The Status of Alternatives to the Theropod Hypothesis

There is no question that the theropod origin of birds is by far the most popular hypothesis on avian ancestry. The question then arises,Are there credible alternatives? In most previous reviews (e.g., Ostrom, 1976; Gauthier, 1986; Wit-mer, 1991; Feduccia, 1996; Padian and Chiappe, 1998b), the debate on avian origins was divided into three competing hypotheses: (1) the theropod hypothesis, (2) the crocodylo-morph hypothesis, and (3) the basal archosauriform or "thecodont" hypothesis. However, in recent years it has become apparent that there really are just two major hy potheses: (1) the theropod hypothesis and (2) the "not-theropod" or, as I have termed it previously, "the alternative ancestry hypothesis." Relationship to crocodylomorphs, originally proposed by Walker (1972), seems to have simply faded away in that earlier advocates, such as Martin (1983, 1991), Walker (1990), and Tarsitano (1991), have not renewed their support. The basal archosauriform hypothesis received a significant boost from Welman (1995), who suggested that Euparkeria shares with Archaeopteryx to the exclusion of theropods and crocodylomorphs a large suite of derived characters in the cranial base. This new "thecodont" hypothesis has received to date no additional adherents and was severely challenged by the detailed analysis of Gower and Weber (1998). Thus, for the present, the crocodylo-morph and basal archosauriform hypotheses no longer appear to merit serious consideration.

Indeed, opposition to the theropod origin of birds has become almost exclusively just that, an argument of opposition rather than an argument of advocacy. Criticism is a necessary and appropriate part of the scientific process, and opponents have published a number of papers taking issue with certain of the characters (Martin et al., 1980; Martin, 1983,1997; Tarsitano, 1991; Feduccia, 1996). It is not my goal here to analyze these criticisms or to provide responses, although a few will be touched on in the next section. My main point here is that opponents have sought to destroy but not build in that they have lost sight of the goal of phylogenetically linking birds to actual taxa. The cladis-tic approach is more constructive in that a particular phy-logenetic hypothesis is refuted not simply by criticizing the characters but rather by offering an alternative that better accounts for the available data, that is, an hypothesis that is more parsimonious, a shorter tree. In 1991, I stated: "At present, supporters of relationships of birds with crocodylomorphs, 'thecodonts,' or mammals have failed to produce a competing cladogram, and in this respect the coelurosaurian hypothesis is uncontested" (Witmer, 1991:457).

That statement still stands today, largely because there are no serious alternative phylogenetic hypotheses. For some time, opponents have offered a variety of small, generally poorly preserved Triassic forms as being relevant to the debate (Martin, 1983, 1991, 1997, 1998; Tarsitano, 1985, 1991; Feduccia and Wild, 1993; Feduccia, 1996). These Trias-sic taxa include Megalancosaurus, Cosesaurus, Sclero-mochlus, and Longisquama. These forms do not constitute a clade but are a hodgepodge of basal archosaurs or basal ar-chosauromorphs. Megalancosaurus and Cosesaurus both pertain to the archosauromorph clade Prolacertiformes (Sanz and Lopez-Martinez, 1984; Renesto, 1994). Sclero-mochlus has been thought to be related to a variety of taxa, most commonly pterosaurs and dinosaurs (Padian, 1984; Gauthier, 1986; Benton, 1999). Longisquama has never been subjected to adequate phylogenetic scrutiny; Sharov (1970) placed it in "Pseudosuchia," and Haubold and Buffetaut (1987) agreed, although Charig (1976:9) argued that "the justification for this assignation is obscure." Tarsitano (1991:549) and Feduccia (1996:86) referred to these taxa as "avimorph thecodonts" in that they regarded them as basically birdlike. The resemblances, however, have never been particularly strong or numerous. Authors such as Feduccia, Martin, and Tarsitano generally have not considered these taxa to be truly ancestral to birds but rather as merely representative of what their hypothesized arboreal proavis was like. In other words, these taxa show that there were small, arboreal, quadrupedal animals running around before Archaeopteryx (although it should be pointed out that the preferred habitat and mode of life of these animals are perhaps not as obvious as commonly portrayed).

But even given that such animals existed and are consistent with the arboreal theory for the origin of avian flight, this does not constitute actual evidence relevant to the ancestry of birds. For example, unless Megalancosaurus is being considered as close to the ancestry of birds, whether or not it has a "straplike scapula" or a "birdlike orbit" (Feduccia and Wild, 1993) is of questionable significance. It is conceivable that viable candidates for avian ancestry could emerge from such a nexus of "avimorph" forms, but such hypotheses will continue to be relegated to the fringe unless they are framed in explicit phylogenetic terms and take head-on the theropod hypothesis on its own terms.

Nevertheless, one of these "avimorph" forms captured broad attention when Jones et al. (2000a:2205) pointed to a number of features of the integumentary appendages of Longisquama that led them to conclude that these structures represent "nonavian feathers, probably homologous to those in birds." The most compelling resemblances center on the presence of a calamuslike base wrapped in a presumably epidermal sheath, indicating that the appendages probably developed in a follicle, as is characteristic of feathers and unlike scales. However, their interpretation of the structures coming off the central axis as separate "barbs" seems overly generous at best. These structures unite distally, forming a continuous ribbon around the periphery of the appendage. This distal union is completely unlike the situation in avian feathers, and even if a few tolerably similar examples—all of which are specialized feathers—can be found among birds, it is clearly not the primitive avian condition (Kellner, Chapter 16 in this volume). It seems more likely that the "barbs" identified by Jones et al. (2000a) are in fact plications or corrugations in a continuous structure, which would be more consistent with a modified scale than a feather. Reisz and Sues (2000) also were critical of the hypothesis of Jones et al. (2000a), advancing many of the same arguments just articulated.

Still, Jones et al. (2000a) regarded the structures as feathers probably homologous to those of birds. The implications of such a hypothesis were not explored in the paper. For example, if feathers are a very basal innovation among archosaurs, then this would constitute strong support for the interpretation of the filaments of, say, Sinosauropteryx as feathers (which would be ironic given that Jones and colleagues were such vocal opponents of feathered dinosaurs). But if feathers are a basal character evolving in the Triassic, then where are all the Triassic and Jurassic fossil feathers? There are abundant unequivocal fossil feathers in the Cretaceous, but none prior to those of Archaeopteryx in the Late Jurassic (see Kellner, Chapter 16 in this volume).

The Jones et al. (2000a) paper carefully avoided any statement on the origin of birds, but the authors were very vocal in the associated media furor (e.g., see Stokstad, 2000), arguing that the finding of feathers in Longisquama refuted the theropod hypothesis and that Longisquama itself is "an ideal bird ancestor" (J. A. Ruben quoted in Stokstad, 2000:2124). This example of disparity between scientific and public statements is just the latest in the long history of the debate on avian origins, and it is best to focus on the scientific evidence. In this case, the paper of Jones et al. (2000a) offered no scientific statement on the ancestry of birds. In fact, their claims of homology of the integumentary appendages of Longisquama with avian feathers was an incidental point of the paper, based basically on their opinions and not on a careful phylogenetic treatment, which is the ultimate arbiter of homology. It is fair to say that Jones et al. (2000a) demonstrated that the integumentary appendages of Longisquama are more interesting and unusual than previously thought. Beyond that—and in the absence of a phylogenetic analysis—Longisquama and its appendages are as irrelevant to the debate on the origin of birds as are the other "avimorph" forms.

In sum, at present there remains no credible alternative to maniraptoran theropod dinosaurs for the origin of birds. Previous tangible alternatives (crocodylomorphs, basal ar-chosauriforms such as Euparkeria) have been refuted or summarily dropped because of lack of interest. What has replaced these are intangible "models" that conform to preconceived notions on how bird flight evolved, that is, taxa that, although not truly related to birds, are "much like what we would expect" the true ancestors to be. In some ways, it seems as if the search for real avian relatives has been supplanted by the mission to discredit both the theropod hypothesis and the cladistic methodology that continues to corroborate the hypothesis. Fossils such as Longisquama may someday emerge as more relevant players in the debate, but if the media hype surrounding the Jones et al. (2000a) paper was any indication, even Longisquama will be just another attempt to develop a rhetorical weapon to attack the theropod hypothesis and cladistics.

0 0

Post a comment