It might look docile lolling among the water lilies, but bad-tempered and surprisingly swift on terra firma, the hippopotamus has a deservedly fearsome public image in its African homeland. It also has a formidable reputation among evolutionary biologists: the beast has defied attempts to pinpoint its origin for nearly two centuries. To that end, recent findings may finally put the hippo in its place.

With its gaping maw, hairless body and eyes that sit high on its head, the semiaquatic hippo is one of the most distinctive members of Africa's mammalian menagerie. Two species exist today--the common Hippopotamus amphibius and the smaller Liberian hippo, Choeropsis liberiensis--and 40 more are known from the fossil record. Experts agree that hippos belong to the mammalian order Artiodactyla, a group of even-toed, hoofed creatures whose extant representatives include camels, pigs and ruminants such as cows. But exactly where hippos sit on the artiodactyl family tree has proved devilishly difficult to discern.

Two hypotheses lead the pack. The first posits that the piglike peccaries, or tayassuids, are the closest relatives of the hippo. The second proposes that extinct swamp-dwelling creatures called anthracotheres own that distinction. To evaluate the two scenarios, Jean-Renaud Boisserie, a postdoctoral researcher at the University of California at Berkeley, and his colleagues scrutinized all the characteristics ever used to support one or the other of these models, incorporating data from 32 artiodactyl species (including new fossil hippos from Chad and Ethiopia).

In presenting their work at a fall meeting and, more recently, in a paper published online January 24 by the Proceedings of the National Academy of Sciences USA, the investigators concluded that many of the putative resemblances between hippos and peccaries--the rounded shape of the muzzle in cross section, for example--are in fact primitive artiodactyl traits and therefore not indicative of a deeper relationship. Other similarities were also judged meaningless. The anthracothere hypothesis fared much better: the team's analysis supports a link between hippos and anthracotheres, pointing to an especially close relationship with a dentally advanced subset of anthracotheres known as the Bothriodontinae. Although hippo teeth look rather different, the two groups have in common a number of features in the skull, lower jaw and limbs.

The results stand to elucidate not only the ancestry of hippos but also that of whales. In 2001 key fossil whale discoveries revealed the ocean dwellers to be descended from artiodactyls [see "The Mammals That Conquered the Seas," by Kate Wong, Scientific American, May 2002]. And several DNA analyses have concluded that whales and hippos in particular share a common ancestor. But some paleontologists have been reluctant to embrace the molecular findings because whereas the oldest known whales date back to more than 53 million years ago, the earliest hippos yet found are only around 15 million years old. The fossil trail of anthracotheres, however, doesn't peter out until some 41 million years ago. An anthracothere origin of hippos could thus reduce the gap between them and whales to just 12 million years. "This is the best work so far to link anthracotheres to hippos," comments fossil-cetacean expert J.G.M. (Hans) Thewissen of the Northeastern Ohio Universities College of Medicine. The challenge now, he adds, will be to identify anthracothere ancestors from the right time and place to bridge the remaining break between hippos and whales.

The hippo will no doubt remain a force to be reckoned with in the wilds of Africa. But paleontologists may have at last wrestled with the river horse and won.