Feathers Don't Make the Bird

By Josh Fischman|Friday, January 01, 1999
The dead do not all rest in peace at the Royal Tyrrell Museum in Alberta. One of them rests on an artist's easel. The three-foot dinosaur has its head pulled back in the rictus of death, the neck and tail bowed toward each other. Extending from the tail is a peacocklike fan of feathers; other feathers form fringes under the arms. The figure is actually a model of the real fossil, which remains in China, where it was discovered and where Tyrrell paleontologist Phil Currie had the privilege of examining it. "It's a beautiful specimen," he says. "You look at the anatomy on that thing, and everything that's been used to separate birds from dinosaurs is there. But it's too primitive to be a bird. It didn't fly. The only thing truly birdlike about it is its feathers."

And that means, Currie says, that feathers don't make the bird--they could make the dinosaur just as easily. The animal, named Caudipteryx, came to light this June when Currie, geologist Ji Qiang of the National Geological Museum of China in Beijing, and their colleagues published a description of it and another feathered fossil named Protarchaeopteryx. For many paleontologists, these overlapping bird and dinosaur features sealed the 30-year argument that birds are the direct descendants of such dinosaurs. And the discovery seemed to push some intractable opponents of this idea farther out on a limb.

For most of this century, scientists thought that birds and dinosaurs had branched off on their separate paths from primitive reptiles more than 250 million years ago. But in 1970, Yale paleontologist John Ostrom was struck by the similarities between the oldest known bird, the 150-million-year-old Archaeopteryx, and a two-legged carnivorous dinosaur named Deinonychus. After Ostrom's insight, scientists began piling up the anatomic similarities between the two groups until, by the mid-1980s, they had a list of about 150 shared features. One they didn't have was feathers, and given the ravages of time and the delicate nature of plumage, finding some seemed unlikely.

Then, in 1996, Ji presented Currie with a gift-wrapped box containing a chicken-size dinosaur dubbed Sinosauropteryx. The fossil was surrounded by a featherlike fringe. Paleontologists are still arguing over just what it is--protofeathers, long scales, or something more exotic--but a year later Ji produced another dinosaur so similar to the earliest bird that he called it Protarchaeopteryx. And he had not just one but three, all from Liaoning Province in northeastern China, that were between 150 and 120 million years old.

Currie hotfooted it back to China to examine the new fossils, and thanks to the fine-grained sediments in which the animals were buried, he could make out their anatomy in stunning detail. "You could definitely see there were feathers attached to the arms and the tail," says Currie. These feathers were long, with tightly interlaced strands splitting off from a central spine, like modern flight feathers; the bodies were covered with fuzzy, downy structures. On close examination, the second and third specimens turned out to be yet another species, which Currie and Ji named Caudipteryx, which means "tail feather."

And then the scientists did what many a birder would do when faced with homeless avians: put them in a tree. Ji, Currie, and their colleagues compared the two new fossils with Velociraptor, a sleek dinosaur considered by many researchers to be a close relative of birds, as well as Archaeopteryx and other primitive birds. The analysis produced a sort of family tree, with its first split coming between the lineage leading to Velociraptor and the line leading to all the other dinosaurs and birds. Protarchaeopteryx branched off this line soon after that, and then Caudipteryx's ancestry split away from that of Archaeopteryx and all modern birds. In other words, the Chinese fossils were dinosaurs. That meant that feathers, apparently, cloaked dinosaurs long before birds arose, perhaps functioning as a layer of insulation or as courtship display. Only later did protobirds adapt them for flight.

Skeptics of the bird-dinosaur link have long pointed out that what appear to be the closest relatives to birds lived 80 million years after Archaeopteryx. "The timing is a real problem," says ornithologist Alan Feduccia of the University of North Carolina. The new fossils don't resolve the gap. Feduccia suggests that the pre-Ostrom idea still holds and that the Chinese fossils are early, flightless birds. "Caudipteryx is really just an old kiwi."

Currie admits at this stage there are still shaky branches on his tree, but the classification of the Chinese finds as feathered dinosaurs is not one of them. Feduccia's scenario creates an even bigger disparity with the fossil record and requires that the Chinese animals independently evolved many dinosaurian traits. "It's good to have a dissenting voice that says, 'Hey, there's more than one possibility here,' because it makes you look at the specimens harder," says Currie. "But some of these arguments are simply drifting into the realm of silliness."


The Old Fish of the Sea Marine biologist Mark Erdmann was strolling through a market in the city of Manado on the Indonesian island of Sulawesi when he was brought up short by a fish. In a cart being wheeled past him was a homely coelacanth, one of the most famous, yet most elusive, fish on the planet. Sporting fleshy, limblike fins, it originated 400 million years ago from the same stock of fish that gave rise to the ancestors of land-dwelling vertebrates called tetrapods. Among living animals, only lungfish are closer kin to tetrapods. But until just 60 years ago, no one knew that coelacanths were among the living--they were thought to have been extinct for some 80 million years. Then in 1938 fishermen hauled in a live specimen off the coast of Cape Town. Since then the fish have been found in a few other patches of the ocean near Madagascar, Mozambique, and the Comoro Islands. On the whole, biologists viewed these populations as tiny remnants of a once-great dynasty.

Thus, to see a coelacanth 6,200 miles away from its known habitat was decidedly odd. Unfortunately Erdmann was unable to buy and study the fish in the cart--he was on his honeymoon at the time--but he was so intrigued that he spent ten months interviewing local fishermen in search of another specimen. (Along the way he learned that on Sulawesi the fish is called raja laut, meaning "king of the sea.") His persistence was rewarded this past July when a four-foot, 64-pound coelacanth (shown here) was caught in a deep-set shark gill net placed off the volcanic island of Manado Tua.

"It's a very important discovery," says Roy Caldwell, Erdmann's postdoctoral sponsor at the University of California at Berkeley, "because coelacanths are in serious trouble in the Comoros--the population has probably halved in the last ten years. People there fish for oilfish and catch coelacanths as a by-product."

Erdmann froze the Indonesian coelacanth shortly after its death, and a genetic analysis of its tissues is now under way to compare it with the African populations to see if it's a separate species or not. "If they are related," says Caldwell, "I'd bet Indonesia is the source population." An ocean current flows from the Philippines past the Sulawesi Sea, through the Makassar Strait, and into the Indian Ocean, where it travels westward, to the Comoros. --Josie Glausiusz


Searching for the First Animal The earliest fossils of many of today's animal lineages appeared about 540 million years ago, in an episode known as the Cambrian explosion. These Cambrian critters must have had predecessors, but they would most likely have been too small and soft to fossilize well. In February, however, two teams of paleontologists reported finding the fossilized remains of some of those ancestors. Both groups found their fossils in the 570-million-year-old Doushantuo formation in southern China. The rocks are rich in phosphate, which diffuses into small organic structures that preserve exceptionally well. "Under the right circumstances, very minute and delicate animal remains can be preserved," says paleontologist Andrew Knoll of Harvard. "That gives the promise of discovering the real roots of animal history."

In the Doushantuo rocks, Chia-Wei Li of the National Tsing Hua University in Taiwan and his colleagues found fossils of primitive sponges--the oldest remains ever found from a living group of animals. Meanwhile another group, which included Knoll, his graduate student Shuhai Xiao, and Yun Zhang of Beijing University, found balls 500 micrometers in diameter consisting of one, two, four, eight, and more cells--in other words, embryos in the earliest stages of division. While the exact identity of the embryos is a mystery, they have a mirrorlike symmetry, as do all animals known as bilaterians--a group ranging from humans to worms to starfish. Embryos of more primitive animals, such as jellyfish, have a radial symmetry.

The common ancestor of both groups, therefore, still remains to be discovered in older rocks. One tantalizing, albeit controversial, clue to how far back those fossils may go emerged in October, when Yale paleontologist Adolph Seilacher reported traces of the tracks of worms in rocks dating back 1 billion years. It's likely that the Chinese fossils, as remarkable as they are, are only the beginning of the story. "In these rocks we are not at, or even near, the base of that animal record," Knoll says. --Kathy A. Svitil


Amber Ants Ants are the Bill Gateses of evolution--their ecological success is so spectacular that they thrive just about everywhere, particularly in rain forests, where they may constitute up to a third of the animal biomass. Yet the path to their triumph is hard to trace because their small, easily degraded bodies don't fossilize well. In 1998, however, their history became much clearer thanks to the discovery of the oldest known ant fossils.

Researchers from the American Museum of Natural History discovered seven 92-million-year-old ants preserved in amber in a rich deposit in central New Jersey. They were sure they had found ants and not some related insect because they could see a structure unique to ants called the metapleural gland. It secretes antibiotic substances that protect ant nests in humid places, such as rotten trees or in the ground, from infestation by bacteria or fungi.

The amber ants provide a suggestive picture of early ant evolution. The insects are divided among a few separate subfamilies of ants, meaning that by 92 million years ago ants must already have achieved a fair amount of diversity. That diversification might well have begun as far back as 130 million years ago, when related insects such as wasps, yellow jackets, and bees began to branch out. Yet the fossil record hints that ants began to prosper only after the end of the Cretaceous Period 65 million years ago, when the dinosaurs went extinct. Why ants needed so long to hit their stride remains an open question. --Michael Hagmann
Comment on this article