rice variety (Oryza sativa ssp. indica), the plant rapidly elongates. Ashikari and his colleagues had previously linked aspects of this response to a section of the rice chromosome 12. Now they have identified the genes that allow this rice plant to rocket upwards when submerged (see page 1026).

It took five years, Ashikari says, to produce sufficient crosses of deepwater and non-deepwater rice to pinpoint the key stretches of DNA, and the roots of the project go back even further. In 1998, Ashikari was working on his doctorate when the Japanese government launched the Rice Genome Research Program. Ashikari joined up because he thought that rice would make an interesting research model — unlike some other plant models, cultivated rice can crossbreed with its wild counterparts. “I thought wild rice might have many agriculturally important traits hidden in its genome,” he recalls. He hoped to be able to restore some such traits to the higher-yielding crop varieties.

In general, traits are not controlled by single genes, but rather by the combined effects of a host of genes, each of which has more than one version. Identifying the stretches of DNA that bear the responsible genes — quantitative trait loci (QTL) — is a matter of statistics and breeding a vast number of plants. “Measurement is important,” Ashikari says. “In the mapping of single mutant genes, the mutation is either there or it isn't, but QTL analysis is not so simple — it maps traits on a continuous spectrum.”

Ultimately, Ashikari and his co-workers used tens of thousands of crosses to narrow down the genes responsible for flooding-induced elongation. Now, they reveal two genes that respond to the natural build-up of ethylene gas in stems that are trapped underwater. The genes' corresponding proteins launch a signalling cascade that causes the stems to shoot skywards.

The authors also bred the identified genes into non-deepwater rice and tested the new variety's growth in a slow flood. Recreating conditions in the lab that mimicked the monsoon-driven flooding common to many rice-growing regions was “laborious”, Ashikari says. “It built muscles.” His team transplanted plants from the field that had reached a height of 30–50 centimetres into 3,000-litre tanks, with only the top third of the plants above water. When the researchers raised the water level by 10 centimetres a day, plants with the deepwater QTL easily outgrew the rising waters. Those without them did not.

After all the work to identify these genes for flood-resistance, Ashikari needed to name them. He settled on SNORKEL1 and SNORKEL2. The rice stem is, after all, a hollow tube that allows an organism to exchange gas while underwater. And, Ashikari notes, the word 'snorkel' sounds similar in English and Japanese.