Forget about kids feeding virtual pets. Two researchers at the University of Nebraska at Lincoln report in today's issue of Nature that they're "feeding" several varieties of virtual moths to live blue jays to test complex evolutionary theories about the relationships of predators and prey.

In the wild, prey come in many varieties. Scientists have believed that predators focus their attention on the most common varieties of a species of prey and often overlook varieties that are more rare. But this theory of "apostatic selection" is difficult to test in the wild.

So the two authors of the Nature paper -- Alan B. Bond, a research professor of biological sciences at Nebraska, and Alan C. Kamil, a professor of biological sciences there -- instead used trained blue jays and virtual moths.

The researchers used a computer to create five different types of the moths, based on photographs of two actual moths. The blue jays, which eat moths in the wild, were trained to peck at the computer screen wherever they saw a virtual moth; they were rewarded with a piece of mealworm if they correctly found a virtual moth.

But the blue jays found them. The six birds used in the experiment were each shown 120 scenes on the computer screen each day for 50 days. As the theory of apostatic selection predicts, the birds tended to find the more common types of virtual moths and ignore the less common types.

Moreover, the blue jays' affinity for various types of moths changed as the abundances of the varieties changed. Each day of the experiment, the researchers created a new "generation" of virtual moths. The relative number of varieties of moths in the new generation was identical to the proportion of the types of moths in the preceding generation that escaped detection by the blue jays on the preceding day.

"We created an ecosystem in a computer," says Mr. Bond.

In each new generation, the jays tended to pick the most common varieties and ignore the less common types. At the start of the experiment, the proportions of different varieties varied sharply from one generation to the next, but by the end of the experiment, the populations of the different varieties were constant from one generation to the next. Mr. Bond says that suggests that apostatic selection may be responsible for maintaining the various proportions of varieties of a prey species.

Mr. Bond says he believes the experiment reflects the birds' behavior in the wild. "This is what they do for a living," he says. "They're extremely good at finding the real animals on real tree trunks."

However, he says that scientists are not yet sure why blue jays and other predators ignore less common varieties of prey. One possibility, he said, is that it is more efficient for a predator to look for the most common varieties of prey. If the predator looked for less common varieties, it would miss the opportunities to eat the more frequent varieties, Mr. Bond says.