Recent work by paleontologists Tomasz Baumiller of the University of Michigan and Forest Gahn of the Smithsonian's National Museum of Natural History offers new insights into the process. A report on their research appears in today's issue of Science.

Biologists long have speculated that predators and prey play a game of evolutionary one-upsmanship, in which an adaptation on the part of one---say, sharper teeth in a predator---prompts a "go-you-one-better" response in the other---tougher hide in the prey, for instance.

Hints that this has occurred are scattered throughout the fossil record, but not evenly, Baumiller said. During one part of the Paleozoic Era known as the Middle Paleozoic Marine Revolution, for example, the diversity of shell-crushing predators increased explosively. Around the same time, some 380 million years ago, mollusks and other shell-bearing marine animals developed better protective devices, such as more spines or more tightly-coiled shells.

Apparently, the prevalence of shell-crushers prompted development of better defenses against them. But simply finding evidence of changes in both predators and prey doesn't prove that one caused the other, Baumiller noted. "You have to provide evidence that they, in fact, were interacting."

To search for such evidence, Baumiller and his former graduate student Gahn, studied fossil crinoids, a group of marine animals related to starfish and sea urchins. Crinoids, also called sea lilies, have feathery arms that they extend to catch bits of plankton or detritus passing by in the current.

Like their starfish cousins and other animals in the group known as echinoderms, crinoids are capable of regenerating lost body parts. Because modern day crinoids usually lose---and regenerate---their arms as a result of attacks by fish, Baumiller and Gahn reasoned that arm regeneration in fossil crinoids would be a good indicator of predator-prey interactions in the geologic past.

To test their idea, they examined more than 2,500 Paleozoic crinoids for evidence of arm regeneration, focusing on fossils from the Ordovician to Pennsylvanian Periods (490 to 290 million years ago). As predicted, they found that the only significant increase in regeneration frequency was during the Middle Paleozoic Marine Revolution.

"Indeed, the frequency of regeneration, which we regard as a proxy for predation intensity, was low during intervals before the Middle Paleozoic Marine Revolution and then there was a sudden increase, coincident with the diversification of predators and the increase in the evolutionary response of the prey," Baumiller said.

That's not the whole story, though. Baumiller and Gahn suspect that crinoids were not the intended targets of the predators that inflicted damage upon them, and that their broken arms were simply "collateral damage." Crinoids, Baumiller explained, play host to a variety of other organisms that take up residence on various parts of their bodies, and the predators were probably going after those creatures.

One way to test this notion would be to look for correlations between the degree of infestation and the rate of regeneration, and Baumiller, Gahn and Carlton Brett, a paleontologist from the University of Cincinnati, have applied for funding to do just that.

The current research was supported by a Fulbright Research Fellowship to Baumiller and by grants from the Geological Society of America, the Scott M. Turner Fund (University of Michigan) and the American Chemical Society.