Still, scientists say that experimental populations evolving in parallel are not identical. For example, the genetic changes underlying identical adaptations in different populations can differ, even if in minute detail. Are these the subtle differences that could send evolution down a different path?
Only the definitive experiment which remains beyond the scope of the National Science Foundation budget can say.
"What we need are about 1,000 worlds to play evolution back correctly," Rieseberg said. "Then we can really find out what would happen."



