At the University of Illinois, using knowledge gained from the sequencing of the bee genome, Robinson's team will try to find which genes in the collapsing colonies are particularly active, perhaps indicating stress from exposure to a toxin or pathogen.

The national research team also quietly began a parallel study in January, financed in part by the National Honey Board, to further determine if something pathogenic could be causing colonies to collapse.

Beekeeper Hackenberg agreed to take his empty bee boxes and other equipment to Food Technology Service, a company in Mulberry, Fla., that uses gamma rays to kill bacteria on medical equipment and some fruits. In early results, the irradiated bee boxes seem to have shown a return to health for colonies repopulated with Australian bees. "This supports the idea that there is a pathogen there," Cox-Foster said. "It would be hard to explain the irradiation getting rid of a chemical."

Still, some environmental substances remain suspicious.

Chris Mullin, a Pennsylvania State University professor and insect toxicologist, recently sent a set of samples to a federal laboratory in Raleigh, N.C., that will screen for 117 chemicals. Of greatest interest are the "systemic" chemicals that are able to pass through a plant's circulatory system and move to the new leaves or the flowers, where they would come in contact with bees.

One such group of compounds is called neonicotinoids, commonly used pesticides that are used to treat corn and other seeds against pests. One of the neonicotinoids, imidacloprid, is commonly used in Europe and the United States to treat seeds, to protect residential foundations against termites and to help keep golf courses and home lawns green.

In the late 1990s, French beekeepers reported large losses of their bees and complained about the use of imidacloprid, sold under the brand name Gaucho. The chemical, while not killing the bees outright, was causing them to be disoriented and stay away from their hives, leading them to die of exposure to the cold, French researchers later found. The beekeepers labeled the syndrome "mad bee disease."

The French government banned the pesticide in 1999 for use on sunflowers, and later for corn, despite protests by the German chemical giant Bayer, which has said its internal research showed the pesticide was not toxic to bees. Subsequent studies by independent French researchers have disagreed with Bayer. Alison Chalmers, an eco-toxicologist for Bayer CropScience, said at the meeting Monday that bee colonies had not recovered in France as beekeepers had expected. "These chemicals are not being used anymore," she said of imidacloprid, "so they certainly were not the only cause."

Among the pesticides being tested in the American bee investigation, the neonicotinoids group "is the No. 1 suspect," Mullin said. He hoped results of the toxicology screening would be ready within a month.