Five hundred million years ago, spineless chordates slunk through Earth's Cambrian oceans. These unassuming creatures would eventually give rise to more complex vertebrates such as fish, dinosaurs and even us, so they are crucial evidence for scientists trying to trace animal evolution's early steps. But because these organisms lacked bones or shells, the soft-tissue features that managed to survive the fossilization process have made the specimens look very primitive, possibly excessively so.

Placing these rare suspects into evolutionary trees based on a few mushy features seemed a little fishy to one group of researchers at the University of Leicester in England. They propose, in a study published online January 31 in Nature, that some of these Cambrian chordates might not have been so simplistic after all. (Scientific American is part of Nature Publishing Group.)

What makes them say that? Countless hours of watching fish rot.

"It is pretty smelly work," says Mark Purnell, of the university's Department of Geology and a co-author of the new study. But "the results are worth the effort. It's telling us something new about how we should take care interpreting the fossil record."

The very first vertebrates had yet to develop a backbone, or any bones at all, making their Cambrian period fossils even rarer than those for later animals with bones, teeth and other hard materials, which would leave "at least a lifelike snapshot of its skeleton," Purnell says. Only on rare occasions were the mineral composition and environmental factors just right for the soft tissue to be fossilized.

But whereas bones can take decades to decay, soft features, such as eyes, gills and organs, can disappear in a matter of months, weeks and even days. So, the researchers thought that a little contemporary forensic sleuthing could help to flesh out approximately what these ancient creatures looked like in life half a billion years ago.

For the dirty job, Purnell and his team chose the larvae of Lampetra and Branchiostoma. No red herrings, these contemporary denizens of the seas are thought to bear close resemblance to Cambrian chordates. After making observations of how the fish larvae decomposed, the researchers deduced that many of the characteristics that were presumed to have been absent from some 500-million-year-old chordates might actually have just faded during the decay process before fossilization occurred.

"The idea that characters are lost—especially in soft-tissue creatures—to decomposition is not a novel idea," Purnell says. But, he notes, "the assumption has been that they are randomly distributed."

Purnell and his team found that, in fact, features disappeared in a distinctive pattern. Some parts, such as the gut and liver, lasted for 130 days, but the eyes were gone after 64 days and the multichambered heart after only about 11. "That the pattern of nonrandom decay in this group was so strong was really a surprise," Purnell said.

Such a consistent putrefaction pattern also comes as a surprise to others on the trail of early animal evolution. "There's a presumption that a lot of these extinct beasts should belong to very primitive and bizarre groups," says Bruce Lieberman, a professor in the Department of Geology at the University of Kansas in Lawrence, who was not involved in the study. In the rare cases when these ancient animals are well preserved, "we can recognize that they have a lot of commonalities with living things," he said. "A lot of the groups do undergo very early evolutionary bursts. After that burst, there's just a basic tinkering of forms."

Some of the features of these protofish that are most helpful in distinguishing them from their more primitive ancestors are, in fact, quick to decay, leading Purnell to caution that, "if it's only got very basal chordate characteristics, it could be that it's a more complex animal that has lost a lot of features because of decomposition…. We need to take that into account when we want to determine what the actual anatomy has been."

The fetid findings led the researchers to propose that misinterpreting decayed fossils has likely lead to frequent "stem-ward slippage," in which organisms are being consistently classified as more basal than they really were. They note that it might be "a widespread but currently unrecognized bias in our understanding of the early evolution of a number of phyla" because failure to recognize that some features might just have faded before fossilization can mean "erroneous evolutionary conclusions"—a stinky situation, indeed.

Reevaluating the evolutionary revamping in these first vertebrates "may change our idea about the rate and order in which characters appear," Purnell says. But Lieberman notes that the findings should not be taken too far in the lab. Just because a more advanced feature might have been lost to decay does not mean that it was, he cautions. "If we can't see them," he says about the speculative features, "there's still only so much we can do without evidence."

Scientists have not been blind to the fact that decay has surely played a role in changing many of the creatures they examine, but, says Lieberman, "It's really hard to take it into account in a quantitative framework." It can be especially challenging, he says, because fossils of one species can look quite different from one another if they come from different locations that have different geologic and mineral compositions.

The rotting fish have not yet provided an answer to this quandary, and Lieberman says that other than trying to get as many specimens as possible, he does not have a solution to the challenges of interpretation, "except to be darn careful."

Like any good Cambrian period caper, the mystery of early animal evolution just keeps getting stickier—and stinkier.