Birds of a Feather: Downy Dinos and the Origins of Aves
When science historians look back at the late 20th Century, they will find a legacy of stunning discoveries and revelations, and no field will yield as breathtaking an assemblage of revolutionary finds, as that of dinosaur paleontology. In the past thirty years we have experienced a renaissance in our understanding of the paleobiology and phylogeny of the most marvelous animals known (Bakker 1974, 1986). Gone forever are the days in which Dinosauria was viewed as a Byzantine taxon, fated to extinction at the hands of the ever-clever Mammalia. A correlate of this enlightenment is perhaps the most sensational discovery in the entire history of dinosaur science: feathered theropods, recovered from the Yixian lagerstatten of China.
Most recently predicted by Ostrom (1973), feathered dinosaurs are prima facie evidence for the theropod origin of birds, and their discovery marks the death-knell of the “thecodont” ancestry argument (e.g., Feduccia 1996). Whenever so momentous a discovery is made, with profound phylogenetic implications, it is of course expected and desired that skepticism and critical review be voiced. Thus, almost as soon as the first of these so-called “downy dinos” was reported (Ji & Ji 1996), the validity of this interpretation of the integumentary structures preserved in these fossils as feathers, or feather antecedents, has been challenged. The debate, like all things regarding the origin and phylogeny of Aves, is sharply demarcated, with the adherents of theropod ancestry to one side, and those advocates of “thecodont” ancestry to the other.
While proponents of the “thecodont” hypothesis have largely dismissed the evidence on the dogmatic a priori assertion that birds are not dinosaurs, it is nonetheless pertinent to review the data concerning the integument of allegedly feathered dinosaurs, and examine the alternative arguments put forth by some researchers (e.g., Geist et al 1997).
There are to date six theropod taxa, which display integumentary derivates regarded as feather homologues and/or antecedents: Sinosauropteryx prima, Protarchaeopteryx robusta, Caudipteryx zoui, Beipiaosaurus inexpectus, Sinornithosaurus millenii, and Microraptor gui (Ji & Ji 1996, Ji et al 1998, Xu, Tang & Wang 1999, Xu, Wang, & Wu 1999, Xu, Zhou, & Wang 2000). The alvarezsaur Shuvuuia deserti (Chiappe, Norell, & Clark 1998) appears to display simple tubular integumentary structures, which according to biochemical analysis are apparently related to feathers (Schweitzer et al 1999). However, as the exact phylogenetic status of Alvarezsauria remains uncertain, and they may in fact be basal Aves (sensu Chiappe 1995), these data have little bearing on the discussion of feather-like integumentary derivatives in Maniraptoriformes (contra Paul 2002).
In what is without question an era of discovery in dinosaur science unmatched since the great bone wars of E. D. Cope and O. C. Marsh, it is not surprising that the pace of published reports often lags behind the work in the field. Within the past two years, spectacular new finds have bolstered the interpretation, offered herein, that some theropods displayed feathery integument. Two new species of small, highly arboreal theropods from the Upper Cretaceous of Liaoning Provice, Scansoripteryx heilmanii and Epidendrosaurus ninchengensis, both display filamentous integumentary derivatives similar to those observed in Sinornithosaurus milleni (Zhou et al 2002). While these taxa are still in need of further description and phylogenetic analysis, and they may indeed be synonymous, they are unquestionably theropod, and offer insight on the very topic matter of this essay.
Feathered Theropods vs. the “Pilt-down” Dinosaur
Of the taxa enumerated, none have received as much attention in this debate as has Sinosauropteryx. As the original “downy dino,” this is only to be expected. Almost as soon as the pictures of the holotype NGMC 2123 were displayed at the 1996 meeting of the Society of Vertebrate Paleontology, objections were voiced to the conclusion that the preserved integumentary structures were feather homologues/pre-cursors. In 1997 an expedition organized by the Philadelphia Academy of Natural Sciences, and consisting of a veritable “who’s who” in avian phylogenetics (John Ostrom, Peter Wellnhofer, Larry Martin, and Alan Brush), traveled with much press-fanfare to China to examine the Sinosauropteryx material. Conclusions were, predictably, mixed.
There has emerged a fundamental dichotomy in the interpretation of the integumentary structures of Sinosauropteryx and indeed all of the feathered Maniraptoriformes. There is a general consensus that in Sinosauropteryx at least, we have evidence of a pelage of basal feather homologues (Ji & Ji 1996, Chen et al 1998, Padian et al in Tanke & Carpenter 2001, Paul 2002). Several researchers have argued that rather than feather homologues, the integumentary derivates of Sinosauropteryx represent dermal collagenous fibers and thus, do not constitute evidence for the presence of feathers of any form in Theropoda (Geist et al 1997, Feduccia 1999, Ruben & Jones 2000). Alan Feduccia, among the more prominent advocates of “thecodont” ancestry, infamously labeled Sinosauropteryx the “Pilt-down dinosaur” (in Morell 1997).
The question remains, which interpretation is supported by the morphology of the integumentary derivatives seen in Sinosauropteryx? Detailed description of these structures can be found in Padian et al (in Tanke & Carpenter 2001) and Witmer (in Chiappe and Witmer 2002), and the reader is referred to these papers. A brief summation of the work presented therein, will be presented in this essay.
The integumentary structures of Sinosauropteryx prima are distributed along the midline of the body, extending from the crown of the skull caudally to the tail, and at scattered locations throughout the rest of the body including the abdomen, the knee, and adjacent to the humerus and ulna (Chen et al 1998). Thus, the conclusion that the structures were strictly limited in distribution to the body midline is falsified (contra Geist et al 1997).
The structures are filamentous, and densely distributed over the body surface. Padian et al (in Tanke & Carpenter 2001) report a range of length for the filaments of 8-40mm, although most of the filaments range from 8-10mm. Approximately 10 filaments are preserved per millimeter, providing an indication of how dense the pelage actually is (contra Feduccia 1999). Compare Plate 1A and 1C in Padian et al (in Tanke & Carpenter 2001) to the figure of dermal collagenous fibers in Varanus (Feduccia 1999, 378). The filaments appear to lack a central rachis, and may or may not be hollow. The preservation of the structures is not sufficient to further clarify their structure, and biochemical testing is infeasible.
From the available data, the suggestion that these fibers are dermal collagenous structures, similar to those seen in aquatic or semi-aquatic reptiles such as some species of Varanus, must be regarded as wanting of substantiation (contra Geist et al 1997, Feduccia 1999, Ruben & Jones 2000). First and foremost, the distribution of the fibers and the sheer density with which they appear, are difficult to reconcile with the argument that they represent collagen fibers. A comparison between Plate 1A and 1C in Padian et al (in Tanke & Carpenter 2001) and the photograph from Geist et al (1997) presented in Feduccia (1999, 378) of a skinned varanid tail, shows striking differences. Combined with the lack of any evidence for an aquatic habitus for Sinosauropteryx or any other compsognathid theropod, the arguments presented by Geist and his colleagues, as well as Feduccia and Ruben, are without merit. No compelling data has yet been advanced to conclusively refute the feather-homologue status of the integumentary derivatives seen in NGMC 2123 and all other Sinosauropteryx material.
What of the other Yixian theropods? As regards integumentary derivative morphology, the filaments seen in Beipiaosaurus inexpectus and Sinornithosaurus millenii are most similar to those observed in the type and referred material of Sinosauropteryx. No coherent argument has been presented whereby these structures are relegated to mere analogues, and indeed, there is no reason to suspect that this is the case. It is impossible to argue credibly that these structures represent a median frill of dermal structures, as they are preserved either on the antebrachial region (in Beipiaosaurus), or uniformly over the body surface (in Sinornithosaurus).
In Protarchaeopteryx robusta and Caudipteryx zoui, symmetric remiges and rectrices are preserved, in addition to plumulaceous feathers associated with the abdomen (Ji et al 1998, Padian et al in Tanke & Carpenter 2001, Paul 2002). These structures are unquestionably feathers, and thus, unable to call their homology into question, proponents of “thecodont” ancestry have instead argued that these taxa are merely neoflightless birds, completely unrelated to Theropoda (Feduccia 1996, 1999, Feduccia in Fischman 1999, Martin and Czerkas 2000, Ruben & Jones 2000). While a possible (though not necessarily plausible) case for the neoflightless, avialian status of Protarchaeopteryx has been made (Feduccia 1996, 1999, Paul 2002) similar arguments advanced on the behalf of Caudipteryx zoui must be called into question, and they are shown untenable by the material available. Since the most recent attack on the dinosaurian affinity of Caudipteryx (Ruben & Jones 2000), new material attributable to the species has been recovered and described (Zhou & Wang 2000, Zhou et al 2000), which has concomitantly led to a more thorough understanding of caudipterygian osteology.
Caudipteryx displays multiple characters synapomorphic of Oviraptorosauria. Contrary to Feduccia’s assertions, said traits are not limited to the articulation of the temporal region. In the analyses offered by Sereno (1999), Zhou & Wang (2000), Zhou et al (2000), and Barsbold et al (2000), the suturing of the quadrate and quadratojugal are stressed as plesiomorphic comparative to Aves. The presence of a robust dorsal process of the jugal, is the only other cranial character cited in these reviews (Witmer in Chiappe & Witmer 2002). Traits allying caudipterygians and oviraptorosaurs in a holophyletic clade (sensu Sereno 1999, Barsbold et al 2000) include the structure of the pubes, the procumbent orientation thereof, the prominent obturator process, the shallow pelvic canal, the contact between the fibula and calcaneum, and the dorsolateral projections of the astragalus, which overlap the calcaneum (Sereno 1999, Barsbold et al 2000). The structure of the mandibles seems to further ally these two groups (Paul 2002), notably, the presence of a mandibular fenestra is a significant plesiomorphy which seems to further argue against the status of this taxon as a neoflightless "Mesozoic kiwi" (Witmer in Chiappe & Witmer 2002).
The argument posed by Feduccia (1999) that herbivory in caudipterygians is irrefutable evidence for aviaan status, is directly falsified by the herbivorous or at least omnivorous paleobiology of other oviraptorosaurs, and particularly therizinosaurs—both theropod clades. In light of the osteology unifying caudipterygians and oviraptorosaurs, arguments that Caudipteryx is merely a neoflightless avialian, are at best unsubstantiated, and at worst entirely specious.
Perhaps the most damning testimonial against the assertion that feathers remain avian autapomorphies, is the feathery integument observed in the specimens of the fabulous Microraptor. While these structures have not yet been extensively described, they appear to be asymmetric remiges. An argument for non-homology is simply indefensible.
Lingham-Soliar with the Rebound
In two recent papers, Theagarajen Lingham-Soliar (2003a, 2003b) has argued that detailed similarities between fossilized ichthyosaur and recently buried bottlenose dolphin (Tursiops truncatus) collagen fibers strongly suggests homology between the two. Lingham-Soliar hones his sights on two studies in particular: Currie & Chen (2001) and Xu et al (2001), and it is so littered with errors that its well worth quoting at some length.
In his introduction (pg. 2), Lingham-Soliar says that the "titles of articles (Xu et al. 2001; Ji et al. 2001) that proclaim "feathered dinosaurs" actually describe integumentary fibers or feather-like structures, which are not the same as avian feathers, as some of these authors recently noted (Norell et al. 2002)." Prum, whose model of feather evolution he cites in the first line and apparently believes to be the sine qua non of research in the field, disagrees. On page 297 of the very paper he references, Prum says "the follicle originated with the cylindrical epidermal invagination around the base of the feather papilla. The undifferentiated tubular collar yielded the first feather-a hollow cylinder that resembles the calamus, or sheath, of a modern feather" [emphasis added]. This conforms to the widely used definition given by Brush (2001), that a feather has two essential properties; it is composed beta keratin, and protrudes from a follicle.
On page six, he continues: "It was not so long ago that pterosaurs covered in hair pervaded reconstructions (e.g., see Wellnhofer 1991), an unfortunate interpretation of Sharov's (1971) description of hair-like structures in Sordes pilosus. More recently, however, the structures were shown to be fibers supporting the wings (Unwin and Bakhurina 1994), straight, closely packed fibers on the outer half of the wing and shorter, loosely packed fibers close to the body." Lingham-Soliar presents Unwin & Bakhurina's paper in a way which makes it appear as if they have concluded all the pterosaur integument was of this sort, which is decidedly not the case.
On the more substantive issues, he makes repeated assertions that the collagen fibers approximate dromaeosaurid integument. A nearly exhaustive sample includes:
"Most [feather] fibers are near-parallel (fig 2) and the few instances of divergence can easily be accounted for by taphonomic professes [associated with collagen fibers]." (2003a, pg. 4)
"In other places on the same icthosaur specimen, fine fibers of one layer are compressed onto thick fibers of another, giving an uncanny resemblace to rachis and barbs" (2003a, pg. 5)
"...'wavy' fossilized fibers described in the theropod dinosaur Sinosauropteryx (Currie and Chen 2001), as 'soft and pliable', conform in nature and behavior to collagen fibers." (2003b, pg. 566)
"Slower mineralization may also account for darker edges and lighter center of integumentary structures (figure 1f in Lingham-Soliar 2003[a]), in contrast to the interpretations of similar features in a dromaeosaur as reflecting hollow integumental structures resembling the hollow rachis of feathers (Currie and Chen 2001)" (2003b, pg. 567)
Whats most interesting about all of this is that the the ichthyosaur collagen fibers can be seen to overlap bone or lie medial to the skin surface (see, for example, fig. 1a in 2003a). This is clearly not the case in feathered dromaeosaurids, where integument is clearly external to, and reaches far beyond, the skin and bone. NGMC 91 (Sinornithosaurus cp.) is one extremely clear example. Additionally, the most feather-like of the collagen is frayed from the cross-fiber patchwork associated with aquatic skin types. Unless the author is suggesting (which he did not), that the sort of skin seen in cetaceans and sharks evolved, for some spectacularly odd reason, on terrestrial vertebrates like dinosaurs, one is left to wonder just how appropriate his observations really are.
If not aquatic dinosaur skin, where did these collagen fibers comes from? We are told that there are "two possible scenarios are envisaged in Sinosauropteryx: either the bundles of tightly strung ligaments broke contact with the vertebrae during post-mortem decay and came to lie alongside the caudal vertebrae, or the skin possessed masses of strengthening fibers or rays vertically orientated to the long axis of the body" (2003a, pg. 6).
As Chen et al (1998) observe, the integument is by no means limited to the midline, so these suggestions are rather difficult to comprehend. One also must wonder if this is supposed to apply to other species as well. As Prum (2002) and Prum & Brush (2002) rightly ask, are we to believe that Beipiaosaurus had a 50-70 mm long ligament on its ulna? That NGMC 91 had a 35 mm ligament on its snout?
Finally, he concludes (in 2003a) that "it is impossible to say that the famous Chinese dromaeosaurs did not possess feathers, or that they are not collagen fibers, which has been suggested as the biological material preserved in these dinosaurs (Feduccia 1999)." It is odd that Dr. Lingham-Soliar references Norell et al (2002), a paper preliminary describing Cryptovolans (=Microraptor gui?) and announcing its unmistakable asymmetrical remiges, but still tells us all dromaeosaurid feathers could be collegen fibers. He seems to at least understand the problem in principle, rhetorically asking if "the discovery of true feathers in non-avian dinosaurs support[s] the filamentous "protofeather" stage in Prum's (1999) model." He suggests it would not, saying "both hair and collagenous fibers have been discovered in mammoths (see Kukhareva and Ileragimob 1981) with obviously no evolutionary connection." Aside from begging the question, this argument is odd in that 1) the presense of remiges certainly had a phylogenetic origin, constituting at least prima facie evidence for the integument being keratinous, and 2) it is precisely the sort of feather Prum predicts (his stage 1). In the other paper, (2003b), Lingham-Soliar does acknowledge the presense of pennaceous feathers reported in Norell et al (2002), but places "modern" in those sneering inverted commas, and suggests, without argument or further comment, that they're "not beyond dispute" (presumably because its "an alleged dromaeosaur"), and repeats his former argument.
It should also be reemphasized that the identical integument of the alvarezsaur Shuvuuia deserti has been shown by Schweitzer et al (1999) to have been composed of beta-keratins, and could therefore not be dermal in origin.
Phylogenetic Double-Takes, Longisquama, and Plucking the Downy Dinos
The discovery of forms such as Microraptor gui, preserving uambiguously closed pennaceous vanes displaying asymmetry has naturally triggered some interesting reactions amongst those who argue against a dinosaurian origin of birds. Since the remigial fans of Microraptor are clearly visible, and their homology to avian feathers undeniable, the phylogenetic back-flips that have resulted in some quarters of this debate are equally stunning. Contradicting 30 years of insistence that theropods are superficially alike birds in every way, it is now common currency amongst advocates of the "thecodont" origin of birds, that dromaeosaurs, particularly Microraptor, are in fact secondarily flightless birds near to the origin of Aves, that have merely converged upon theropods (Feduccia 2002). In the process these maniraptorans have been stuck in phylogenetic limbo, being neither dinosaur nor bird because neither classification can be adequatley defended by opponents to the theropod origin of birds. Our incredulity aside, let us examine the fundamental argument of this proposition.
Adherents of this new phylogenetic interpretation (e.g., Martin, Feduccia) argue that forms such as Microraptor are sufficiently avian that they cannot be grouped with theropod taxa, and are thus not coelurosaurs at all. In one respect, this argument is without flaw. Mircoraptor is astonishingly avian, perhaps more so than any other single theropod ever discovered. Undoubtedly this is a form very close to the origin of birds. Yet is the second assertion at all valid? Even a cursory examination of the skeleton of Microraptor demonstrates that however avian it is, it can in fact be classified as a dromaeosaur on the basis of multiple synapomorphies of this clade (Xu et al. 2003), including:
a) Chevrons and prezygapophyses extremely elongate
b) Manual phalanges III-1 longer than III-2
c) Pedal digit II modified in typical dromaeosaur fashion
d) Metatarsal V long
It is further worth noting that the intellectual structure of this objection to the theropod hypothesis, is bankrupt, as it denies the mosaic nature of evolution. For instance, following the sort of argument by taxonomic provision that Feduccia et al. have recently taken up, it could be asserted that as Archaeopteryx is a bird, it cannot reveal anything about the origin and early evolution of birds. Essentially, the statement that Microraptor is a bird and therefore is meaningless to the debate on avian ancestry is in effect, anti-evolutionary, regardless of whether Microraptor is a bird, or a feathered dromaeosaur.
And what of alternatives to these "downy dinos," which have been advanced as ideal avian ancestors? The hopes and dreams of the "thecodont" camp are pinned upon an enigmatic Upper Triassic reptile from Kyrgyzstan, described over thirty years ago by A. G. Sharov (1970), Longisquama insignis. Surely to be doomed to obscurity in a curator's draw but for the nature of its bizarre dorsal integumental appendages, Longisquama has been debated since its discovery and initial description. Jones et al. (2000) argued that the integumentary structures of Longisquama were homologous with feathers, and thus concluded that Longisquama presented potent evidence for a "thecodont" ancestry of birds. The history of varying interpretations of these structures, is, however, of interest in that researchers have been nearly unanimous in the past in stating that they are not feathers (e.g., Feduccia 1985). Futhermore, their very association with the skeleton preserved on the main slab of the type has been questioned, as these appendages bear a rather striking assemblage to plant material (Feduccia 1985, Paul 2002). Thus it was a surprisingly iconoclastic paper submitted to Science by Jones et al., as it contradicted the very observations some of its own authors had offered concerning Longisquama in years past. These details aside, it is much to the credit of Jones et al. (2000) that they did undertake a very thorough examination of the integumentary structures in an attempt to determine if they bore any relation to feathers. Their conclusion that they are feather homologs rested on several assertions.
Jones et al. argued that the blade of each individual appendage was pinnate, with distal fusion similar to that seen in some extant birds. They further argued that these appendages were derived from tubular, hollow bases arising through follices, as in avian feathers. This argument rested on the allegedly calamus-like structure of the basal region of each individual appendage, which it was claimed consisted of an ensheathed dermal pulp and cornified pulp cavities (e.g., Fig. 6, Jones et al. 2000).
However, these conclusions have been sharply criticized (e.g., Prum 2000, 2002, Paul 2002), and there are significant problems with the interpretations of Longisquama integumentary morphology offered by Jones et al. (2000). Perhaps most seriously they based their conclusion that the vane of each appendage was pinnate on scant evidence, relying on a subjective determination that the matrix quality of a homogenous surface with plications would have been more consistent than that observed in Longisquama. As Prum (2000) noted, this hardly seems a compelling basis for their profound assertions about the homologies of these appendages. Debate concerning the relative distribution of such characters as the distal fusion of the alleged pinnae, emphasized by Prum (2000), are as noted by Jones et al. (2000b), irrelevant to determining whether or not these structures are related to feathers, and thus shall not be discussed at length here. The structure of the vane itself in these appendages is not consistent with that of feathers, most glaringly in the orientation of the supposed pinnae, which face inward and not outward toward the central shaft of the appendage. This is entirey incongruous with what is known of pinna and rachis development in feathers, which follows a helical growth pattern, followed by obverse fusion of the pinnae to form a rachis (Prum 2002). The structure of the basal region of the appendages, which is apparently cylindrical, is however in little other respect similar to the calamus of birds, and Unwin & Benton (2000) noted that allegedly avian structures such as cornified pulp caps, are equally if not more readily epxlicable as preservational defects. Given these data, the conclusions reached by Jones et al. (2000) seem largely unsubstantiated.
An equally serious problem in the discussion of the phylogenetic relevance of Longisquama is its very status as an archosaur, which is open to interpretation. Unwin & Benton (2000) noted that the pattern of cranial fenestration cannot adequately be determined, and the apparent mandibular fenestra is not particularly congruent with that of other basal archosaurs. Most interestingly the dentition of Longisquama is not thecodontian, plesiomorphic for archosaurs, but rather acrodont, typical of Lepidosauria. If Longisquama is in fact not an archosaur but an aberrant lepidosaur, it stretches credulity to envision it as playing any significant part in the ancestry of birds.
Conclusive data falsifying the feather-homologue status of the integumentary derivates preserved in several exceptional theropod fossils from the Yixian lagerstatten of China has yet to be advanced. On the contrary, there is evidence to suggest that these structures are both feather antecedents, and homologous thereto.
- Bakker, R. T. 1974. Dinosaur bioenergetics—a reply to Bennet and Dalzell, and Feduccia. Evolution 28: 497-503.
- Bakker, R. T. 1986. The Dinosaur Heresies: New Theories Unlocking The Mystery of the Dinosaurs and Their Extinction. William Morrow, New York.
- Barsbold et al. 2000. A pygostyle from a non-avian theropod. Nature 403: 155-156.
- Chen et al. 1998. An exceptionally well preserved theropod dinosaur from the Yixian Formation of China. Nature 391: 147-152.
- Chiappe, L. M. 1995. The first 85 million years of avian evolution. Nature 378: 349-355.
- Chiappe, L., Norell, M., & Wu. 1998. The skull of a relative of the stem-group bird Mononykus. Nature 392: 275-278.
- Feduccia, A. 1996. The Origin and Evolution of Birds, First Edition. Yale University Press, New Haven.
- Feduccia, A. 1999. The Origin and Evolution of Birds, Second Edition. Yale University Press, New Haven.
- Feduccia, A. 2002. Birds are dinosaurs: Simple answer to a complex problem. The Auk 119(4): 1187-1201.
- Fischman, J. 1999. Feathers don’t make the bird. Discover 20(1): 48-49.
- Geist et al. 1997. Implications of soft-tissue preservation in the compsognathid dinosaur, Sinosauropteryx. Journal of Vertebrate Paleontology 7 (supplement to no. 3): 48A.
- Ji, Q. & Ji. S. 1996. On the discovery of the earliest bird fossil in China and the origin of birds. Chinese Geology 233: 30-33.
- Ji et al. 1998. Two feathered dinosaurs from northeastern China. Nature 393: 753-761.
- Jones, T. D., Ruben, J. A., Martin, D., Kurochkin, E., Feduccia, A., Maderson, P. F. A., Hillenius, W., Geist, N., & Alifanov, V. 2000. Nonavian feathers in a Late Triassic archosaur. Science 288 (5474): 2202-2208.
- Jones, T. D., Ruben, J., Maderson, P. F. A., & Martin, L. 2000. Longisquama fossil and feather morphology, Response. Science 291 (5510): 1899c.
- Lingham-Soliar, T. 2003a. Evolution of birds: ichthyosaur integumental fibers conform to dromaeosaur protofeathers. Naturwissenschaften 90: 428-432.
- Lingham-Soliar, T. 2003b. The dinosaurian origin of feathers: perspectives from dolphin (Cetacea) collagen fibers. Naturwissenschaften 90: 563-567.
- Martin, L. & Czerkas, S. A. 2000. The fossil record of feather evolution in the Mesozoic. American Zoologist 40(4): 687-694.
- Morell, V. 1997. The origin of birds: the dinosaur debate. Audubon 99: 10-11.
- Ostrom, J. 1973. The ancestry of birds. Nature 242: 136.
- Padian et al. 2001. Feathered dinosaurs and the origin of flight. In: Tanke, D. H. & K. C. Carpenter (eds.), Mesozoic Vertebrate Life: 117-133.
- Paul, G. S. 2002. Dinosaurs of the Air: The Evolution and Loss of Flight in Dinosaurs and Birds. Johns Hopkins University Press, Baltimore.
- Prum, R. 2000. Longisquama fossil and feather morphology. Science 291 (5510): 1899c.
- Prum, R. 2002. The evolutionary origin of feathers and diversification of feathers. The Quaterly Review of Biology 77(3): 261-295.
- Ruben, J. & Jones, T. D. 2000. Selective factors associated with the origin of fur and feathers. American Zoologist 40(4): 585-596.
- Schweitzer et al. 1999. Beta-keratin specific immunological reactivity in feather-like structures of the Cretaceous alvarezsaurid, Shuvuuia deserti. Journal of Experimental Zoology 285: 146-157.
- Sereno, P. 1999. The evolution of dinosaurs. Science 284: 2137-2147.
- Unwin, D. M. & Benton, M. J. Longisquama fossil and feather morphology. Science 291 (5510): 1899c.
- Witmer, L. The debate on avian ancestry. In: Chiappe, L. & Witmer, L (eds.), Mesozoic Birds: Above the Heads of Dinosaurs, 3-31.
- Xu, X., Wang, X., & Wu, X. 1999. A dromaeosaurid dinosaur with a filamentous integument from the Yixian Formation of China. Nature 401: 262-266.
- Xu, X., Tang, Z., & Wang, X. 1999. A therizinosaurid dinosaur with integumentary structures from China. Nature 399: 350-354.
- Xu, X., Zhou, Z., & Wang, X. 2000. The smallest known non-avian theropod dinosaur. Nature 408: 705-708.
- Xu, X., Zhou, Z., Wang, X., Kuang, X., Zhang, F. & Du, X. Four-winged dinosaurs from China. Nature 421: 335-340.
- Zhou, Z. & Wang, X. 2000. A new species of Caudipteryx from the Yixian Formation of Liaoning, northwest China. Vertebrata PalAsiatica 38: 111-127.
- Zhou et al. 2000. Important features of Caudipteryx—Evidence from two nearly complete new specimens. Vertebrata PalAsiatica 38: 241-254.
- Zhou et al. 2002. A juvenile coelurosaurian theropod from China indicates arboreal habits. Naturewissenschaften (2002)89: 394-298.