The problem of bird origins and early avian evolution

J. Ornitho!. I42 Sonderheft 1,139-147 (2001) Deutsche Ornithologen-Gesellschaft/Blackwell Wissenschafts-Verlag, Berlin ISSN 0021-8375 The problem of bird origins and early avian evolution Alan Feduccia Department of Biology, University of North Carolina, Chapel Hill, NC 27 514 USA; Email: feduccia@bio.unc.edu Summary First discovered in 1860, Archaeopteryx has played a prominent role in discussions of avian origin especially during the 20 th century. One of the most dramatic discoveries in modem ornithology involved the revelation of a completely new clade of Mesozoic birds, the enantiornithines, first described in 1981. Among the most spectacular of these early birds is Confuciusornis from the early Cretaceous of China. Enantiornithines are now known from every continent except Antarctica, and from the Lower to Upper Cretaceous. They were the predominant landbirds of the Mesozoic Era, and produced a vast adaptive radiation of varied morphological types. Zusammenfassung Das Problem der Entstehung der V'6gel und ihrer friihen Evolution Nach seiner erstmaligen Entdeckung im Jahr 1860 spielte Archaeopteryx eine herausragende Rolle in der Disknssion um die Entstehung der heutigen V6gel, insbesondere im 20. Jahrhnndert. Eine der aufsehenerregendsten Entdeckungen der modernen Oruithologie war die Entdeckung einer neuen Klasse yon V6geln, den Enantiornithines, die 1981 erstmalig beschfieben wurden. Unter ihnen erregte am meisten Aufsehen Confuciusomis aus der frtihen Kreidezeit Chinas. Enantiornithines sind mit Ausnahme der Antarktis yon allen Kontinenten und aus der gesamten Kreidezeit bekannt. Sie waren die dominierenden Landv6gel des Mesozoikums und zeigten eine enorme Formenaufspaltung. It has been a great privilege to be able to study birds most of my life time. Birds are diurnal and easily studied and are beautiful, their study is what Ernst Mayr has called the scientia amabalis, the beautiful science (Mayr 1963). As a consequence, birds have come under the constant scrutiny of both amateur and professional alike, and because of this birds have contributed disproportionately to myriad fields of biology, including animal behavior and ecology. We need only mention Nobel Laureates Niko Tinbergen and Konrad Lorenz to illustrate this point. Certainly the field of evolutionary biology owes a tremendous debt to the study of birds. Charles Darwin's theory of evolution by natural selection was based to a large extent on the study of birds. The infinite variability of natural populations was vividly illustrated to Darwin by studies of domestic pigeons in England, and the transformation of natural populations by the Gal~pagos mock-thrushes and finches, now commonly termed Darwin's finches. Yet, this careful examination of birds over several hundred years means primarily that the species are very well known, perhaps the best known of any group of vertebrate animals. But, I would claim, as did Erwin Stresemann in one of his late papers in 1959, entitled U.S. Copyright Clearance Center Code Statement: 0021-8375/2001/142[Suppl. 01]-0139 $ 15.00/0

140 Journal ftir Ornithologie 142 Sonderheft i, 2001 "avian systematics and its unsolved problems", that we still know very little about the relationships of the higher categories of birds, and may never understand the intricacies of these relationships. Given all the studies ranging from anatomy to molecular biology that have been brought to bear on the problem of the relationships of modern avian orders, it is astounding that we still understand so very little. I shall later propose a reason for this enigma. I would claim also, in agreement with Walter Bock (1999), that despite what one reads in the popular presses, we still have a very incomplete understanding of the origin of birds, still no solid evidence for a suitable avian ancestor, at the proper geologic time. Archaeopteryx still at the center of the controversy The most famous vertebrate fossil discovered, the late Jurassic Solnhofen Archaeopteryx, is considered by many to be a Rosetta Stone of evolution because it was among the first fossils to link two major groups of living vertebrates, reptiles and birds. First discovered from a single feather in 1860, and later a complete skeleton (the London specimen) in 1861, Archaeopteryx has played a prominent role in discussions of avian origin especially during the 20 th century. However, Huxley's theory of bird origins from dinosaurs dealt more with the similarities of the hind quarters of the chicken and the theropod dinosaur Megalosaurus than with Archaeopteryx. Indeed, the similarity noted by Huxley between birds and dinosaurs related to similarities of secondarily flightless birds, the ratites, and dinosaurs, and this comparison has carried into the present day. Huxley's view was challenged by the turn of the 19 th century with the discovery of pseudosuchian thecodonts, especially Euparkeria, from the late Triassic Karoo of South Africa. Then, in 1926, Gerard Heilmann solidified the "thecodont" origin of birds, along with its corollary, the arboreal, or trees-down theory for the origin of flight, in his well-written book "The Origin of Birds". And, the generally held view that Archaeopteryx was a primitive "bird" held sway until the early t970s when John Ostrom (1973) rekindled the dinosaurian origin of birds, based on overall similarities he had noted in the Cretaceous man-sized theropod Deinonychus, he had discovered in the 1960s. In 1970, while in Europe studying pterosaurs Ostrom discovered another specimen of Archaeoperyx (the Tyler specimen) that had been misidentified as a distinctive species of pterosanr. Strangely, this discovery never led to the pterosaurian origin of birds. Nevertheless, the Ostrom theory of dinosaurian origin of birds was a strange amalgam of overall similarity (old fashioned systematics), physiology and evolutionary scenario, and dealt not only with the theropod origin of birds, but also with hot-blooded (endothermic) dinosaurs needing feathers for insulation, and then a complicated scenario to elongate feathers on the forelimbs as insect swatters. Once elongated, the flight feathers of these endothermic dinosaurs would serve nicely for a ground-up (cursorial) origin of birds. Thus, the dogma of the day (Bakker 1975: 69) was that "In spite of its very birdlike appearance, Archaeopteryx was closely related to certain small dinosaurs and could not fly," and later, "It is my contention thatarchaeopteryx was still learning to fly... from the ground up." (Ostrom 1985: 169). For the new "theropodists", then, Archaeopteryx was and is little more than a feathered theropod, learning to fly from the ground-up; for most ornithologists, Archaeopteryx is a primitive bird, albeit a "bird" in the modern sense. It is interesting here to note that the "new" theropod origin of birds was based entirely on overall similarity and evolutionary scenarios; it was only much later that the supposed bird-theropod nexus was placed into a cladistic framework (Ganthier 1986). Contrarily, the Berlin Archaeopteyx, with outstretched wings, compares beautifully in proportions with modern magpies and coucals, as illustrated in a brilliant article by Heinroth, published in this same journal in 1923 (Fig. 1

A. Feduccia - Early avian evolution 141 and 2). In outline, Archaeopteryx has the elliptical wing of a modem woodland bird, and the flight feathers, both primaries and secondaries, are strongly asymmetric, indicating an aerodynamic function (Feduccia & Tordoff 1979). Amazingly, the single feather, an asymmetric secondary found in 1860, proved a hundred and forty years ago, that Archaeopteryx was a flying bird and not an earth-bound dinosaur! By almost any measure, including, interestingly, wingloading (Norberg 1990), Archaeopteryx fits well within the framework of modem birds. The feathers, down to their microscopic detail, are those of a modern bird, with well developed barbs and barbules. The foot, too, is that of a bird, not a dinosaur, with a well-devel- Fig. 1. Comparison of the Berlin Archaeopteryx to a pheasant cuckoo (plate 4 from Heinroth 1923). Abb. 1. Vergleich des Berlin Archaeopteryx mit einem Fasanenkuckuck (Tafel 4 aus Heinroth 1923). Fig. 2. Comparison of a modem magpie to the pheasant cuckoo (plate 3 from Heinroth 1923). Abb. 2. Vergleich einer modemen Elster mit einem Fasanenkuckuck (Tafet 3 aus Heinroth 1923).

142 Journal ftir Ornithologie 142 Sonderheft l, 2001 oped hallux and highly recurved claws that match up nicely with those of modern perching birds, not the flattened claws of ground-dwelling birds. Too, as Derek Yalden of the University of Manchester has shown, the wing claws of Archaeopteryx are quite similar to those of trunk climbing birds and mammals, and exhibit extreme lateral compression. Likewise, the foot claws are laterally compressed and are quite unlike the hoof-like, broad claws of Compsognathus and other earth-bound dinosaurs. It is simply inconceivable to believe that Archaeopteryx was not a primitive, but perching and flying bird. Therefore, it is astounding that Archaeopteryx has been the focus of all theories relating to the origin of avian flight. However flight evolved, Archaeopteryx is so far removed, structurally and temporally, from that event as to be of little relevance to such discussions. Studies of Opposite Birds Illustrate the Problem One of the most dramatic discoveries in modern ornithology involved the revelation of a completely new clade of Mesozoic birds, the enantiornithines or "opposite birds", so-called because they are distinctive from modern or ornithurine birds, and the tarsometatarsus fuses in the direction opposite that of modern birds, that is, from proximal to distal. Their distinctiveness is illustrated by the presence of a long, fleshy tail, with an elongate pygostyle, a distinctive sternal keel and the fact that the triosseal canal is formed differently from that of modern birds. First described in 1981, the opposite birds are now known from literally dozens of localities, from every continent except Antarctica, and from the Lower to Upper Cretaceous; they were beyond doubt, the predominant landbirds of the Mesozoic Era, and produced a vast adaptive radiation of varied morphological types. Best known perhaps are those from the early Cretaceous of Spain, and somewhat similar forms from the late, early Cretaceous of China, forms such as Cathyornis that have been shown to have a skull extremely similar to that of the classic Bavarian late Jurassic bird Archaeopteryx. Indeed, most evidence now indicates that the opposite birds and Archaeoperyx form a single avian clade, the subclass Sanriuriae, while all other birds (including all modern Tertiary birds) constitute the subclass Ornithurae. It therefore appears that, unexpectedly, there was an early avian dichotomy, as we now have both opposite birds and well-developed ornithurine birds from deposits of early Cretaceous age. This view is in stark contrast to the classical view of modern birds evolving in orthogenic fashion, directly from Archaeopteryx (Feduccia 1999). Among the most spectacular of these early birds is Confuciusornis, which is now known from literally hundreds of complete specimens from the early Cretaceous of China. Regretfully, many specimens have found their way into the fossil black market and are now sold regularly in Europe and the United States. However, given the number of complete specimens, there can be little doubt concerning its morphology. In many ways it is an interesting evolutionary mosaic, with a modern beak and the tail of an opposite bird, by in other ways it is more primitive than Archaeopte~Tx, with a diapsid temporal region and a well-developed postorbital bone. It also shows extremely asymmetric flight feathers, indicate considerable flight capability. Too, unlike the tail of Archaeopteryx, with paired tail feathers emanating from each caudal vertebra, the tail of Confuciusornis sports two elongate, scale-like feathers, devoid of barbs which are easily visible in the tail of Archaeopteryx, Like Archaeopteryx, however, Confuciusornis has the foot of a modern perching bird, devoid of barbs, with a well-developed hallux and exhibits very highly curved claws, similar to those of modern perching birds. I emphasize this fact because it beautifully illustrates the totally different glasses through which the two camps of bird evolution view the origin of birds.

A. Feduccia Early avian evolution 143 To the ornithologist, virtually every detail of the anatomical structure of Confuciusornis indicates an arboreal lifestyle and powered flight. Among these features are long, pointed wings with highly asymmetric primary feathers, many specimens with two extraordinarily long central tail feathers. It has a well-developed hallux, and all foot toes have highly recurved claws. Too, the pelvis is the standard avian, retropubic construction, and short tarsus is similar to that of perching coraciiforrn birds. Yet, the paleontological view, as divined by Padian & Chiappe (1998), and illustrated on the cover of Scientific American, shows Confuciusornis as an awkward earth-bound, theropodlike predator, with arms outstretched as if to trap some prey. The pubis is illustrated as a vertical theropod condition, observed in none of the hundreds of specimens now available. The heavy, flat toes are shown with claws with little curvature, and overall, the illustration can only be compared to that comic view of a turkey that has fallen from a poultry truck and is wandering around in a dazed condition. One can only wonder why such a gross misinterpretation of Confuciusornis was produced, but it becomes clear that the "theropodists" desire is for all early Cretaceous birds to have been terrestrial (Olson 2000), just having evolved from earthbound dinosaurs, of course, from the groundup. Thus, one can only guess that this misinterpretation is produced to reinforce the now faltering cursorial origin of avian flight, considered by almost all biologists to be a biophysical impossibility (Norberg, 1990). Theropod Origin of Birds With the above backdrop, it is interesting now to consider some of the major flaws in the current dogma of the theropod origin of birds, any one of which should raise a major red flag. I can briefly summarize these problems in three parts: 1) the improbability of the cursorial origin of flight; 2) the temporal problem associated with a bird origin from theropods; and 3) the problem of homology of supposed shared, derived anatomical structures or synapomorphies. Flight origins First, let's consider the problems involved in the ground-up origin of flight. One might correctly ask why one would worry about such evolutionary scenarios since the ultimate test for the ancestor of birds must rest with the correct phylogeny and not the mechanism by which flight evolved. However, since the inception of the Ostrom version of the dinosaurian origin of birds in 1973, the cursorial origin of flight has been inextricably linked to the dinosaurian origin of birds. Virtually everything we know about flight (Norberg 1990) indicates that there are two requisites for the origin of flight in vertebrates: small size and high places. Small size is necessary because only in small animals will any flattening of the body or integumentary projections have any adaptive advantage in producing drag and slowing a parachuter or jumper. Interestingly, feathers projecting from the arms or elsewhere will produce drag and slow the fall of a small arboreal organism; and in a running form the same will happen. That is, a running animal will be slowed down by integumentary projections: just the opposite effect of that desired[ High places are necessary to take advantage of the cheap energy provided by gravity. Otherwise an organism is constantly fighting gravity and even if jumping gets an animal into the air, where does the necessary energy come from to maintain it airborn? If one examines vertebrates that have taken to the air, either in the form of parachuters, gliders or active fliers, aside from birds, the list is very long and includes members of all the classes of terrestrial vertebrates: amphibians, reptiles and mammals. In all cases, we see the same general theme, that the organisms are small to modest in size, and in all cases, without exception, volant organisms are also arboreal. Ulla Norberg (1990) has elegantly shown, along with other authors, the high level of improbability of flight origin from the ground-up;

144 Journal ftir Ornithologie 142 Sonderheft 1, 2001 and contrarily, the ease with which flight can originate and has originated from the treesdown. As a final note, the entire bauplan of early (late Triassic) theropods is exactly the opposite of what one would desire for the origin of flight. Theropods such as Herrerasaurus, Syntarsus and Coelophysis are typical, and exhibit the typical theropod body plan, with a laterally compressed body, featuring a large, balancing tail, and forelimbs shortened to half the length of the hindlimbs. There may exist theropods suitable for an arboreal (or cursorial) flight origin from this critical time zone, but if they existed they have yet to be discovered. The temporal paradox The discovery of massive numbers of beautifully preserved vertebrates, especially birds and dinosaurs from the Mesozoic of China has refocused the argument for a theropod origin of birds. However, despite various attempts to push the clock back on these deposits, there no longer appears to be any question that these fossils are from lake beds dating from about 120-124 million years ago, that is, from the early Cretaceous, some 26 to 30 million years after the 150 million year old Solnhofen deposits of the late Jurassic of Germany, that produced the oldest known bird. To put it in perspective, that's about half the time period of the entire Age of Mammals! To account for the presence of so-called "feathered dinosaurs" and supposed avian ancestral prototypes in these deposits, the deposits are considered to be unique, "a refugium for relicts" (Luo 1999), and the fossils are considered to be relict lineages of the Jurassic. Be that as it may, the curious aspect of these arguments is that the most acceptable ancestors of birds among theropods are not the late Triassic forms such as Eoraptor, Herrerasaurus, Syntarsus and Coelophysis, which have no avian features, but rather the dromaeosaurs, a group confined to the Cretaceous, and a group that becomes more and more superficially birdlike as one approaches the latest Cretaceous. Indeed among the avian prototypes now cited as among the most important are Deinonychus, some 30 million years beyond Archaeopteryx, Unenlagia, some 60 million years beyond Archaeopteryx, and Velociraptor and Bambiraptor (the most birdlike), from the latest Cretaceous, some 75-80 million years beyond Archaeopteryx. Given the fact that Archaeopteryx was an already well-developed bird, in the modem sense, by 150 million years ago, it would seem to be an incredible stretch of credulity to find the most birdlike dromaeosaur Bambiraptor staring at the meteor at the close of the Cretaceous. Indeed, this type of temporal paradox is generally associated with the well-known phenomenon of convergent evolution, an insidious trap that lies baited and waiting for the unsuspecting paleontologist. Perhaps, as the paleontologists tell us, there exist ghost lineages of these superficially birdlike dromaeosaurs yet to be discovered in the mid-to-early Jurassic, but if they exist they certainly have not been uncovered to date. The ancestors of birds will some day be discovered, not in the late Cretaceous, but some 150 or so million years earlier, somewhere in the Triassic or early Jurassic Periods. The character mismatch Despite the red flags associated with the scenario of a ground-up origin of flight and the fact that the time line for a theropod origin of birds is lawed, a theropod origin will be validated or invalidated based on the veracity of anatomical comparisons. Here again, we see numerous inconsistencies and major problems. I noted earlier the superficial similarities of the hindlimbs of ratites and chickens to theropod dinosaurs and how this comparison impressed Thomas Huxley, but these comparisons still hold today, and it is among the Cretaceous dromaeosaurs that the comparison is perhaps the best. However, the dromaeosaur skull is devoid of any striking similarities to birds (Elzanowski 1999), and it is among the Cretaceous oviraptisaurs that the skull comparisons are most favorable.

A. Feduccia. Early avian evolution 145 By this method of "cherry-picking'characters from among all of the theropods, then, a ctadistic argument has been made for a dinosaurian origin of birds. However, as Professor Larry Martin of the University of Kansas, frequently notes, "every phylogeny begins with a single character: let's have one". When we examine characters one by one, problems arise with virtually every one; everywhere we turn we are being asked to place a round peg into a square hole. Let's consider several important features. As noted earlier there is little similarity between the avian skull, particularly the jaws and palate, and that of dromaeosaurs, and particularly disturbing is the mismatch of tooth morphology and striking differences in tooth replacement. Professor Larry Martin has demonstrated the complicated differences in tooth replacement between birds and dinosaurs (Martin & Stewart 1999). However, we are all familiar with the typical theropod tooth, characterized by museum favorites such as Allosaurus and Tyrannosaurus. It is a laterally flattened, recurved tooth, devoid of an expanded root crown, and with serrations running up and down the front and back sides. Contrarily, avian teeth, exemplified by Archaeopteryx and Hesperornis, are quite unlike those of theropods and show a striking resemblance to those of crocodilians, with a flattened, unserrated crown that becomes constricted as it approaches the crown/root juncture. The tooth narrows and then expands into a enlarged cement-covered root. To simplify, they are peglike, nonserrated teeth, with expanded root crowns. Several authors in attempting to derive the avian tooth from that of theropods have argued that such a transition could take place through a paedomorphic process, but when we view the juvenile late Cretaceous putative bird ancestor Bambiraptor we find a typical theropod tooth, with well developed serrations. One could go on and on about the mismatch of characters between birds and theropods, but one feature that has almost universally linked birds and theropods is that both groups have a hand reduced to three digits, that is, a tridactyl hand. Thus, the question is: what is the identity of the three fingers? To be brief, there is no dispute that the theropod is a grasping, raking hand composed of digits I-1I-III, the identity being confirmed by the presence of vestigial digits IVand Vin the well-preserved late Triassic basal theropod Herrerasaurus. It has been shown repeatedly by studies of embryology from a variety of authors using different approaches that the hand of birds is composed of digits II-III-IV. Among the latest demonstrations are studies by Richard Hinchliffe (1985), and later by Ann Burke and myself (Burke & Feduccia 1997, Hinchliffe 1997). In our study we show that all anmiotes have a very stereotyped pattern of cartilage condensation. A primary axis of cartilage condensation is prominent and runs through the humerus to digit IV. When compared to serially homologous elements of the hindlimb, one can confirm that the retained digits of the avian hand are II-III-IV. Gerd Mtiller and Pere Albrech reached the same conclusion in 1990 (160): "the developmental evidence overwhelmingly supports the II-III-IV theory of the wing skeleton in birds". To continue, if the bird hand is II-III-IV and the theropod hand is I-II-III, then it follows that the semilunate carpal thought to be a critical link between theropods and birds simply cannot under any circumstances be homologous! The concept of dismantling the entire developmental program of the theropod hand to produce an avian hand with a different set of digits, involving the disappearance of digit I and the reappearance of digit IV is truly a stretch of credulity. Finally, the concept of feathered dinosaurs has been dismantled piece by piece, and the sole remaining feathered dinosaur, Caudipteryx, has now been shown to be a secondarily flightless bird, a "Mesozoic kiwi" (Jones et al. 2000). To seek the appropriate ancestor of birds we must go back into the earlier sediments of the Jurassic and Triassic Periods, and find archosaurs that do not possess the anatomical bag-

146 Joumal fiir Omithologie 142 Sonderheft 1, 2001 gage of Cretaceous theropods. Would such an animal be a dinosaur? I suppose that all depends on the definition of "dinosaur", but certainly it would involve an archosaur that is not committed to a I-II-III hand and a bauplan that involves foreshortened forelimbs and a large balancing tail with a see-saw theropod pelvic region. It would most likely be a basal archosaur, a small, arboreal quadruped, capable of evolving flight from the trees-down. As Walter Bock (1999: 568) recently noted, "it is best to consider birds as part of the great archosaurian radiation without being more specific, as has been agreed by zoologists for more than a century." Back to the thecodonts! Explosive evolution of Tertiary birds Finally, according to the "theropodists" the dinosaurs did not become extinct at the K-T boundary, but live today as the nine thousand or so species of living birds. In a sense, then, as we have seen, the dinosaurian origin of birds is a delusional fantasy by which one can study dinosaurs at the backyard birdfeeder. That 95 % of all vertebrates became extinct from the K-T cataclysm, while birds, the most sensitive of the earth's environmental indicators, representing the vertible "Miner's Canary", passed through such an event unscathed, is truly delusional. Many paleontologists still deny a sudden extinction event for birds (Padian& Chiappe 1998), and a number of molecular studies claim to support an origin of the major orders deep within the Cretaceous, at a time when the continents split (Hedges et al. 1996, Cooper & Penny 1997), but their techniques have serious faults. New estimates indicate that the bolide impact may have struck earth with an impact equal to some 10,000 times the force of the current entire current world's nuclear arsenal, whatever that is. Nevertheless, it is almost astounding that any birds made it through the K-T cataclysm. One thing is certain, given the literally dozens of localities, nearly worldwide that have produced late Cretaceous enantiornithine landbirds, as well as ocean deposits that have produced ichthyornitids and hesperornithiforms, the absence of modern avian orders from all of these deposits, if present, would be phenomenal. From the Ichthyornis localities, where are the pelecaniforms, procellariiforms, and charadriiforms (gulls and terns), if present? And, why are there no enantionithines, ichthyorniforms, or hesperornithifoms beyond the K-T boundary. The answer is becoming clear: there was a huge avian extinction event that coincided with the extinction of the dinosaurs. Then, by the Lower Eocene (54 million years) we see all of the major orders of modern birds present. In terms of numbers, that means that there is some 15 million years during which all of the major orders of modern birds evolved. This can only be described as an explosive evolutionary event, and one that paralleled a similarly explosive evolutionary pattern in mammals (Feduccia 1995). If this hypothesis proves correct,,and the divergence of most modern avian orders is tightly clustered within a time period of some 5 to 10 million years, then an explanation would be provided for Stresemann's lamentation of 1959, that we may never fully understand the interrelationships of the major orders. References Bakker, R.T. (1975): Dinosaur renaissance. Scientific American 232(4): 58-78. Bock. W.J. (1999): Review of, The Mistaken Extinction. Dinosaur Evolution and the Origin of Birds. By L. Dingus and T. Rowe. Auk 166: 566-568. Burke, A. C. & Feduccia, A. (1997): Developmental patterns and the identification of homologies in the avianhand. Science 278: 666-668. Cooper, A. & Penny, D. (t997): Mass survival of birds across the Cretaceous-Tertiary boundary: molecular evidence. Science 275:1109-1113. Elzanowski, A. 1999. A comparison of the jaw skeleton in theropods and birds, with a description of the palate in the Oviraptotridae. Smithsonian Contributions to Paleobiology 89:311-323. Feduccia, A. & Tordoff, H.B. (1979): Feathers of Archaeopteryx indicate aerodynamic function. Science 203: 1021-1022.

A. Feduccia. Early avian evolution 147 Feduccia, A. (1995): Explosive evolution in Tertiary birds and mammals. Science 267: 637-638. Feduccia, A. (1999): The Origin and Evolution of Birds. 2 nd ed., New Haven. Gauthier, J.A. (1986): Saurischian monophyly and the origin of birds. Memoires of the California Academy of Sciences 8: 1-55. Hedges, S. B., Parker, R H., Sibley, C. G. & Kumar, S. (1996): Continental breakup and the ordinal diversification of birds and mammals. Nature 381: 226-229. Heinroth, O. (1923): Die Fltigel yon Archaeopteryx. J. Ornithol. 71: 277-283. Hinchliffe, J.R., (1985): "One, two three" or "two, three, four": an embryologist's view of the homologies of the avian digits and carpus of modern birds. In: Hecht, M. K., Ostrom, J. H., Viohl, G. & Wellnhofer, R (Eds.): The beginnings of birds:!41-147. Eichst~itt. Hinchliffe, J.R. (1997): The forward march of the bird-dinosaurs halted. Science 278: 596-597. Jones, T. D., Farlow, J. O., Ruben, J. A., Henderson, D.M. & Hillenius, W.J. (2000): Cursoriality in bipedal archosaurs. Nature 406:716-718. Lou, Z. (1999): A refugium for relicts. Nature 400: 23-24. Martin, L.D. & Stewart, J.D. (1999): Implantation and replacement of bird teeth. Smithsonian Contributions to Paleobiology 89: 295-300. Mayr, E. (1963): The role of ornithological research in biology. Proceed. XIII International Ornithological Congress, 27-38. Mfiller, G. B. & Alberch, R (1990): Ontogeny of the limb skeleton in Alligator mississippiensis: developmental invariance and change in the evolution of archosaur limbs. J. Morphol. 203:151-164. Norberg, U. M. (1990): Vertebrate flight. Berlin. Olson, S. L. (2000): Review of: Chiappe, L. M., et al. 1999. Anatomy and systematics of the Confuciusornithidae (Theropoda:Aves) from the Late Mesozoic of northeastern China. Auk 117: 836-839. Ostrom, J.H. (1973): The ancestry of birds. Nature 24: 136. Ostrom, J.H. (1985): The meaning of Archaeopteryx. In: Hecht, M.K., Ostrom, J.H., Viohl, G. & Wellnhofer, R (Eds.): The beginnings of birds: 161-176. EichstStt. Padian, K. & Chiappe, L. M. (1998): On the origin of birds and their flight. Scientific American 278(2): 38-47. Stresemann, E. (1959): The status of avian systematics and its unsolved problems. Auk 76: 269-280.