Bt broccoli test: Refuges cut pest resistance.

A field test with broccoli genetically engineered to make its own Bt pesticide has confirmed the value of maintaining nearby stands of the traditional crop to curb pest resistance, according to an international research team.

"This is really the first field test of a resistance-management strategy for a Bt crop," says Anthony M. Shelton of Cornell University. Starting in 1995, the Environmental Protection Agency registered brands of corn, cotton, and potatoes that carry genes from the pesticide-producing bacterium Bacillus thuringiensis.

To ease fears that widespread use of Bt crops will promote insect resistance, EPA already requires that crops deliver a high dose of Bt and that growers plant Bt-free refuges for insects. The original refuge requirements for Bt cotton--4 percent of acreage not sprayed with pesticides at all and 20 percent of sprayed acreage--come up for renewal January 2001. Last January, EPA mandated that Bt corn growers designate 20 percent of the areas planted as refuges.

The current strategy comes from the observation that insects need two copies of a resistance gene to laugh off Bt. A refuge is expected to provide a haven for Bt-susceptible moths. When they mate with resistant pests, the mixed-susceptibility offspring perish when they encounter Bt, knocking the resistant parent's genes out of circulation.

However, "testing resistance management is extraordinarily difficult," explains coauthor Richard T Roush of the University of Adelaide, Waite Campus, in Australia. Ethics concerns complicate outdoor research on insects bred to be resistant, and natural resistance rarely evolves quickly enough to study.

"We could only do this because of a unique system," says Shelton. He and his colleagues found an ideal test insect, the widespread diamondback moth, which evolved resistance to Bt sprays in Hawaii. The moths regularly hitchhike to the test area in New York but can't survive the winter. Coauthor Elizabeth D. Earle of Cornell then engineered Bt broccoli for the insects to attack, more as a test vegetable than a commercial breakthrough.

The researchers released populations of moths with known frequencies of resistance genes into various arrangements of Bt and non-Bt broccoli. Randomly mixing the two types of plants in the same field didn't preserve susceptibility as well as planting the types in separate patches did, the researchers report in the March NATURE BIOTECHNOLOGY.

Farmers, who get cranky about growing crops just to breed bugs, sometimes spray refuges with pesticides other than Bt. The researchers found that as models predict, sprayed refuges did not work as well as unsprayed refuges of the same size. Spraying the refuge will reduce the frequency of genes for susceptibility in the insect population, Shelton says.

The broccoli study is only a first step, cautions Fred Gould of North Carolina State University in Raleigh in a commentary in the same journal. However, he welcomes confirmation of the shortcomings of the mixed-plant or sprayed refuges. "Field studies like this one are essential for developing public confidence," he says.

Jane Rissler of the Union of Concerned Scientists in Washington, D.C., also applauds the study despite its limitations. "I think it's research that we need a lot more of," she says. "We needed it before 1995, when they started approving these crops."