The origin of birds remains uncertain, as does the status of feathers on dinosaurs. Review of "Riddle of the Feathered Dragons", by Alan Feduccia.
As regular readers can surmise, I was raised (scientifically) in an empirical, experimental tradition- that of molecular biology. In that field there is little drama, since any dispute can be taken back to the lab for adjudication. No titanic battles of conflicting interpretations happen, and extremely high standards pervade the field, since any lapse is easily discovered and replicated. Despite the dominant position of molecular biology in the major journals, due to its high productivity, it is thus rarely in the public spotlight. It has been a bit of a culture shock to realize that other areas of science have significantly different standards and epistemology. Many fields (such as astronomy or paleontology) are at heart observational rather than experimental, or have other restrictions or conflicts (medical science, nutrition studies) that impair their ability to find the truth, leading to a great deal more interpretation, drama, and sometimes, rampant speculation.
Paleontology, and the study of the past in general, has an intrinsic lack of data. If the fossils are missing, what can one do but to wonder and speculate what could have happened during that gap? And when fossils do turn up, they still lack alot of information about their unfortunate contributor- they are only the bare bones, after all. They may be in bad condition and particularly hard to interpret. Whole genera or above may be represented by a tooth or single bone. Millions of years may go by with nothing to show for it. No wonder speculation fills the gap- but is that science? Incidentally, I have to thank the Discovery Insitute, with its keen nose for scientific controversy, for pointing me to today's author, who disputes the now-conventional view that birds arose from dinosaurs. While Alan Feduccia has nothing to do with Creationism and its offshoots, and is a perfectly respectable paleontologist, he is, in the course of at least four books on the subject, (of which this is the third), clearly frustrated with the reigning interpretations of his field, which has jumped to what he regards as unwarranted conclusions that have led to a flurry of portrayals of feathered dinosaurs.
 |
| The Archaeopteryx fossil, Berlin specimen, which dates to roughly 150 million years ago. |
The first bird, more or less, was Archaeopteryx. Its Berlin fossil, found about 1875, and dating to roughly 150 million years ago (mya) is perhaps the most beautiful, and informative, fossil ever found- a complete bird, with fully spread flight feathers on its arms and legs, a long tail, claws on its hands,and teeth in its mouth. It had the precisely the in-between characteristics of both reptiles and birds that gave immediate validation to Charles Darwin's theory of evolution by gradual change and natural selection. But where did it come from? That is the big question. While a great many other fossil birds have been found, none substantially predate Archaeopteryx, (other than perhaps Anchiornis, very similar to Archaeopterix, and dating to roughly 160 mya), and thus we really do not know (as yet) from fossils how birds originated.
Feathers are one diagnostic feature in this lineage. Archaeopterix and many later birds found in China and Mongolia from the Cretaceous have feathers, clearly marking them as birds and as lineally related. But other allied fossils have been found, nominally described as dinosaurs, which are described to have feathers as well. One is shown below.
 |
| Fossil of Sinosauropteryx, which dates roughly to 124 million years ago. |
 |
| Closeup of the hair-like impressions on the tail of Sinosauropteryx. |
Whether these structures are feathers, or related to them, is quite debatable. They look more like hairs, and Feduccia claims that they do not even occur outside the body wall. Some experimentation by others has shown that sub-dermal collagen can form this kind of hair-like fuzz during some forms of decay and fossilization, given proper squashing. Current conventional wisdom, however, describes them as filament-like feathers used for insulation or display. My take, looking closely at these pictures, is that they are not feathers, but are outside the body wall, which, on the tail certainly, would have been very close to the bone. Additionally, as these specimens are all rather late, they could easily be descended from birds, while being large and flightless. Feduccia points out that while land-based animals have never gained/regained flight, flightlessness has evolved many times through the bird lineages. Similarly, extensive lineages of secondarily flightless birds may have developed in the Mesozoic, that conventional paleontologists call dinosaurs, (often with feathers), and posit as evidence that the reverse happened- that birds evolved from dinosaurs. For example, the conventional view of dinosaurs and birds draws on many later fossils from the Cretaceous (such as Deinonychus), which had both bird-like and dinosaur-like features: the "raptors".
Another character at issue is flight itself. If birds are basal- that is, they arose prior to or separately from the other dinosaurs- then they could easily have developed from small arboreal lizards that learned to glide from place to place. On the other hand, dinosaurs are all relatively large and bipedal. So conventional paleontologists have labored to come up with ways that flight could have developed "from the ground up". Such theories as insect trapping by nascent small wings, or occasional tree climbing with tiny wings, to escape predators, have been invoked as rationales for feathers and wings to develop in terrestraial bipedal dinosaurs. Feduccia counters that in the whole history of flight, all animals (birds, bats, squirrels, others) have developed flight from gliding, not from the ground up. Indeed, there are countless flightless birds, and none of them have resumed flight, despite presumably having much of the genetic wherewithal to do so.
Given patchy data, the leading method to make sense of it and organize organisms from the fossil record into a phylogenetic story is the cladistic method. Practitioners choose a wide range of "characters", (such as the lengths, angles, holes, and other morphologies in the available bones) and tabluate their values from all the proposed species. Then they can mathematically just total up who is more distant from whom. Feducci emphasizes that this is an excellent method for ordering closely related genera and species. But over the long run, evolution repeats itself alot, making numerous flightless birds, for example, or similarly shaped swimming animals, not all of which are as closely related as they might look morphologically. Cladistics is a classic case of garbage-in-garbage-out analysis, and has routinely been overturned by molecular evidence when, among extant species, genomic data is available. Sadly, genomic data is not available for the fossils from the Mesozoic (the age of the dinosaurs, which encompasses, in order, the Triassic (245 mya to 208 mya), the Jurassic (208 to 144 mya) and the Cretaceous (144 to 65 mya) periods), nor from any living descendants of the dinosaurs... other than their putative decendants, birds.
As an aside, molecular phylogenies are also at heart cladistic in their theory and method. They just have a lot more "characters"- i.e. the letters of the DNA sequences in homologous / aligned sequences. But even more importantly, since a large proportion of these characters are neutral, (to natural selection), and thus vary (in sort-of clock-like fashion) no matter what convergent evolution might happen morphologically, molecular phylogenies can easily resolve difficult questions of phylogeny on the short to medium geologic terms. When it comes to the deepest phylogenies, however, going over a billion years, neutral characters become wholly useless due to homogenization by the vast times that have passed, so for such time periods these methods become less incisive.
 |
| Crude cladogram illustrating the alternative hypotheses- that birds are descended from theropod dinosaurs, or that birds arise from a basal lineage of their own, directly from the common stem of archosaurs. In the latter hypothesis, numerous bird-like lineages currently construed as dinosaurs might be secondarily flightless birds. |
It is cladistics (along with other evidence) that has enshrined birds within the dinosaur lineage, finding that theropods came first, and the avians came later on. (With a contrasting view, and a critique of the contrasting view.) Theropods and birds are certainly similar, compared to their crocodilian / archosaur antecedents. They are bipedal, with similar hip structures, neck structures, and hands/feet reduced from five to three toes. But if much of what we take to be the dinosaurs (those with feathers and the whole so-called "raptor" class), are actually secondarily flightless birds, then one can make a lot of sense of some of these similarities, while casting the origin of birds quite a bit father back in time, more or less co-incident with the origin of true dinosaurs. Such as in the diagram above.
The problem with all this is again time. The early Jurassic and Triassic, amounting to almost one hundred million years before Archaeopterix, provide a lot of evidence for dinosaurs. They first appear roughly 240 mya, and flourish after the major exinction event that ended the Triassic, at 201 mya. The stark lack of evidence for birds, and widespread evidence for dinosaurs, including the lineage (theropods) that are most related to birds, suggests strongly that birds did not originate back in the Triassic, in parallel with the core dinosaur lineages. It suggests, rather, that among the many theropod dinosaurs during the ten or twenty million years before Archaeopterix were some small enough to take to the trees, grow longer arms, and be in position for flight. There were doubtless plenty of insects up there, at least until just about this time of the late-Jurassic, when birds started to eat them! Fossil record gaps are treacherous things, but this one indicates strongly that birds evolved in the middle Jurassic, along with (and within) the wider adaptive radiation of dinosaurs.
"Yet Archaeopteryx is still the classic urvogel- the oldest well-studied bird yet discovered, perhaps some 25 or more million years older than most of the Early Cretaceous Chinese fossils. As we saw in chapter 3, the Solnhofen urvogel is a mosaic of reptilian and avian features, a true bird, and the more it is studied, the more and more birdlike it is revealed to be. Ignoring the element of geologic time, however, many paleontologists have proposed that the Liaoning fossils provide evidence for all the stages of the evolution not only of birds and bird flight but also of feathers, from fiberlike protofeathers to pennaceous, asymmetrical flight remiges. Such a claim is remarkable and would be astounding in any fauna, but is especially so for a fauna so temporally removed from the time of avian origins, presumably before the Middle Jurassic and perhaps well back into the Triassic.
University of Pennsylvania paleontologist Peter Dodson, remarking on the inadequacies of cladistic methodology, tells us: 'To maintain that the problem of the origin of birds has been solved when the fossil record of the Middle or Late Jurassic bird ancestors is nearly a complete blank is completely absurd. The contemporary obsession with readily available computer-assisted algorithms that yield seemingly precise results that obviate the need for clear-headed analysis diverts attention away fron the effort that is needed to discover the very fossils that may be the true ancestors of birds. When such fossils are found, will cladistics be able to recognize them? Probably not.'"
Feducci makes a lot of insightful points and hits some sensitive marks, in addition to all the trash-talk. Cladistics has problems, hairs are not feathers, and Cretaceous birds don't tell us much about the evolution of bird flight, which doubtless began as gliding between trees tens of millions of years earlier. And he is right that the hunt for clear antecedents of Archaeopterix, whether far in the past or near, should be the focus of this field. But overall, it is hard to fully credit the "birds early" story.
