American Scientist

Lead Review (Full Text)

March-April 1997

The Continuing Debate Over Avian Origins

The Origin and Evolution of Birds. Alan Feduccia. xii + 420 pp. Yale University Press, 1996. $55.

The popular press seems perpetually confused by the intense debate over how birds and their flight evolved. Did birds originate within small carnivorous dinosaurs or from an as yet unknown, more distant reptilian group? Did they climb trees and glide on the way to flight, or did they run and jump from the ground, extending their leaps by flapping their incipiently feathered forelimbs? Is there any real evidence bearing on these questions, or are they just matters of acrimonious debate among scientists, each with part of the puzzle? Is any progress being made?

In fact, this problem is not different from most others in science, and the solution is the same. Science depends not on authoritarianism, not on hegemony of single fields, and not on majority votes, but on method. Progress is made not only as more evidence accumulates, but also as methods of analyzing evidence are developed, tested and standardized. The difficulty of the present debate is that the methods used by the rest of the community are not used by a few who still object not only to those methods but to their results. In their place, however, the dissenters offer no methods and no opposing theory that can be tested; their theory amounts to an objection to the accepted theory. So it is very difficult for the rest of the community to evaluate or test these objections. But to those not familiar with the evidence and methods, the whole issue can appear to be only a pointless squabble.

In The Age of Birds (Harvard University Press, 1980), Alan Feduccia first laid out his arguments that birds did not descend from theropod dinosaurs, as commonly accepted, and that flight must have evolved from the trees down. In The Origin and Evolution of Birds Feduccia fights hard against accepted relationships and the methodologies used to reach them-now generally preferred by the systematic community, including reviewers for the systematic division of the National Science Foundation. How well he succeeds may depend on how much the individual reader knows about the issues and methods to begin with.

The book can be seen in three parts: the origin of birds and their flight, the Cretaceous bird radiation and the evolution of the living groups of birds in the Tertiary. The first topic is the most controversial and has attracted the most notice for the book. However, the treatment of virtually all topics in the book eventually recurs to the question of the use of scientific method.

John O'Neill's rendering of a small accipitrid hawk, Horusornis vianeyliaudae (late Eocene of France and early Oligocene of North America). From The Origin and Evolution of Birds.

The success of phylogenetic systematics (cladistics) applied to nearly all branches of life in the past two decades has been revolutionary, although not infallible. It has provided a universal, testable method for establishing relationships among organisms, and a framework for anchoring and testing other evolutionary questions, including those of adaptation, ecology, behavior and biochemical change. Feduccia dismisses all these results because he cannot accept the theory that birds descended from theropod dinosaurs. Why? Because he knows that flight must have begun in trees, and so the first birds were arboreal, unlike the terrestrial predatory dinosaurs of the Mesozoic Era. How does he know this? Well, most birds today live in trees, and they are adapted for flight. And theoretically it is easier to evolve flight if you can already glide from an arboreal perch.

In this and many other issues in the book, Feduccia conflates separate problems and seldom explains the evidence that is not on his side. The phylogenetic origin of birds, for example, is a different question from how they evolved flight. Feduccia cites mathematical studies that explain (unsurprisingly) why it is easier to gain lift and airspeed if you jump from a tree rather than leap from the ground. He quotes prominent bioaerodynamicist J. M. V. Rayner, casting doubt on the origin of flight from a running, leaping model because the energetic demands on both forelimb and hindlimb systems would have been high.

But models do not tell us what happened in any given case, and nature sometimes has a way of not following the easiest path. Rayner later noted that the terrestrial model would have been historically more likely because the avian ancestors were terrestrial, there is no real evidence that Archaeopteryx was a climber, and there were no trees in the vicinity of the Solnhofen lagoon where the first known birds are preserved. Phylogenetic, paleontological and paleoecological evidence, then, supports the view that birds evolved from small, terrestrial dinosaurs. Most small four-legged animals can get into trees (goats come to mind), but plausibility is not evidence for arboreal habits. Feduccia did a study purporting to show that the claws of Archaeopteryx were curved like those of arboreal birds, but studies by Ostrom, Stephan, and Peters and Görgner disagreed. (In any case, Tyrannosaurus claws are similarly curved; this is common to carnivorous dinosaurs.) Feduccia discusses none of those studies and surprisingly provides no evidence to support the arboreal theory; it is simply assumed, and most of that part of his chapter is about feathers.

Why not accept that birds evolved from dinosaurs? Beginning in the preface, Feduccia makes it clear that this is impossible, and his impatience and rage with those who disagree with him leap constantly from the page. As I read it, Feduccia is working against two disadvantages: He is not familiar with the paleontology of reptiles or any animals other than Tertiary birds, and he cannot accept cladistic analyses because they keep coming up with birds emerging from theropods. For Feduccia, cladistic analyses must be methodologically flawed because the unique anatomical features that unite birds with theropods must be de facto convergences, primitive states or not homologous. He does not explain to his audience that cladistic analysis is the only method currently known that explicitly deduces probable convergences, primitive states and nonhomologous characteristics, and he offers no method in its place. The closest he comes, in several places, is to rely on logically circular "pseudophylogenies" based on the presumed adaptive value of various features. But one is left to imagine transitional animals that had these traits.

Instead, Feduccia seizes the biologically and taxonomically vacuous term "avimorph thecodonts" (supposed archosaurs with one or more apparently birdlike features) as prime candidates for bird origins. Of the disparate Triassic animals he mentions, Cosesaurus is an aquatic prolacertiform, not even an archosaur; Longisquama is so poorly preserved that it cannot even be classified confidently within the diapsid reptiles; Scleromochlus has emerged from four separate analyses as the most likely sister group to the flying pterosaurs; and Megalancosaurus is an aquatic protorosaur that would require the reversal of 135 shared evolutionary features (according to data available even 10 years ago) to place it near Archaeopteryx and the birds. Recent cladistic analyses of the relationships of archosaurs have progressively separated these long-problematic reptiles into those closer to crocodiles on the one hand and birds on the other. Feduccia regards this work as "tremendous confusion," but it appears that the only confusion is his. Unfortunately, his readers are neither given the evidence necessary to judge for themselves nor referred to the abundant literature that would help.

Feduccia attempts to paint the controversy about bird origins as "paleontological" versus "ornithological" views, but this is misleading. It is simply a controversy between workers who use the methods of comparative biology and those who do not. Since the "theropod origin" was proposed in its modern version in the early 1970s, only a half dozen workers have continued to oppose it, and all of them except Feduccia are paleontologists. Furthermore, none of them has ever produced a cladistic phylogeny using all the available evidence, and they do not seem to agree about what animals, if any, are closer to birds than are theropods. In contrast, every cladogram done by workers not prejudiced against theropod origins has located birds squarely among theropods; but perhaps more significantly, this result has also been reached by other workers who do not use cladistic analysis, so is it likely that the method itself is at fault? Feduccia's contentious and repetitive attacks on cladistics and cladists never bother to get at the ten dozen or so (by some analyses, 200 or more) shared derived characteristics that support this placement. Can they all be convergences, primitive states or mistaken homologies?

Feduccia's animadversion to cladistics and cladists extends throughout the book to the detriment of his audience, who might have been expected to judge the full evidence for themselves. The most abundant and widespread group of Cretaceous birds now known is the Enantiornithes ("opposite birds"), but they were not discovered until 1981. Feduccia omits all discussion of their interrelationships, notably the cladistically based work done by their foremost student, Luis Chiappe. He also omits discussion of cladistic analyses of the systematic position of the Enantiornithes, as well as of the Early Cretaceous Spanish and Chinese birds that have appeared in abundance in the past few years, among other birds and archosaurs. He will not accept that the flightless Mononykus is a Cretaceous bird because it did not fly. Curiously, he groups all flightless birds in a single chapter, but apart from the unremarkable generalizations that flightlessness was independently derived, and that it might be linked in some groups to large size and developmental heterochrony, there is little to offer here. Ironically, phylogenetic analysis of each taxon and its closest flying relatives might have helped to show what mechanisms operated in each case.

For similar lack of methodology, Feduccia sheds little light on whether ratites are monophyletic (a natural group); he presents only confusion about the paleognathous palate that differentiates them from the other, neognathous birds, because he cannot accept that this is a primitive state also shared by dinosaurs (as are the slender scapulocoracoid, the furcula, the three-toed foot with reversed hallux, the semilunate wrist bones and the homologous first three fingers of the hand). Readers may wonder why fossils with some characters of two living bird groups are often advanced as "intermediates" between them--with no sense of which group may have evolved from which, whether the fossil in question is a common ancestor or perhaps just a chimaera of convergence. For Feduccia, there is simply no way to decide these questions, because he does not subscribe to the comparative method.

Feduccia is at his best tracing the fossil records of living bird groups as far back as they will go, and he is sensible about not accepting fragmentary material or isolated bones as reliable evidence: Most individual bird bones are not very diagnostic taxonomically. Many of the illustrations are nice but are undermined by the off-white pages that dull their contrast. Readers with an interest in bird groups will enjoy looking up the fossil records of their favorites, but evolutionary biologists will have to look elsewhere for current understanding of the origin and evolution of birds.--Kevin Padian, Integrative Biology and Museum of Paleontology, University of California, Berkeley

©American Scientist 1997

Home Page | Subscriptions | Permission Forms | Email to the Editors