The findings, by Jonathan B. Losos of Harvard University and colleagues, are detailed this week in the journal Science. Losos did much of the work before joining Harvard earlier this year from Washington University in St. Louis.
"Because of its epochal scope, evolutionary biology is often caricatured as incompatible with controlled experimentation," says Losos, professor of organismic and evolutionary biology in Harvard's Faculty of Arts and Sciences and curator in herpetology at the Harvard Museum of Comparative Zoology. "Recent work has shown, however, that evolutionary biology can be studied on short time scales and that predictions about it can be tested experimentally. We predicted, and then demonstrated, a reversal in the direction of natural selection acting on limb length in a population of lizards."
Losos and colleagues studied populations of the lizard Anolis sagrei on minuscule islands, or cays, in the Bahamas. They introduced to six of these cays a larger, predatory lizard (Leiocephalus carinatus) commonly found on nearby islands and known as a natural colonizer of small cays. The scientists kept six other control cays predator-free and exhaustively counted, marked, and measured lizards on all 12 isles.
Anolis sagrei spends much of its time on the ground, but previous research has shown that when a terrestrial predator is introduced, these lizards take to trees and shrubs, becoming increasingly arboreal over time. Losos and his colleagues hypothesized that immediately fol