According to a slew of news reports last week, a study of dust mites has just proven that evolution can reverse. One headline in the Daily Beast even declared that "Evolution Bites the Dust". Cute, but wrong.

The scientific paper that prompted these stories was based on a study showing dust mites evolved from free living to parasitic to free living again. Evolution seemed to have taken a U-Turn. I didn't see why this was surprising, so I asked some biologists.

The scientists weren't surprised either. They said that in fact there's no reason to doubt that evolution might reverse course and an organism might lose a specialized trait that had once been adaptive. The way science has understood evolution since Darwin, there's no ladder of progress. Living things evolve to become better adapted to their environments, and because conditions are always changing, fitness is a moving target. And so my sources said they'd be more surprised to find that evolution was not reversible.

"Environments are constantly changing, so what is good in one environment won't be good in the next one," said Penn biologist Joshua Plotkin. "The question of whether evolution is reversible presupposes that there's a direction to begin with."

Animals have lost all kinds of traits – fur, flight, vision, ability to parasitize certain hosts. It's still evolution and not devolution because new environments may render those traits more costly than beneficial.

The stories that played up the reverse evolution claimed it was a violation of something called Dollo's Law, which was formulated by Belgian biologist Louis Dollo more than a century ago, and stated that evolution did have a direction and organisms couldn't revert to a former state. But according Plotkin and other biologists I consulted, this is just an antiquated bit of history – not a "law" anyone takes seriously. And certainly any example of reverse evolution would have absolutely no impact on the validity of Darwin's theory, as the absurd Daily Beast headline implied.

One the other hand, Plotkin said he was genuinely surprised by a result that came out several years ago showing that on a molecular scale evolution might not be reversible. That study, by Joseph Thornton of the University of Oregon, was written up in this very nice story in the New York Times in 2009.

The study showed that changes in evolution happen through a complex combination of genetic changes that are very unlikely to happen in the reverse order. Often, for example, neutral mutations randomly spread through some fraction of a population, causing no effect until a second mutation creates an advantage but only in those that carry the first, neutral change. There's a lot of randomness in the system – many steps to an advantageous trait are not beneficial by themselves.

So traits may be lost, but probably not through the reverse sequence of the same genetic changes that brought them on. An animal may look like it's reverted to an ancestral state, but it will be genetically different. "In some sense this study is a molecular resurrection of Dollo's Law," Plotkin said. "That's really cool."