It is a shrew-like creature that no one has ever laid eyes on and whose fossil remains have never been found. Yet it must have lived once, rustling in the underbrush over 60 million years ago, because without it we wouldn't be here now.
So say researchers behind a massive effort to trace the family tree of all placental mammals, a fantastically diverse set of species that includes everything from bats to whales to humans. In a landmark study published Thursday in the journal Science, the team provides both a reconstruction of our tiny fur-clad forebear and fixes it to a crucial period in paleo-history: the first few hundred thousand years after a devastating asteroid impact emptied Earth of its dominant life forms, the dinosaurs.
The finding promises to heighten a growing debate about precisely when mammalian diversity first took off and why. But what has scientists just as excited is where the ancestor was unearthed: not in a paleontologist's bone bed but in a killer web application dubbed Morphobank.
"This is seriously cool," says team member Mary Silcox, a biological anthropologist at the University of Toronto, Scarborough, who helped to build the interactive database that merges information drawn from a broad range of species to reveal the evolutionary links between them. "It's really going to change the way we work."
Maureen O'Leary, a paleontologist at the State University of New York at Stony Brook and one of the leaders of the study, agrees the approach marks the arrival of "big data" to a science that is better known for its museum displays and narrow specialties.
"I think it's a tool that can really break things open," Prof. O'Leary says. "It allows people to work together in big teams and attack big questions as they do in other areas of science."
To get at the question of mammal diversity, Prof. O'Leary and her colleagues spent six years developing and feeding MorphoBank with data on 4,500 anatomical traits across 86 representative mammal species, both existing and extinct. Nearly a third of the traits involved differences in mammal teeth – that part of the study Prof. Silcox oversaw. "Teeth are important because they are the most common fossils and they can really tell you a lot about what an animal was doing," she says.
The study focused on placental mammals, which give birth to live young, but lack pouches, unlike Australia's marsupials. It is known that placentals diversified explosively at some point, but when is a matter of dispute. Some studies based on accumulated changes in DNA over many generations suggest that placental mammals diversified long before the dinosaurs disappeared.
Prof. O'Leary and her collaborators found otherwise. With all the information they'd gathered in one easy-to-manipulate database, the team worked out the simplest way to relate placental mammal species to one another and create a profile of the hypothetical animal that gave rise to them all.
The result is a small, insect-eating progenitor, the last common ancestor of all placental mammals, that they say was still present 65 million years ago when all dinosaurs except birds went extinct. That creature went on to diversify and colonize many of the ecological niches the dinosaurs left vacant, the team argues.
Others caution that in the absence of a real fossil the proposed creature may be difficult to reconcile with other data. "We're never going to have the full story," says Craig Scott, a paleontologist and curator of fossil mammals at the Royal Tyrrell Museum in Drumheller, Alta.
Prof. O'Leary says that while it's unlikely the exact ancestor their study points to will be uncovered at a fossil dig, "we may well find something that will help us reinforce our picture."