It seems like such an elegant use of genetic engineering.
Splice the genes for a naturally pest-killing microbe into a crop plant so that the corn or wheat or potato plants themselves become lethal to the pests. No more wasteful -- and expensive -- spraying of toxic pesticides. Instead, the bugs taste the leaves and go off to die.
Indisputably, it works. The genes of Bacillus thuringiensis, or Bt, have been added to corn, cotton, potatoes, and other crops, with the desired results. Scientists are aware, however, that any effort to manipulate the ecological balance must reckon with the evolutionary consequences. So even before the first Bt crops were planted, scientists, farmers, and regulators began worrying about unintended effects.
For example, insects that feed on the Bt-enhanced plants are continuously exposed to the lethal Bt toxin, not intermittently exposed, as when Bt is sprayed. The plants kill most of the insects, but a few bugs will survive because of random mutations that make them resistant.
They'll not only survive, but they'll gain a reproductive edge, breeding with each other to create strains of Bt-resistant bugs that could dominate an area. The greatest fear is that the resistant insects would prevail to such an extent that Btin the form of a spray, used by organic farmers for more than 50 years, might become useless too.
To forestall this catastrophe, the federal Environmental Protection Agency called for farmers to create "refuges" -- fields planted with the original, non-Bt-modified crops. In these refuges the pests' reproduction will be free of the selection pressures favoring resistant mutants. The few Bt-immune individuals will be overwhelmed by the wild-type, Bt-susceptible insects. And when the resistant bugs mate with the wild-type individuals, their offspring will be vulnerable to the Bt toxin.
There's general agreement refuges should work. The questions arise over how large the refuges must be (as a percentage of the farmer's total crop planting) and whether the requirement will be economically disadvantageous to farmers.
For example, they would have to spray the refuge crops with traditional Bt pesticide -- so why not just spray all the crops with it? The debate goes on, and the problem remains unresolved.