Little old whales. Skulls of 50-million-year-old pakicetid whales, with a coyote skull (bottom) for scale.
Dennis Normile
Rallying Round the Tertiary Radiation
In recent years, researchers who determine how long species have been diverging based on
differences in their DNA have pushed back the dates of emergence of modern mammals--the
predecessors of everything from whales to tree shrews--to as much as 100 million years
ago. That's far earlier than fossils suggest, but the DNA researchers blame the
discrepancy on the notoriously incomplete fossil record. At the meeting, however, two
paleontologists went on the offensive, claiming that a close look at the fossil record
shows that it is complete enough to date the origin of the modern mammalian orders. If the
DNA "clocks" can't agree with the fossils, says the author of one study,
paleontologist J. David Archibald of San Diego State University, then "the problem is
with the molecular clock."
Paleontologists have long held that the modern mammalian orders emerged and differentiated into families, genera, and species after the Cretaceous-Tertiary (K-T) extinction 65 million years ago. That event wiped out the dinosaurs and presumably gave mammals more evolutionary breathing room. But many in the molecular camp have argued that several orders of mammals, including primates and rodents, arose more than 35 million years before the K-T boundary. Divergence within a few if not many of the modern orders was well under way during the Cretaceous, they say (Science, 1 May, p. 675).
Now Archibald and ecological modeler Douglas Deutschman, also of San Diego State University, have surveyed the fossil record and report renewed support for the traditional paleontological view. Archibald notes that 15 of the 18 extant orders of placental mammals first appear in the fossil record during an evolutionarily short period of 16 million years in the early Tertiary. Statistically, says Archibald, the probability of this clustering occurring randomly due to gaps in the fossil record is "vanishingly small."
What's more, Archibald found no evidence that paleontologists have neglected the period before the K-T, creating a gap in the fossil record that might result in an apparent explosion of diversity later. He found roughly equal numbers of fossil sites and specimens for the 5-million-year periods immediately before and after the K-T boundary. Yet only 11 genera of placental mammals have been found in the fossil record in the period before the boundary, while 139 placental genera have turned up in the 5 million years after it. "Something happened to cause this explosion of speciation," he says.
The reality of the early Tertiary radiation was echoed by John Allroy, a paleontologist at the National Museum of Natural History in Washington, D.C., who did a similar analysis focused just on North American mammals and came to a similar conclusion. If molecular analysts can't find evidence of this explosion in speciation, he says, then they "don't know anything about the evolutionary process." Archibald suggests that something happened at the time to cause molecular clocks to speed up, making the splits among mammals appear earlier than they actually were.
The paleontologists' arguments were "a real eye-opener," says Michael Stanhope, a molecular biologist at Queen's University of Belfast, showing that the fossil record cannot be lightly dismissed. "I think there is a good chance we're missing something about the way DNA sequences evolve."
But not everyone was convinced. Peter Waddell, a phylogeneticist at the Institute of Statistical Mathematics in Tokyo, says that fossil evidence of the ancestors of aardvarks, tree shrews, and rabbits, among others, is missing. "The fossil record is not picking up things we know are there," he says, "so why close the book on [other] missing modern forms in the late Cretaceous?"
Others caution that the molecular evidence may never exactly fit the fossil data. Molecular phylogeneticists can only work with extant species, for example, and so will "never be able to reconstruct all those species which arose but died out," says molecular biologist David Mindell of the University of Michigan, Ann Arbor. There may also be a gap between the moment two DNA sequences begin to differ and the moment a species actually divides into two.
"We are hoping the molecular and morphological data converge," Waddell says, "but molecular people wouldn't necessarily be unhappy if they don't."
Whale-Ungulate Link Strengthens
Judged by its DNA, a whale is just an overgrown hippopotamus with an unusual lifestyle.
Researchers who learn how living animals are related by studying their DNA have tended to
group the cetaceans--whales, dolphins, and porpoises--with the even-toed ungulates, or
artiodactyls, which include cows, pigs, and hippos. By some analyses, hippos are the
closest living whale relatives. But to paleontologists, who study fossils, that conclusion
has long been anathema. Instead, they contend that cetaceans descended from extinct
hyenalike mammals called mesonychians. Now the fossil record may be opening the door to a
whale-ungulate connection.
At the meeting, Hans Thewissen, a paleontologist at Northeastern Ohio Universities College of Medicine in Rootstown and an expert on whale evolution, described analyses of new specimens of early whales and whale ancestors his team collected in Pakistan. The new specimens weaken the link between the whales and the mesonychians, which was primarily based on similarities in the teeth. But they support the idea that whales are cousins of the ungulates, if not actual members of that group, he reported. "I think there is no doubt that they are very closely related to artiodactyls," says Thewissen.
Little old whales. Skulls of 50-million-year-old pakicetid whales, with a
coyote skull (bottom) for scale.
J.G.M. THEWISSEN
One blow to the mesonychian link came from two specimens of a 50-million-year-old whale, a member of the family Pakicetidae. Analysis by a colleague of Thewissen's, Maureen O'Leary of the State University of New York, Stony Brook, showed that its teeth are not as highly evolved as those of the mesonychians, making it unlikely that whales are the descendants of that group. But on the question of whether the cetaceans are an actual subgroup of the Artiodactyla, as the molecular biologists think, this and other fossil whales don't give a clear answer.
Thewissen says that five morphological features of the early whales, including features of the skull, upper teeth, and feet, are "not inconsistent" with the hippo hypothesis. In particular, the new pakicetid skulls have holes over the eye sockets, known as supraorbital foramina. These features are not known in modern whales but are common to all artiodactyls.
But the last molar on the lower jaw, which has three sections in artiodactyls, has just two in whales. And in artiodactyls, the astragalus, one of the anklebones, has a rounded head and other characteristics that make the ankle much more flexible than it is in any other mammal. Thewissen recently discovered an anklebone from an early whale ancestor that still had legs. It lacks the rounded head, although in other respects it is similar to an artiodactyl astragalus.
Still, Thewissen thinks the morphological evidence, although mixed, opens the door to some kind of relation between the whales and the ungulates. He adds that there is now "considerable doubt" that cetaceans are closely related to mesonychians. That conclusion got a thumbs up from paleontologists at the meeting. For example, John Allroy of the National Museum of Natural History in Washington, D.C., says pulling the mesonychians out of the picture makes a closer cetacean-artiodactyl link plausible. But O'Leary says "it's [still] difficult to connect hippos with whales in the fossil record."
The molecular camp, for its part, viewed Thewissen's conclusion as just a first step toward ultimate vindication. As Norihiro Okada, a molecular biologist at Tokyo Institute of Technology, put it: "I think paleontologists may discover more [features common to early cetaceans and early hippos] in the near future."