Science: Where there's a quill there's a way

A recent rich find of dinosaur fossils in China is casting new, if confusing, light on the evolution of flight. By Pat Shipman

PALONTOLOGISTS, WHO study fossils for a living, have got into an unseemly flap over the discovery of six stunning specimens of early bird-like dinosaurs from Sihetun in China. These fossils, which went on public display in Washington DC this week, hold the key to one of the biggest mysteries in evolution - how did birds take to the air and fly?

Despite years of study, the questions seem as insurmountable as ever:

l Did birds conquer the air from the ground up or the trees down?

l How could flight evolve at all, since half-flying is obviously dangerous?

l What good is part of a wing or a single feather?

l Are birds simply advanced dinosaurs who took to the skies - or are they descendants of some earlier, more general reptile?

Controversies began in 1861, when the first fossil skeleton of Archopteryx was discovered in the Solnhofen limestone of Germany two years after Charles Darwin published Origin of Species. This gorgeous half-bird, half-reptile is 150 million years old and yet has feathered wings, the sine qua non of birds.

The feathers themselves have a modern structure, with a quill or rachis dividing the feather into two unequal vanes for aerodynamic efficiency. But Archopteryx is not wholly bird-like; it has three wickedly clawed fingers on each wing, a long, bony, reptilian tail, toothy jaws, and reptilian shoulders that could not flap its wings in a modern fashion.

Dozens of studies have focused on the seven known partial skeletons of Archopteryx; dozens of clever scientists have theorized, argued, and theorized yet again about how bird flight evolved. The Solnhofen site has been key because of the extraordinary preservation of its fossils. Not only is there Archopteryx, there are beetles, fish, crabs, dragonflies, jellyfish, crocodiles, plesiosaurs, small dinosaurs, pterodactyls complete with impressions of their skin wings, many kinds of plants, leaves, ferns, and succulents, and much more. Solnhofen is the clearest window we have into the world of 150 million years ago.

Now Sihetun, a site in Liaoning province in northeastern China, is beginning to rival Solnhofen in importance. The Sihetun fossils are older than 120 million years (not as ancient as Solnhofen's fossils) but they have proven as pivotal to the debates.

The Sihetun frenzy began in 1994, when a peasant farmer discovered a beautifully-preserved fossil bird. Called Confuciusornis sanctus - the sacred Confucius-bird - this specimen has small feathered wings, the earliest known beak, and a pygostyle (the reduced bony tail of modern birds). Confuciusornis flew much more adeptly than Archopteryx and, like many modern birds and some dinosaurs, nested in colonies. Male specimens show a pair of elongated tail feathers, fabulous equipment for rial mating displays. But Confuciusornis still has clawed fingers on each feathered "hand" and other archaic features.

In 1996, Sihetun fossils really shook things up in the scientific world. That year saw the discovery of the first "feathered dinosaur" specimen. Sinosauropteryx prima ("first Chinese dinosaur-wing") was a small (one metre long), fast, two-legged theropod dinosaur. Inside the rib cage of one specimen is a pair of eggs, ready for laying; in the stomach of another are the bony remains of a small mammal, its last meal. And there were feather-like fibres or bristles of some kind running along its spine from nose to tail, and probably down its body, arms and legs.

When this find was first announced, scientists who believe birds are simply evolved dinosaurs crowed loudly that the debate was over. If dinosaurs and birds were the only two feathered creatures - out of all of the animals past and present known in the world - there was an ironclad, evolutionary link between them.

Unfortunately, the fibres on Sinosauropteryx lack a quill and the intricate structure of feathers. The skeptics who reject the clustering of birds with dinosaurs were delighted to see their opponents with egg on their faces.

What were these fibres for if they are not true feathers? They may be forerunners of flight feathers or they may have evolved as insulation, to keep the little dinosaur suitably warm or cool.

Thermoregulation is an essential part of the hot-blooded metabolism of birds or mammals. Since flying is an incredibly expensive way to get around the world, it is tempting to think that bird ancestors must have had a fast, hot-blooded metabolism in place before beginning to evolve flight.

If these fibres insulated Sinosauropteryx, they are one more piece of evidence that birds are really evolved dinosaurs. f they are something else - like the deteriorated remains of skin frills, similar to those found on living iguanas - they imply only that birds are descended from reptiles. Finding Sinosauropteryx provoked a lot of squawking but little resolution.

Last year, scientific feathers were ruffled again by another remarkable dinosaur from Sihetun, Protarchopteryx robusta. Its name gives the story away: this fossil looks like a good ancestor for Archopteryx, even though it is not ancient enough to be its actual ancestor.

Protarchopteryx is the size of a turkey, with big legs and true feathers on its hands, body, and tail. The feathers on the arms and hands are too few and too short to be functional wings and they are symmetrical, not aerodynamic, like feathers of flightless birds.

Are these primitive feathers, persisting in bird-like dinosaurs long after Archopteryx had already evolved aerodynamic feathers? Or are they degenerate flight feathers? How can we decide?

The latest fossil find from Sihetun is another dinosaur, an enigma named Caudipteryx zoui (Zou's tail-feather). About the size of Protarchopteryx, Caudipteryx is covered in symmetrical feathers, with a striking fan of plumes arising from its bony tail.

Metaphorically, Caudipteryx is poised between the truly bird-like Archopteryx and the more dinosaur-like Protarchopteryx. It has wings that are too small to propel such a large animal through the air. It also has teeth and a bony tail which is short for a dinosaur, long for a bird.

All of the Sihetun fossils are too recent to be ancestral to Archopteryx and later birds. Still, looking at the fossils from Sihetun is like looking through a kaleidoscope. These fossils blur the hard-and-fast distinctions between birds and dinosaurs; they rearrange the categories in our minds. It is as if all of the elements of birds and dinosaurs were fractured, shuffled, and reassembled according to the throw of a pair of primordial dice.

So far, ancient Sihetun has shown us a true bird with elaborate tail plumes (Confuciusornis), an extremely primitive bird-like dinosaur with feathers but no ability to fly (Protarchopteryx), another even more bird- like feathered dinosaur with useless wings (Caudipteryx), and a small, running dinosaur with fibres-that-might-be-insulation (Sinosauropteryx). What a bewildering array of possibilities.

Do we know yet if birds are simply evolved dinosaurs? No, but the links grow daily more convincing, as more and more dinosaurs with feathers are found.

Do we know if feathers evolved for thermoregulation rather than for flight? No. The oldest feather (on Archopteryx) is unquestionably a flight feather. The Sihetun feathers include flight feathers, feathers like those on flightless birds, and (maybe) down-like insulation.

These finds hint that the function of half a wing or a single feather may have more to do with controlling body temperature than flying, but they do not show us the very beginning of the story.

Do we know if flight evolved from the ground up or the trees down? No, again. But the more we find small, running, earthbound dinosaurs that look like birds, the more probable the "ground up" hypothesis becomes. Sihetun promises a nestful of surprising discoveries but the questions still hatch faster than the answers.

Pat Shipman is an anthropologist at Penn State University. Her latest book, `Taking Wing' (Weidenfeld & Nicolson, pounds 20), is published this month