Arguments Against The Dinosaurian Ancestry of Birds

Bird Ancestors: A Shrinking Temporal Discordance

Several creationist authors have argued that the bird-like theropod taxa are actually 20 million years younger than Archaeopteryx, and that this is evidence against the theropod ancestry of birds. A review of how this time discordance argument has evolved over the past decade reveals its weakness. It is true that until rather recently, bird-like dinosaurs were known only the mid-late Cretaceous (e.g. Velociraptor mongoliensis), in deposits much younger than those containing Archaeopteryx.  For instance, as recently as the mid 90's, opponents of the theropod-bird hypothesis were saying that "most of the supposed similarities between urvogel and dinosaurs are seen in birdlike dinosaurs that lived 80 to 100 million years later" (Feduccia, 1994, p. 32). By 1996, they were musing that, to theropod-bird proponents "it is inconsequential that birdlike dinosaurs occur some 75 million or more years" after Archaeopteryx (Feduccia, 1996, vii). And by 1999, the gap closes even further, and Feduccia (1999, p.4740) wonders "why these superficially birdlike theropods only occur in the fossil record 30 to 80 million years after the appearance of the earliest known bird." Now, creationists and other opponents of the theropod-bird hypothesis are arguing for a discordance of 20 million years. And not surprisingly, more recent finds show that even this is a slight overestimate, since Sinovenator, a very bird-like theropod from from the lowest part of the Yixian Formation, dates to 128Ma, making it 17 million years younger than the Archaeopteryx specimens from the Jurassic of Germany (Xu et al, 2002). So, the time discordance has been dramatically reduced, from 80-100 million years, to 75 million years, to 30 million years, to 17 million years.

And despite the rarity of fossil lagerstatten, and particularly of well-preserved fossil birds in general (note that both Archaeopteryx and the Liaoning theropods are from rare lagerstatten), there is now suggestive but incomplete fossil evidence that some of the bird-like theropod groups, such as dromaeosaurs (the same family of theropods as many of the Liaoning 'dino-birds'), did in fact exist prior to Archaeopteryx. Witmer (2002, p. 19) notes:
There are much greater time discordances in the dinosaur fossil record (Sereno 1997b, 1999a) than this one. But, moreover, there are a variety of of fragmentary specimens (mostly teeth) of animals that closely resemble those of dromaeosaurids and troodontids recovered from Middle Jurassic deposits that predate Archaeopteryx by 20My (Evans and Milner, 1994; Metcalf and Walker, 1994). Similarly, Zinke (1998) reported an extensive collection of theropod teeth from deposits perhaps just slightly older than Archaeopteryx; Zinke made firm assignments of these teeth to Dromaeosauridae (29 teeth), Troodontidae (14 teeth), and Tyrannosauridae (3 teeth). Finally, Jensen and Padian (1989) described fragmentary but provocative skeletal material of maniraptoran theropods from the Late Jurassic Morrison Formation.
Also, there are some curious features of the critics' positions that should be pointed out. For instance, from an 'old earth' perspective, which accepts the validity of radiometric dating and stratigraphic correlation, the weakness of the argument is demonstrated by ghost lineages and lazarus taxa. These are taxa that are known from time A and from time C, but are not known from intervening time B. Naish (1998) gives the example of the champosaurs, for which there are ghost lineages of at least 45 million years long, from the late Triassic to the middle Jurassic. And then of course there is the infamous coelacanth, for which there is a ghost lineage of about 65 million years. So unless Ross and other old-earth creationits hold that these taxa were created, then went extinct, and then were created once again tens of millions of years later -- unless they want to take that position, which seems unlikely- then they would have to acknowledge that whatever time discordance does exist for the theropod-bird hypothesis is not good evidence against that hypothesis.

Finally, for Feduccia to complain of a temporal discordance is somewhat hypocritical. As Brochu and Norell (2000) point out, when you consider all of the proposed models for the ancestry of birds, including those of Feduccia, and consider the consistency of all of the nodes in the cladogram with the stratigraphic record, then the theropod hypothesis actually compares favorably to all the alternatives.

The 'Collagen Fibers' of Sinosauropteryx?

One rather strange idea that seems to be widely accepted amongst creationists is that the fibers of Sinosauropteryx are not integumentary structures at all, but frayed collagen fibers that in life were part of a dermal frill. According to one creationist article:
A remarkable fossil find from the Yixian formation in China revealed the theropod dinosaur called Sinosauropteryx, which was nicknamed the "feathered dinosaur." Subsequent studies have indicated that the feathers were probably "frayed collagenous fibers beneath the skin."6 (Demise of the 'Birds are Dinoaurs' Theory).
Let's start with the reference, which is wrong. Reference 6 is to a commentary in the journal Science by Gibbons, and does not include the quoted text. However, it does make a reference to the now-refuted hypothesis (Geist 1997; Feduccia, 1999), that the feathery structures of Sinosauopteryx were not feathery at all, but rather were "frayed collagenous fibers from beneath the skin" that were originally part of a dermal frill running along the dorsal midline of the body. This was proposed following the discovery of the first specimen of Sinosauropteryx prima, a dromaeosaurid theropod with filamentous structures evident around the body. This claim was highly questionable even when it was proposed, and more recent specimens from the same deposits show that such a interpretation today is wholly untenable. For one thing, even in the first described specimen, the feathery fibers were not limited to the animal's sagittal plane, since the animal's head is rotated slightly to the right, and the fibers thus interesect the left dorsolateral surface of the skull rather than the sagittal plane. Moreover, subsequently discovered specimens of Sinosauropteryx described by Chen et al (1998) showed the filamentous structures not only along the dorsal surface of the neck and back, as might be expected for a dermal frill, but also along the margins of the tail, with patches on the side of the skull, humerous and ulna.

In addition, other nonavian theropods from the same deposits, for instance the therizinosaur Beipiaosarus inexpectus, have the same type of simple filamentous structures as Sinosauroptryx prima present on both fore- and hindlimbs, a distribution which again is clearly inconsistent with their having been part of a sagitally-placed dermal frill. While these structures are not identical to feathers of modern birds, they may well be feather homologues and/or represent a primitve stage of feather evolution. These structures are in fact integumentary structures, they were on the outside of the skin, they clearly did not run down only the dorsal midline of the body like a dermal frill, and both macroscopically and under magnification they look nothing like collagen fibers. Also interesting in this context is the report by Schweitzer et al (1999) that the filamentous structures preserved around the head of Shuvuuia deserti were both hollow and composed of beta keratin. If both observations are correct, then these structures may well have been feather homologues, since feather rachi are the only type of integumental cover that is both hollow and composed of beta keratin. At present it is not possible to say how widely distributed such structures were amongst dinosaurs, but they may have been more widely distributed than previously assumed.

More importantly, recently discovered theropods show a more complex integumentary structure, with complex branching, in some cases covering nearly the entire body (e.g. Ji et al, 2001; Xu et al, 2001). Examples include Sinornithosaurus millenii, Microraptor zhaoianus, and Microraptor gui (Norell, 2002; Xu et al, 1999; Xu et al, 2000; Xu et al, 2003). Regarding the integumentary structures present on the dromaeosaurid theropod Sinornithosaurus milenni, Xu et al (2001, p. 200) noted the presence of "two types of branching structure that are unique to avian feathers: filaments joined in a basal tuft, and filaments joined at their bases in series along a central filament." Feathers are the only known branching integumental structures in vertebrates, and the structure of the 'feathers' of Sinornithosaurus millenii match extremely well stages II and IIIA predicted by Prum's (1999) developmental model of feather evolution. Also, though barbules are not evident in these specimens, they may well have been present, given how well-ordered the barbs are (without barbules, the barbs would be much more clumped together). The newly described specimens of the newly described species Microraptor gui, specimens IVPP V13352 and TNP00996, not to be confused with Microraptor zhaoianus, are even more compelling (Xu et al, 2003; Prum, 2003). Here we have a theropod which can be "unequivocally referred to Dromaeosauridae" (Xu et al, 2003, p. 336) possessing unequivocal 'avian' feathers (see for instance, figure 2e in Xu et al, 2003). One of the interesting features of is the presence of feathers on both forelimbs and hindlimbs, making it the only known 'tetrapteryx.' This provides evidence that Microraptor gui may well have been capable of gliding, consistent with an arboreal origin of bird flight, making Microraptor gui signifcant as both a morphological and a functional intermediate form.

Dinosaur Lungs

Based on Ruben et al (1997, 1999), many creationist websites claim that theropod dinosaurs had crocodile-like pelvovisceral muscle pump (aka 'hepatic piston') rather than the air-sac complex typical of modern birds, that a transition from a crocodile-like lung to a bird-like lung is impossible, and therefore that theropod dinosaurs could not have been ancestral to Archaeopteryx and to later birds. As we'll see, however, the skeletal morphology morphology of theropods, far from demonstrating the presence of a croc-style pelvovisceral pump, actually strongly suggests just the opposite. Finally, the soft-tissue evidence for pelvovisceral pump ventilation in theropods is entirely unconvincing.

Pelvic Morphology

Ruben et al are apparently the source for the rather absurd claim that the pelves of crocodiles and theropods are very similar to each other but are dissimilar to the pelves of Archaeopteryx- the point being that the orientation of the pubis in theropods suggests possession of the pelvovisceral pump lung, while the orientation of the pubis in Archaeopteryx is inconsistent with this taxon having possessed this type of lung. One creationist author writes:
. . . the pelvic bones of the theropod dinosaurs look nothing like that of either modern birds or Archaeopteryx, but look very similar to that of modern reptiles, such as the crocodile. There is no way for the pubis of modern reptiles or the theropod dinosaurs to serve as an attachment point for suprapubic muscles to serve in assisting breathing during perching. Since there are no "intermediate" theropod which possesses a pelvic structure similar to Archaeopteryx, it seems unlikely that they could have given rise to Archaeopteryx (Demise of the 'Birds are Dinosaurs' Theory).
This is inaccurate on essentially every count. Below are three pelves. From left to right, they belong to the dromaeosaur Deinonychus, Archaeopteryx, and Aquila (Golden Eagle).

 

The reconstruction of Archaeopteryx's pelvis in Ruben et al (1997), which is the source for several of the claims made in their paper and on the website quoted above, is inaccurate. Ruben et al's figure 6b shows the pubis of Archaeopteryx rotated backwards and nearly parallel with the ischium, very similar to the condition seen in modern birds. But this reconstruction of Archaeopteryx's pelvis is inaccurate and outdated, based on the London and Berlin specimens. In all 3 of the specimens of Archaeopteryx in which the pubis is well-preserved, the pubis is oriented nearly vertically, or offset only slightly posterodorsally, as is typical of most theropods, and is not nearly as retroverted as shown as Ruben et al (1997). Elzanowski (2002) in his recent review of the Archaeopterygidae comments that in the Maxberg, Eichstatt, and Munich specimens, the pubis is nearly vertical. This is signficant, because the pelvis of Munich specimen is clearly well preserved and the pubis is not broken from the ilium. Elzanowski (2002, p. 144) writes:
The well-preserved pelvis in the Munich specimen (Wellnhofer, 1993) confirmed that the pubis was directed nearly vertically downward as in Unenlagia (Novas and Puerta, 1997) and Rahonavis (Forster et al, 1998). The pubis was only slightly retroverted, forming a cranial angle of 110d with the long axis of the ilium (Wellnhofen, 1985). The caudal orientation in the London and Berlin specimens is due to a dislocation in the acetabulum . . . Whatever the exact angle, the orientation of the pubis is intermediate between neornithines and the majority of nonavian theropods, except for therizinosaurids and dromaeosaurids, which have the pubis strongly retroverted (Norell and Makovicky, 1997). Ruben et al's (1997) speculations about pulmonary ventilation in archaeopterygids being different from that of the nonavian theropods are based on a reconstruction that exaggerated the backward slant of the pubis in archaeopterygids.
This reconstruction is bolstered by the discovery of  early birds with well-preserved pelves, such as Rahonavis ostromi, in which the pubis is indeed oriented vertically relative to the long axis of the ilium. Interestingly, in footnotes 17 of their article Ruben et al admit that "[t]he position of the pubis in Archaeopteryx has occasionally been interpreted as having been vertical." Actually, no one today would reconstruct the pelvis of Archaeopteryx as Ruben et al do, with the pubis effectively parallel to the ischium. Clearly the pelvic morphology of Archaeopteryx unites it with rather than divides it from theropod dinosaurs.

Also contary to Ruben et al, the pelvic morphology of crocodilians is very different from that of theropods. The pubis of crococilians is short and broad, not "elongate" as Ruben et al (1997) suggest, there is no pubic boot as in theropods, the pubis becomes distally transversely broad forming a plate that partially supports the abdomen, and crocodilian pubes are actually uniquely mobile -- they swing back and forth to accomodate the posterior displacement of the viscera during breathing (Paul, 2002, p. 349-350). In theropods and early birds, by contrast, the pubis is never mobile, the pubis is elongate and transversely narrow, the pubis is oriented vertically or is retroverted as in birds, there is no distal transverse broadening of the pubis, and the pubis is not mobile. In fact, the slenderness of many theropod pubes, and the highly variable orientation of the pubis even in closely related theropod taxa, is not consistent with thier having anchored such a pump. 

Other Adaptations Associated with Pelvovisceral Pumps

Other features of theropod anatomy argue against the presence of a pelvovisceral pump in theropods as well. For instance, in all crocodilians there is a well-defined, rib-free lumbar region. This lumbar region allows the abdomen to expand during the inspiration phase, similar to human breathing. Theropods lack a rib-free lumbar region. Paul (2002, p. 352-354):
Contrary to the opinions of Ruben et al (1997b, 1998) and Hengst (1998), the trunks of avepod dinosaurs and crocodilians could hardly have been more different. The double heading of the entire avepod dorsal rib series is in sharp contrast to the crocodilians' winglike transverse processesses (contra Hengst, 1998). The avepods' long, gracile pubes differ dramatically from the broader, stouter crocodilian structure.  .  . Appedix table 3 shows that the two groups do not share a single osteological adaptation that supports the presence of a pelvodiaphragmatic muscle pump in any predatory dinosaur. Quite the contrary, the operation of a  pelvodiaphragmatic pump appears to have been impossible in the latter. The absence of a smooth, bony rib cage ceiling means that the ceiling was not specialized to accomodate, and may have hindered, the strong back-and-forth movement of an expandable lung, even if intracostal muscles and other tissues lined the ceiling of the thorax. The lack of either a lumbar region or mobile pubes would have hindered the abdominal volume changes inherent in the operation of a  pelvodiaphragmatic muscle pump. Because the abdominal surfaces of dino-avepod pubes were usually narrow, especially the most gracile examples, they appear too transversly narrow to anchor large,  pelvodiaphragmatic muscles (see Hutchinson [2001] for additional comments). Strongly retroverted pubes were especially poorly suited for supporting these respiratory muscles. Of the derived examples, the pubes of alvarezsaurs were too slender and weak to anchor and resist the pull of  pelvodiaphragmatic muscles. For that matter, it is difficult to see how the pubis could have been a part of the respiratory complex when its orientation and boot development was so variable within the group as a whole, and even among close relatives.

Ruben et al (1997, p. 1258) claimed also that "theropods lacked avianlike jointed or hinged ribs and an expansive sternum (9), structures without which proper ventilatory airflow cannot be maintained in the modern bird lung." However, recently described bird-like dromaeosaurs, for instance Sinornithosaurus (Xu et al, 1999), do indeed posesss several sternal ribs and sternocostal joints as well. Second, it is clear that some theropods, such as Velociraptor, Sinornithosaurus and Bambiraptor, had large sternal plates, some almost half as long as the ribcage, relatively much longer than those of Archaeopteryx, in which the sternum is about 12% the length of the rib cage (Paul, 2002, p. 356). Paul (1997) observes:
. . . large ossified sternal plates -- at least as large as those of kiwis -- were described and figured in dromaeosaurs and oviraptors by Barsbold in 1983, and have been discussed and figured in many other publications! In an odd way this denial makes sense, in that Ruben et al 1997b use an out of date cross-section of the ribcage of a dromaeosaur -- based on incomplete disarticulated remains -- in a futile effort to deny/ignore the evidence provided by complete ribcages, that these near birds have large sternal plates. These big sternal plates articulated with the coracoids via a transverely long hinge joint which allowed the sterna to help ventilate the antero-ventral air-sacs. Ruben et al also ignore the ossified dromaeosaur sternal ribs published by Ostrom in 1969. They fail to mention the ossified uncinate processes present on the fighting Velociraptor. Why do not Ruben et al make any mention of the work by Britt (1994) or Reid (1996) showing that the pneumatic vertebrae of theropods are strongly indicative of the presence of pulmonary air-sacs? Fact is that advanced theropods had the most bird-like trunks of any tetrapods.
Furthermore, Leahy (1998) points out that in "some juvenile precocial birds, the sternum is very small, entirely cartilaginous, and the sternal ribs do not yet articulate with the thoracic ribs (Fig. 7 in Olson, 1973), yet these birds are fully capable of  ventilating their lungs." So, neither the sterna nor the ribs of theropods rule out the existence of bird-like lung, any more than it rules out a bird-like lung in kiwis and precocial chicks.

Ruben et al (1997, p. 1268) also claimed that the "well-developed gastralia" of theropods are evidence for crocodilian hepatic piston lungs. Yet, Archaeopteryx, Confuciusornis and other early birds, unlike modern birds, also possessed gastralia like those possessed by theropods. Caudipteryx, which most researchers consider an oviraptorid but which Ruben and colleagues consider a flightless bird, also possessed gastralia, as did Protarchaeopteryx and Microraptor.

Soft-Tissue Evidence for Pelvovisceral Pumps in Theropods

Ruben and colleagues have argued that rare soft-tissue evidence from fossil theropods supports their having possessed pelvovisceral pump lungs. Both Ruben et al and creationist commentators on Ruben et al have characterized such as evidence as disproof of the theropod ancestry of birds. As we will see, however, none of the soft-tissue evidence for pelvovisceral pumps in theropods is convincing. Ruben et al (1997) write:
Recently described Early Cretaceous theropod specimens [Sinosauropteryx (11)] retain preserved outlines of much of the visceral cavity. The cavity exhibits complete thoracic-abdominal separation, defined by a remarkably crocodilianlike vertically oriented partition coincident with the apparent dome-shaped anterior surface of the liver (Fig. 5).
In fact, nothing in their figure 5 shows "complete thoracic-abdominal seperation." This is one possible intepretation, but not the only one nor even a particularly compelling one. The specimen in question does show a darkened area, which may or may not be a flattened liver. The darkened area does span the entire body cavity, but this does not prove a seperation of thoracic and abdominal cavities, and in some birds the liver is also so large as to nearly span the entire body cavity from sternum to vertebrae (Paul, 2002, p. 347). Nor is it clear this darkened area represents the liver, nor, if it is a liver, that the it spanned the body cavity during life; it could be transversely flattened like a pancake. The alleged "dome-shaped anterior surface of the liver" appears to be nothing but a preservational artifact. An actual 'partition' (ie. a septum) is not seen on the fossil, it is simply inferred from the shape of the putative liver. In fact, according to Paul, Ruben et al (1997) actually misidentified sediment breakage as soft-tissue, perhaps because they were working from photographs only and had not actually seen the Sinosauropteryx specimen first-hand.  Paul (1997) wrote:
It is important to understand that Ruben et al have not actually seen any of the Sinosauropteryx specimens, they are making guesses based on photos of just one specimen, photos that fail to show its 3-D complexity. In Fig. 5A they use a low resolution, out of focus photo to contend that there is a semi-circular anterior border to the abdominal cavity. There is no such thing. Examination of higher quality, larger format photos on the cover of the April 97 Audubon (counterslab) and Nov 14 97 Science (main slab) and the March 97 Episode (both slabs) show that much of the supposed border of the abdominal cavity is really an irregular break in the sediment! This is especially obvious in the superbly detailed Audubon photo, where the shallow rim of a large, semi-circular, light colored break under the spinal column is clearly delineated both by cast shadows, and obvious breakage of the ribs at that location, the rest of the ribs are complete. Ruben et al saw this photo since they cited it in their paper, yet they make no mention of the damage (either they missed it, or thought it unimportant). Yet their arrow points to this break as the septum -- i.e. they misidentify sediment damage as soft tissue anatomy. On the main slab there also appears to be small, subrectangular break in the sediment projecting ventrally from the larger area, perhaps paralleling a rib. If so, then this break forms more of the border of the supposed abdominal cavity. Also, on the mainslab, a small part of the dark area extends more anteriorly than they indicate, resulting in a more pointed apex to the dark region than the nice gentle curve they indicate. The rest of the "abdominal cavity" just consists of vague, irregular, darkish stains with no particular pattern to them. Who knows what they represent. Perhaps thoroughly degraded abdominal tissue flattened to paper thinness, it will require careful examination and analysis of the specimen to determine so or otherwise, there may never be definitive results. The claim that the specimen shows a croc-like separation between the lung and belly cavities via a septum is absurd, it shows nothing of the sort.
And Paul (2002, p. 342-343):
Direct examination of the main slab shows extensive breakage at the dorsoanterior edge of the carbonized material. The ribs are also broken at this location. This break occurred when a thin layer of sediment dislodged from the main slab and remained attached to its counterpart. The dorsal arrow in Ruben et al (1997a) points directly to the edge of this break. A large, complex set of breakage astride and running perpendicular to the anterocentral edge of the dark tissue is present on both slabs. It is partially filled with cement, which appears to have been colored to better match the dark material. A narrow, irregular zone of dark material appears to lie immediately forward of the crack. The material's anterior extent is further obscured by the presence of a rib, but it appears to extent too far anterodorsally to conform to the smooth convex arc described by Ruben et al . On both slabs, the ventro-anterior border of the dark material is another illusion created by an irregular zone of flakage of the superficial layer of sediment: the ventral arrows in Ruben et al (1997a, 1999) point toward this pseudoborder. . . Because more than 60% of the anterior edge of the dark material consists of breakage and and the preserved edge is irregularly formed, there is no well-formed, semicircular structure present on either slab.

Ruben et al (1999) presented new observations from a new, exceptionally well-preserved theropod specimen (Scipionx) in support of their hypothesis that theropods possessed pelvovisceral pump ventilation. Those observations include a trachea set low in the neck, a high-set colon, and putative pelvodiaphragmatic muscle traces. The specimen is crushed, which complicates inferences about the position of various organs in life. For instance, though in birds the trachea typically is positioned dorsally just under the vertebrae, and in crocodilians it runs more ventrally, the ventrally-placed trachea may well have been displaced relative to life position. In fact, the 10th cervical vertebra clearly has been displaced ventrally (Paul, 2002, p. 348), which may have well resulted in ventral displacement of the trachea. The putative pelvodiaphragmatic muscle traces which Ruben et al call "probable remnants of the diaphragmatic muscles" do not extend to the liver, are poorly preserved, and may represent abdominal muscles such as M. obliquus or M. rectus (ibid). The colon of Scipionyx does appear to be positioned dorsally, more like that of crocodilians than those of modern birds. However, the uncrushed theropod described by Martill et al (2000), the ony other theropod with a preserved colon, appears to be ventrally placed, more like those of birds. Martill et al write (p. 898):
The intestine extends well ventral to the vertebral column in SMNK 2349 PAL, rather than in the dorsal abdominal cavity as restroed for Scipionyx by Ruben et al (1999) by analogy with present-day crocodilians. The reconstruction by Ruben et al was based on the holotype of S. samniticus, which was crushed flat and thus, unlike the uncrushed material described in this paper, cannot be interepreted as reliably preserving the original three-dimensional confuguration of the viscera.
What's more, the theropod described by Martill et al, which they assigned to compsognathidae (same theropod group as Sinosauropteryx), preserved a vacuity behind behind the pubis which may represent a post-pubic air sac, which would directly contradict the argument of Ruben et al. Martill et al write (p. 898):
Based on the preservation of the small vacuities inside the lithified intestinal tract, this large space represents the remnant of an original body cavity. . . This space may have originally been filled with either liquid or air. Conceivably a urinary bladder or ovaries could have occupied this space, but, as in present-day crocodilians, these organs were probably restricted to the abdominal cavity anterior to or in the gap between the sacrum and the pubic apron. The most plausible candidate for a structure occupying this vacuity is a postpubic air sac, which would have extended into the space between the posterior surface of the pubic shaft, the dorsal aspect of the pubic boot, and the ischia. Such an air sac could have been ventilated by a dorsal pneumatic duct passing through the left side of the gap between the sacrum and pubes and a ventral one passing through the distal opening in the pubic apron.


References

Brochu, C.A. and Norell, M.A. 2000. Temporal congruence and the origin of birds. Journal of Vertebrate Paleontology 20(1): 197-200.

Chen et al, 1998. An exceptionally well-preserved theropod dinosaur from the Yixian Formation of China. Nature 391, 147 - 152.

Elzanowski, 2002. Archaeopterygidae (Upper Jurassic of Germany). In: Mesozoic Birds: Above the Heads of Dinosaurs, L. M. Chiappe and L. M. Witmer (eds.), University of California Press, Berkeley.

Evans, S.E. and Milner, A.R. 1994. Middle Jurassic microvertebrate assemblages from the British Isles. In: N.C. Fraser and H.-D. Sues (eds.), In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods, Cambridge University Press, Cambridge, 303-321.

Feduccia, 1994. The great dinosaur debate. Living Bird 13, 29-33.

Feduccia, 1996. The Origin and Evolution of Birds. Yale Univ. Press, New Haven, CT.

Feduccia, 1999. 1,2,3,=2,3,4: accomodating the cladogram. Proceedings of the National Academy of Sciences 96, 4740-4742.

Geist et al, 1997. Implications of soft tissue preservation in the composognathid dinosaur, Sinosauropteryx. Journal of Vertebrate Paleontology 7 (suppl. to no. 3): 48A.

Gibbons, 1997. Plucking the feathered dinosaur. Science 278, 1229-1230.

Jensen and Padian, 1989. Small pterosaurs and dinosaurs from the Uncompahgre fauna (Brushy Basin Member, Jurassic Formation:? Tithonian), Late Jurassic, Western Colorado. Journal of Paleontology 63, 364-373.

Ji et al, 1998. Two feathered dinosaurs from northeastern China. Nature 393, 753-761.

Ji et al, 2001. The distribution of integumentary structures in a feathered dinosaur. Nature 410: 1084-1088.

Leahy, 1998. Response to Ruben et al (1997). Internet posting: http://www.cmnh.org/dinoarch/1998Jan/msg00418.html

Martill et al, 2000. Skeletal remains of a small theropod dinosaur with associated soft structures from the Lower Cretaceous Santana Formation of northeastern Brazil. Canadian Journal Earth Sciences 37, 891-900.

Metcalf and Walker, 1994. A new Bathonian microverterate locality in the English Midlands. In: In: N.C. Fraser and H.-D. Sues (eds.), In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods, Cambridge University Press, Cambridge, 322-331.

Naish, 1998. Ghost lineage. Accessed online 6/25/03 at: http://www.dinosauria.com/jdp/evol/ghost.html

Norell et al, 2002. Palaeontology: 'Modern' feathers on a non-avian dinosaur. Nature 416, 36 - 37.

Paul, 1997. Theropod lung unreality. Internet posting: http://www.cmnh.org/dinoarch/1997Nov/msg00450.html

Paul, 2002. Dinosaurs of the air. Johns Hopkins University Press, Baltimore, Maryland.

Prum, 1999. Development and evolutionary origin of feathers. Journal of Experimental Zoology 285, 291-306.

Prum, 2003. Dinosaurs take to the air. Nature 421, 323-324.

Ruben et al, 1997. Lung structure and ventilation in theropod dinosaurs and early birds. Science 278, 1267-1270.

Ruben et al, 1999. Pulmonary function and metabolic physiology of theropod dinosaurs. Science 283, 514-516.

Schweitzer et al, 1999. Beta-keratin specific immunological reactivity in feather-like structures of the Cretaceous alvarezsaurid, Shuvuuia deserti. Journal of Experimental Zoology (Mol Dev Evol) 285, 146-157.

Sumida and Brochu (2000). Phylogenetic context for the origin of feathers. American Zoologist 40, 486-503.

Witmer, 2002. The debate on avian ancestry: phylogeny, function, and fossils. In: Mesozoic Birds: Above the Heads of Dinosaurs, L. M. Chiappe and L. M. Witmer (eds.), University of California Press, Berkeley, 3-30.

Xu et al, 1999. A therizinosauroid dinosaur with integumentary structures from China. Nature 399, 350-354.

Xu et al, 1999. A dromaeosaurid dinosaur with a filamentous integument from the Yixian Formation of China. Nature 401, 262-266.

Xu et al, 2000. The smallest known nonavian theropod dinosaur. Nature 408, 705 - 708.

Xu et al, 2001. Branched integumental structures in Sinornithosaurus and the origin of feathers. Nature 410, 200 - 204.

Xu et al, 2002. A basal troodontid from the Early Cretaceous of China. Nature 415, 780-784.

Xu et al, 2003. Four-winged dinosaurs from China. Nature 421, 335-340.

Zhang and Zhou, 2000. A primitive enantiornithine bird and the origin of feathers. Science 290, 1955-1959.

Zinke, 1998. Small theropod teeth from the Upper Jurassic coal mine of Guimarota (Portugal). Palaontologische Zeitschrift, 72, 1/2, 179-189.