Ancient fungal farmers.

Long before ancient humans picked up a hoe, ants cultivated food. About 50 million years ago, these six-legged agriculturists began developing a strong, mutually beneficial partnership, or symbiosis, with a species of fungus. Today, many of the 200 species of attine ants still farm that fungus, even though the ants have spread north to New Jersey and south to Argentina, says Ulrich G. Mueller, an evolutionary ecologist at Cornell University.

The ants pass the fungus along by allowing new queen ants to pack some starter fungi in their mouths. They use this package to clone a new crop in their own nests. Worker ants quickly expand the crop to a room-size "field" that provides essential nourishment for larvae, if not for all the ants, Mueller says. The fungus does not propagate on its own.

Though entomologists have known about these ant farmers since 1874, biologists are only now figuring out the history of this symbiosis. At Cornell, Ted R. Schultz and Rudolf Meier examined and categorized 51 attine ant species. Then the Cornell group and Ignacio H. Chapela of the U.S. Department of Agriculture in Beltsville, Md., used genetic techniques to work out the taxonomic relationships of 19 fungi from ant nests and 16 samples of free-living fungi from the wild.

The 200 ant species share a common fungus-farming ancestor, Chapela and his Cornell collaborators report in the Dec. 9 SCIENCE.

The genealogy of the fungi is not as clear. In many, but not all, cases, the original fungus evolved into new species in parallel with their ant farmers. Today, the fungi divide into three groups, which coincide well with groupings of ants based on the shape and structure of the ant larvae, the researchers note.

The most recently evolved group consists of the leaf-cutting species, infamous for defoliating 20 percent of the tropical forests in the Western Hemisphere. Only their fungi can use fresh plant material, and both the ants and their fungi seem to come from a single ancestor, Mueller explains.

Fungi in the two other groups rely on decaying organic debris. One group contains ants and fungi descended from an ant that definitely found a new -- and different -- fungal partner at some point. The final group includes ants that seem closely related to the early symbiotic ants. The partnerships in this last group exhibit less parallel evolution than those in the other two, Mueller notes.

In the same issue of SCIENCE, a second team independently reported similar results. Gregory Hinkle and Mitchell L. Sogin of the Marine Biological Laboratory in Woods Hole, Mass., and their colleagues analyzed fungi from five attine ants and two samples of free-living fungi. They studied a different gene, one that evolves more slowly than the one analyzed at Cornell. The existence of this ancient partnership implies that concerns about species preservation must consider not only organisms big enough to see, but their microbial partners, too, Sogin asserts.