Feathers, scutes and the origin of birds

Jeff Poling

Where feathers came from is one of the greatest mysteries of science today. Many theories have been advanced over the years, but few, if any, have even come close to being proved. Intimately linked to this mystery is the origin of the birds themselves: whether they share a common ancestor with dinosaurs or are dinosaurs. Recent experiments by biologists studying webbed toes may have inadvertently provided an important clue.

The feet of tetrapods are inherently webbed. At a certain point in the development of a fetus, the tissue between the digits recedes, leaving the toes free. The exception to this is the webbed feet of ducks.

Drs. Hongyan Zou and Lee Niswander wondered what biological process allowed duck feet to remain webbed. Through experimentation they learned that the absence of certain proteins in the webbing of ducks allowed the webbing to remain throughout the fetus' development and entire life-cycle. The experimentation was carried out on the fetuses of cute, furry or feathered animals. At various stages of development, fetuses were injected with a virus that blocked development of a specific set of proteins in one of their limbs. In chicken embryos, the webbing of the toes were not absorbed and webbed feet were retained.

The lack of the proteins also caused the scutes on the foot to develop into feathers.

Scutes are the thick scales on the top of a bird's foot (see figure at right). There are smaller scutes on the back of the foot, called scutellae, and scales on the bottom of the digits, called reticulae. Analyses by Alan Brush have shown that bird scutes, scuttelae, claw sheathes, beak sheathes, and scales around the eyes are of the same chemical composition as feathers, and are controlled by the same genes. The reticulae have been shown to be identical to crocodilian scales both in composition and their location on the DNA strand.

In all cases where a chick was infected with the inhibitor virus at days 15-18 of development, at least some of the scutes developed into feathers. The feather development ranged from thickening of the edge of the scute, to short, fat feathers, to long, thin feather filaments (see figures at left and right; click on the images to see larger hi-res picture). These feathers contained the barbs characteristic of normal feathers, although the barbs were more numerous. The scutellae also developed into feathers to various degrees.

The reticulae did not develop into feathers. This is not surprising given Brush's research showing the reptilian nature of these scales. Other research has shown that reptilian scales do contain, in small amounts, the same protein family as avian feathers and experiments by Dhouailly, Hardy and Sengel (1980) were able to convert reticulae into feathers after treatment with retinoic acid. This conversion, however, was accomplished with the lowest frequency of the three types of scales.

It has long been believed that dinosaurs and birds are closely related. As the fossil evidence has accumulated, many, if not most, dinosaur paleontologists (dinosaurologists) consider birds to actually be dinosaurs. The main monkey wrench in this theory was the question of where the feathers of birds came from. As reptiles, dinosaurs probably had scales. Skin impressions of hadrosaurs and tyrannosaurs bear this out. Many researchers hypothesized that feathers developed from reptilian scales. However, the research of Alan Brush, as mentioned above, clearly shows that feathers and reptilian scales are chemically and genetically different. The results of the Zou and Niswander experiments failing to turn reticulae into scales bears this out, Dhouailly et al's results not withstanding. It is doubtful that feathers evolved from reptilian scales.

But what of the second type of avian scale, the scute? Crocodiles, a sister group of dinosaurs, have scutes, and they show a similar, though not exact, chemical composition with bird scutes. Dinosaurs had scutes, too. Paleontologists would give their eye teeth to know the chemical composition of dinosaur scutes, because a logical question given Brush's research is "did feathers evolve from scutes?"

Brush's research certainly suggests this may be possible. However, Zou and Niswander's research suggests a different conclusion. It does this by raising an even more intriguing question: "did scutes evolve from feathers?"

The results of these experiments suggest this certainly may be the case. Generally, when a new characteristic develops from another, certain proteins, chemicals or genetic signals are required to express this new character and suppress the old one. When these proteins, chemicals or genetic signals are blocked, the old characteristic will often express itself. In birds this has already been demonstrated by growing bird fetuses with reptilian tails and teeth just like those of early birds such as Archaeopteryx. When Zou and Niswander blocked certain proteins in embryonic chicks, feathers developed instead of scutes, strongly suggesting that feathers are the primitive characteristic. Dhouailly et al's results of also converting reticulae were probably more an artifact of the acid treatment than genetics, given the difficulty of the conversion.

The results of all these experiments have widespread implications for the dinosaur-bird debate. For years one of the strongest objections to avian dinosaur ancestry was that feathers could not and did not develop from reptilian scales, and thus the scaly dinosaurs could not be their progenitors. However, dinosaurs had scutes as well as scales, and if bird scutes developed from feathers, could not dinosaur scutes have developed from feathers as well?

If dinosaur scutes developed from feathers then feathers are a primitive characteristic of the dinosauria as a whole, as most of the lineages of dinosaurs had scutes. While later, larger dinosaurs such as hadrosaurs and tyrannosaurs could have lost their feathers, much as hippos and elephants have lost most of their hair, the earliest dinosaurs would have had to have feathers. These feathers probably would not resemble the specialized flight feathers of birds, but could resemble non-flight feathers such as down.

Recent fossil discoveries support the conclusion that feathers are a primitive feature for the dinosauria. Fossil skin impressions of the new ornithomimid dinosaur Pelicanimimus suggest this dinosaur had hairlike integument similar to Kiwi feathers. Dr. Phil Currie of the Tyrrell museum in Canada has examined a recent Chinese discovery named Sinosauropteryx, a specimen he describes as a compsognathid dinosaur skeleton with clear feather impressions in the rock. These impressions have been described as looking like a bird's down feathers. No formal descriptions have been published for either dinosaur yet, so no formal conclusions can be drawn. If these dinosaurs' integuments are indeed feathers, it is the strongest evidence yet of feathers in at least the theropod branch of the dinosauria.

However, scutes could have evolved from feathers before the dinosaurs evolved, thus making feathers primitive not only for the Dinosauria, but the entire Archosauria (the group formed from the common ancestor of the Dinosauria, Pterosauria and Crocodylia). It has been suspected for years that pterosaurs had "fur," and recent discoveries may prove it beyond reasonable doubt. The thecodont Longisquama, which is a good candidate for the archosaurian common ancestor, had some sort of scale structure that some have interpreted as protofeathers.

Unfortunately, this evidence can also support the theory of many ornithologists that birds share a common ancestor with the dinosaurs, rather than descend directly from the dinosaurs (thus making the birds the fourth group of the archosauria). If feathers are primitive for the group as a whole, there is no inherent reason to think birds had to descend from dinosaurs.

The debate on the subject of bird origins will continue, with ornithologists and dinosaurologists continuing to argue the significance of shared characters. However, this debate will include one new factor, the similarity of dinosaur feathers to bird feathers. Should the feathers of Sinosauropteryx prove to be very close to the down of birds, it may at last bring the doubters of the dinosaurian ancestry of birds into the fold.

The experiments of Zou and Niswander, and Alan Brush, suggest that scutes evolved from feathers. Although the research does not provide hints as to the origin of feathers, it does remove the impediment to the dinosaur-bird theory by showing that while feathers probably did not evolve from scales, scutes, a character shared by dinosaurs, may have evolved from feathers. Recent finds suggest, if not outright prove, that dinosaurs had feathers. Whether feathers are primitive characteristics of the archosaurs is a question that will continue to fuel the debate of whether birds are dinosaurs or a sister group to the dinosaurs. However, the results of the research of Brush and Zou and Niswander, and new finds such as the feathered dinosaur Sinosauropteryx, help strengthen the relationship between dinosaurs and birds.

1. Brush, Alan. bird scales. Public e-mail, Dinosaur Mailing List, 1995.
2. Brush, Alan. croc s&s. Public e-mail, Dinosaur Mailing List, 1995.
3. Brush, Alan. scales-feathers. Public e-mail, Dinosaur Mailing List, 1995.
4. Niswander, Lee. Re: Repercussions of the Science article on scale formation. Private e-mail, 1996.
5. Zou, Hongyan and Lee Niswander. 1996. Requirement for BMP Signaling in Interdigital Apoptosis and Scale Formation. Science 272: 738-741