In the never-ending war between farmers and insects, the ultimate weapon on the growers' side is a protein produced by a soil bacterium, Bacillus thuringiensis - Bt for short. Bt massacres certain major crop pests, but is environmentally correct. Bt is harmless to humans and most forms of life.

When it comes to cotton crops, that enemy insect is the pink bollworm - Pectinophora gossypiella by name. "A pink bollworm is a caterpillar," explained entomologist Bruce Tabashnik at the University of Arizona in Tucson, "so when it grows up it's a moth - that is, a lepidopteran.

"Bt toxin has a cloned bacterial gene, which encodes the Bt toxin. It's lethal to some caterpillars - the larval stage of lepidoptera. So this Bt toxin has no toxicity to beetles, bees, mosquitoes, mites and the overwhelming majority of insects, let alone other kinds of organisms, which are not affected by the toxin directly.

"A conventional insecticide, a majority of which are neurotoxins," Tabashnik pointed out, "would kill not only most insects but at some concentration would be toxic to people, birds, fish. That's not the case with Bt toxin, which specializes in Bt cotton pitted against the pink bollworm."

Tabashnik is co-senior author of a paper in the Proceedings of the National Academy of Sciences (PNAS), released online Feb. 4, 2003. Its title: "Long-term regional suppression of pink bollworm by Bacillus thuringiensis cotton." Its first author is Yves Carriere.

"Our overall finding in this report," Tabashnik told BioWorld Today, "is that Bt cotton caused long-term suppression of pink bollworm, a major cotton pest in the southwestern U.S. During the study, we divided Arizona into 15 different regions, based on the clustering of cotton in various areas. The total cotton-growing acreage in Arizona is on the order of 250,000 acres per year. These 15 regions include all the cotton in the state. In any of the regions we could ask, What was the percentage of Bt cotton in this region, compared with non-Bt cotton?'"

Great Potential As Crop Protection Tool

"This was a 10-year study across Arizona," he recounted, "using data from thousands of cotton fields, showing regions of the state where Bt cotton is critical. It's not surprising at all that Bt cotton would control a pest during a single growing season. That's why it's useful. Growers plant Bt cotton because it has such a great potential as a crop protection tool. But what wasn't known before was how long the protection would last before resistance set in.

"What this study demonstrated," Tabashnik continued, "is density of pink bollworm over a five-year period, compared with the previous five years before Bt cotton was available commercially. This only happened, though, in regions where Bt cotton was used extensively. There is a threshold around 65 percent of Bt cotton acreage. So the places that used 65 percent or more of Bt cotton during the second five years of the study showed significant declines in pink bollworm population, whereas those areas that used less than about 65 percent did not experience the declines.

"One possibility - and it's still a major concern - is that resistance could evolve quickly in pest populations, rendering Bt crops useless. That's the worst-case scenario. And many of us expected that. However, Bt cotton in Arizona vs. pink bollworm, the worst-case scenario was not materialized. So far," Tabashnik added, "there are no documented examples of resistance evolving in the field to any Bt crop anywhere in the world.

"What led us to expect this was going to happen was that insects have evolved resistance to virtually all insecticides used to control them. There are more than 500 species of insects that have populations resistant to one or more insecticides. Insecticides are typically used intermittently, whereas Bt crops expose pests to a single toxin throughout the growing season. This is intense selection, much more so than the exposure to conventional insecticides. The other part of the puzzle is that many insects have been selected in the laboratory to acquire Bt resistance. So they have the genes in their populations to evolve resistance. And with pink bollworm in particular we have selected quickly in the lab strains that can survive on Bt cotton - are resistant to Bt cotton."

Bollworm Owns Resistance Genes But Holds Back

"So it's a paradox they have not found resistance as one might expect, because they have the genetic capability, and they're under intense selection pressure for the past five years," Tabashnik said. "Why are they not evolving? One of the essential ingredients appears to be the presence of crop refuges - substantial acreage of non-Bt cotton. That's a requirement imposed by the EPA - Environmental Protection Agency. So in the refuge areas, the susceptible insects not only can survive but they're at a competitive advantage over the resistant insects.

"In some of these regions," he went on, "the proportion of Bt cotton acreage is greater than 80 percent; in others, less than 20 percent. EPA requires that the percentage of Bt cotton grown by an individual farmer not exceed 96 percent. For the most part, the non-Bt-cotton refuges are where they need to be, which is within a mile or less of the Bt cotton fields. We can't say precisely that the refuges are working as designed, but everything we know now suggests that they are a key ingredient for delaying resistance."

Tabashnik and his co-authors are planning "to rigorously determine the relationship between the abundance and spatial distribution of refuges and the rate of resistance development to Bt cotton. We've mapped the Bt cotton fields throughout the state, and we're going to test for resistance by two methods: first, bioassays, in which the larvae, the caterpillars, are exposed to Bt toxin in an artificial diet, and second, we recently developed a DNA-based method where we can actually check for the resistance genes. So we're going to track over the next three or four years the extent to which resistance increases, and use statistical models," he concluded, "to evaluate how much resistance evolution depends on the distribution and abundance of refuges."