The Significance of Protoavis

A particularly difficult question is whether this ensemble cast should rightly include the Texas fossils known as P.tex-ensis. These fossils, whose discovery was announced only in 1986, have had a troubled and controversial history. In a long series of published works, Sankar Chatterjee (1987, 1988, 1991,1995,1997a, 1998b, 1999) argued that the Protoavis fossils are not only those of a bird but from a bird that lived 75 million years before Archaeopteryx! The fossils are generally attributed to two individuals excavated from the Early Norian Cooper Member of the Dockum Group of western Texas (Chatterjee, 1991), although other material from a different formation and county also was later referred to P.tex-ensis (Chatterjee, 1995,1997a, 1999; justification for this referral has not been presented, and I will not consider that material here). Although the report of any new Mesozoic bird is greeted with great interest, the reception of Protoavis was unique, largely because of the implications that a Trias-sic bird holds. Perhaps surprisingly, despite the great age of the fossils, Chatterjee has never argued for any major changes in the general notion of avian ancestry from dro-maeosaurlike coelurosaurian dinosaurs; in other words, Archaeopteryx remains the basal bird, and the Ostrom/ Gauthier hypothesis of theropod relationships is not challenged. The irony that emerges is that Protoavis perhaps should have relatively little relevance for the origin of birds in that, according to Chatterjee's cladograms, Protoavis is nested well within Aves.

Skepticism about the avian status of Protoavis was immediate and did not necessarily follow along lines of allegiance to any particular theory of avian origins. For example, Feduccia (1996) and Martin (1998), on the one hand, and Ostrom (1987, 1991, 1996), Wellnhofer (1992, 1994), Chiappe (1995, 1998), and Sereno (1997b, 1999a), on the other hand, have all expressed doubt that the fossils of Protoavis are adequate to substantiate the claim of a Triassic bird. At the same time, Chatterjee has had some supporters, including Peters (1994), Kurochkin (1995), and Bock (1997, 1999). Why the controversy? A fuller critical appraisal of the status of Protoavis is presented elsewhere (Witmer, 2001), but a brief analysis is presented here.

Detractors of Protoavis have raised a variety of complaints, the most important of which relate to the taphon-omy of the specimens and their preservation and preparation. With regard to the taphonomy, there is a widespread concern that P. texensis is a chimera, that is, a mixture of more than one species. Indeed, the quarry from which the specimens derive is a multispecific bonebed that has recorded many taxa (Chatterjee, 1985), and thus mixing is a possibility, as has already been suggested for other taxa from the same quarry (e.g., Postosuchus kirkpatricki; Long and Murry, 1995). Chatterjee (1991,1998b) has steadfastly maintained the association of the holotype and paratype skeletons of Protoavis, and Kurochkin (1995) offered his support based on his study of the original material. Nevertheless, the specimens were collected inadvertently while removing overburden with a jackhammer, and hence we can never be completely sure of the taphonomic setting. The possibility that Protoavis is a composite of several species is commonly voiced but, even if true, does not rule out the chance that some of the included bones are avian (Witmer, 1997c). However, as Chiappe (1998) correctly pointed out, the chimera problem presents itself most insidiously during phylogenetic analysis, the ultimate arbiter of avian origins, in that the mixture of taxa means a mixture of characters, all of which leads to phylogenetic nonsense. Thus, the taphonomic question— Is Protoavis a species or a fauna?—is a critical one, and one that is not likely to go away until new material is discovered.

And, according to many of those who have studied the specimens, new material is needed desperately. In other words, a major concern has been that the Protoavis specimens are simply too poorly preserved, too scrappy, to be diagnostic. Unlike Archaeopteryx and the Cretaceous birds from Spain and China, Protoavis is not a "slab animal"; that is, it is not preserved in situ, but rather all the bones have been prepared free of the matrix. Thus, without the aid of the positional information that slab animals preserve, the identification of isolated elements is difficult and can lead to widely different interpretations. The other side of this coin is that all sides of the elements are available for study rather than being half entombed in stone, as is the case for slab birds such as Archaeopteryx. Nevertheless, it has been difficult to confirm not only many of the structures but even some of the bone identifications made by Chatterjee (Witmer, 2001b). Perhaps Padian and Chiappe (1998^13) best characterized the situation by noting that the "material has become a paleontological Rorschach test of one's training, theoretical bias, and predisposition." Coupled with this is the problem that the specimens are extensively reconstructed with plaster and epoxy, and it often seems that the published descriptions are of these reconstructed composites rather than of the fossils themselves.

But ultimately—even given the gravity of these and other concerns—it comes down to the fossils and their structures. Are there clearly interpretable anatomical clues revealing the phylogenetic relationships of the beast? Again, a more comprehensive skeletal analysis is presented elsewhere (Witmer, 2001b), but it is worthwhile to examine here a few of the more important anatomical systems. For Chatterjee (1991, 1995, 1997a, 1998b), the skull, particularly the temporal region, is the most critical, because he regarded Protoavis as possessing the ornithurine condition: that is, loss of the postorbital bone, leading to confluence of the orbit, dorsotemporal fenestra, and laterotemporal fenestra. Given that Archaeopteryx, Confuciusornis sanctus (Peters and Ji, 1998; Chiappe et al., 1999; Hou et al., 1999), and at least some enantiornithines (e.g., the Catalan hatchling, Sanz et al., 1997; Protopteryx fengningensis, Zhang and Zhou, 2000) retain both the postorbital bone and its contact with the squamosal, presence of the advanced ornithurine condition in Protoavis indeed would be highly significant and would, in fact, argue for a higher phylogenetic position within birds than that advocated by Chatterjee (1998b). Unfortunately, the temporal region of the holotype skull has been assembled from disarticulated pieces, and thus the identity and positions of elements are not certain. The most telling clues are found in the squamosal and quadrate; the absence of the postorbital is negative evidence and hence hard to evaluate. The squamosal identification is key, because the element would lack an articular surface for the postorbital, which implies absence of the postorbital bone itself. Similarly, the putative quadrates are important, because they would be drastically modified along the lines of birds, presumably, according to Chatterjee, in association with avian craniofacial kinesis. The squamosal identification is defensible in that the element in question has a cotyla that could receive a quadrate, but other interpretations are possible. The quadrate identifications are less certain (see Witmer, 2001b). My general impression of the reconstructed temporal region is that Chatterjee's view is understandable and justifiable but is not sufficiently clear to merit drawing firm phylogenetic conclusions.

Without question, the braincase is the most easily interpreted part of the skull, at least with respect to bone identifications. It is doubtful that there are any indisputably avian apomorphies in the braincase. However, it is indeed the braincase of a coelurosaur in that it possesses cranial pneumatic recesses (including the caudal tympanic recess, which thus far is not known outside Coelurosauria), a large cere-bellar auricular fossa, a metotic strut, and a vagal canal opening onto the occiput (Chatterjee, 1991,1998b; Witmer, i997d, 2001b), and thus the braincase may pertain to the oldest known coelurosaur.

Postcranially, little is unambiguously avian. Exceptions are the cervical vertebrae, which are truly heterocoelous, if only incipiently so; have prominent ventral processes (hy-papophyses); and have large vertebral foramina. Of course, heterocoely has a fairly homoplastic distribution within birds (Martin, 1983; Chiappe, 1996), and Protoavis is not as heterocoelous as Hesperornis or most neornithines, but the vertebral structure nevertheless represents one of the few bona fide avian suites of Protoavis. There are many problems in the thoracic appendage (Witmer, 2001b), not least of which is the coracoid, which, although having a generally advanced avian shape, seems positively minuscule in comparison with the rest of the skeleton. I cannot confirm the remigial papillae on the ulna or manus. In fact, the identification of the four-digit manus itself has been called into question, with Sereno (1997b) regarding it as the foot of an archosaur. The pelvic appendage likewise is not particularly birdlike. Perhaps the most avian feature is a medial fossa within the os coxae regarded by Chatterjee (1995,1998b) as a renal fossa; however, no other Mesozoic bird has a renal fossa, and the structure in Protoavis differs somewhat from that in neornithines (Witmer, 2001b).

It is probably fair to state that the case for the avian status of P. texensis is not as clear as generally portrayed by Chatterjee. The temporal configuration and vertebral morphology might argue for a position near Ornithurae, yet the long tail, the archaic ankle, the four-digit manus, and other features would argue for a basal position, probably well outside Aves. The braincase is more or less coelurosaurian. The taphonomic problems and the possibility that this is a chimera may make this an intractable problem. An option is simply to take Chatterjee's analyses at face value and proceed (as done by Dyke and Thorley, 1998), but this seems naive at best. Critical study of the original material is absolutely necessary, but even here, in the absence of newly collected material of known association, firm conclusions will likely be elusive.

It would thus seem that Protoavis bears little significance for solving the riddle of the origin of birds. The specimens themselves are problematic, and so we rightly should be skeptical. But even assuming that Chatterjee has interpreted them 100% correctly, Protoavis would have little impact on the phylogenetic pattern of avian origins (Witmer, 1997c, 2001b; Dyke and Thorley, 1998). Then why the often vitriolic controversy? The reasons are complex, but probably a major component is that Chatterjee's acceptance of the Ostrom/Gauthier orthodoxy would require that virtually all theropod cladogenesis had taken place well back in the Tri-assic, at least in the Norian, if not earlier—that is, right at the very dawning of the dinosaurs. I have elsewhere (Witmer, 2001b) referred to this (somewhat whimsically) as the "Norian Explosion." The problem is that we have no real evidence of such an explosion: no Norian tyrannosaurids, no Norian oviraptorosaurs, etc. Thus, many people simply have not accepted this proposition. This incongruity has not gone unnoticed and has been exploited by opponents of the idea of theropod relationships. For example, Martin (1988), Tarsitano (1991), and Bock (1997) were receptive to the avian status of Protoavis and pointed out that a Triassic bird would essentially disprove the prevailing notion of thero-pod relationships.

But a long view is appropriate. The origins of many theropod groups are constantly being pushed back further in time. Therizinosauroids have been reported from the Early Jurassic of China (Zhao and Xu, 1998; Xu et al., 2001a), and Chatterjee (1993) reported an ornithomimosaur from the Late Triassic of Texas, although such claims generally are controversial (Rauhut, 1997). As mentioned, Protoavis itself represents a temporal range extension for Coelurosauria. Whether the idea of a Triassic bird will ever be more palatable is hard to predict, but stranger things have happened in the history of science, and Protoavis may yet prove to be a key player. For the present, however, it is probably both prudent and justifiable to minimize the role that Protoavis plays in any discussions of avian ancestry.

+1 0

Post a comment