Palaeos Vertebrates: Aves

This page covers the first dinosaurs that we would clearly recognize as birds: feathered creatures who flew. The earliest archaeornithines scarcely recognizable as birds when the feathers are removed. In fact, fossils of Archaeopteryx have repeatedly been mistaken for small coelurosaurian dinosaurs. Under the skin they have almost none of the adaptations we associate with birds. In Archaeopteryx, the bones of the skull table are not crowded onto the posterior skull, there is no beak, there is a full set of sharp teeth, the coracoid is not enlarged, the sternum is somewhat enlarged but lacks a keel (carina), the bones of the ankle and wrist are not fused, the pygostyle and notarium (fused tail and back vertebrae, respectively) are absent. The entire flight apparatus of the arms is large, but not so very different from that of other dinosaurs.

In fact, there is still some uncertainty about whether Archaeopteryx was even capable of flight. Full, powered flight -- as suggested by the presence of highly specialized pectoral structures -- actually seems to have been a relatively late development compared to most of the other avian adaptations. So we may speculate that their usual mode of locomotion was something different: jumping, gliding, running, or even diving. The commitment to flight seems to have been a gradual matter, and perhaps all of these strategies played a part in its development.

The Sequence of Flight Adaptations

Garner et al. (1999) is a new and very worthwhile effort at explaining the phenomenon of flying birds. The paper comes very close to being a classic, "must-read" study. Before reciting the many excellent features of the paper, I will whine briefly about style.

What is one to do about the English language when English-born Oxford dons, writing in the Proceedings of the Royal Society of London, cannot seem to get grammatical numbers to agree or to apply verb tenses in the appropriate mode? Beyond purely stylistic questions, what does one do with a sentence such as: "At its simplest level, vane asymmetry is necessary for self-stability of any feather the rachis of which is oriented across the airflow." In addition to stylistic awkwardness, the analytical framework is unclear. What, exactly, does "its" refer to? Why is this the "simplest level"? What would be a more complex level? Would any secondary effects offset the described effect? What is "self-stability" as opposed to simple stability? This mode of writing is full of ambiguities, uncertain references and strained language. Perhaps the authors meant: "The effects of vane asymmetry are often complex. However, we may consider a simple model in which the feather is assumed to be an isolated, flat, rigid object in a uniform airstream. In this simplest case, the following rule applies: if the rachis of the feather is oriented across the direction of air flow, then vane asymmetry is necessary to maintain aerodynamic stability."

Then again, perhaps the authors meant something entirely different. One can't really tell. If this "translation" is correct, the paper ignores a key issue: should feathers be considered as isolated aerodynamic units? How realistic are these assumptions? For what it may be worth, I personally believe this is a very useful and appropriate model; but fuzzy writing, as in the original, only leaves a vague impression of fuzzy thinking without allowing the reader to grapple with the real issues.

With all that said, this is still a great paper. The authors analyze each of the major theories of the origin of flight by showing how each has implications for the sequence in which flight characters evolved. They compare these results to the stratigraphic record and to the "predictions" made by a new model in which powered flight developed from ambush predation.

Garner et al. probably doom the acceptance of their own theory by naming it the "Pouncing Proavis" model. Still, the idea is sensible enough. They postulate that the first birds were small dinosaurs which practiced ambush predation in the manner of some felines: hiding in an elevated position not necessarily a tree) and jumping down onto prey. The authors do not address the issue in precisely this way, but it may be significant that the ancestral theropod was a biped. Unlike a cat, it makes a great deal of difference how a biped lands, because a biped can be thrown off balance much more easily than a quadruped. In any event, Garner suggests that these small theropods developed plumage as a way of maintaining drag-based aerodynamic control -- not to fly or glide, but simply to land at the right place at the right time in the right orientation.

The study compares this Leaping Lizard approach to the standard models -- the arboreal glider, the cursorial flapper, and run-jump-glide hypotheses -- with regard to (a) placement of flight feathers; (b) feather asymmetry; (c) adaptations to minimize weight and (d) avian (as opposed to theropod) posture and gait. They conclude that only the Saltatory Saurischian correctly predicts the observed evolutionary sequence. That sequence, they assert, is that fully functional feathers appear first, that flight surfaces appeared first distally and later proximally along the arm, that vane asymmetry became pronounced only after the formation of a complete wing and that weight reduction and avian posture came last, after the beginning of powered flight.

Allowing for the still spotty stratigraphic record, this is quite reasonable. At the least, the Aerodynamic Archosaur is more consistent with the fossils than the cursorial theories, which require that powered flight occur almost all at once. The same may be said of the arboreal glider theory, which involves a suite of adaptations which are as specialized as, but quite different from, the adaptations needed for flight. It also makes sense of the alvarezsaurs, oviraptors, Caudipteryx, and other large, (probably) feathered, flightless, bird-like things. If Garner & Co. are correct these are simply early Dropping Dinos without real flight surfaces. Instead, they merely used distal feathers, without vane asymmetry, to create enough drag to control attitude during a leap or fall.

Over time, the Vaulting Velociraptors would likely evolve a set of stylized arm movements during the leap to give greater range and control. Given fully formed feathers already perfected for drag-based control, the step to powered flight would not be difficult. Ultimately, the Bounding Bipeds would face selection for weight reduction and weight redistribution to complete the transition to the ornithurine condition. ATW001029

The Bird is the Word

Few now remain who will recall the horrid tune of this name from those dark days of the mid-1960's when music had reached a bleak, abyssal plain not seen again until the flaccid benthic ooze passed off as song today. From those few, perpetually scarred, survivors, I ask forgiveness for dredging up so painful a memory. For the rest, this brief note will relate only to the nomenclature used to mean "bird."

Back in the glory days of the late 1980's when cladistics was starting to remake the paleontological world, Jacques Gauthier and his colleagues were writing critical papers redefining all of the major taxa. A remarkable number of these definitions have not only stood the test of time, but have been consistently useful landmarks -- with one exception: the taxon Aves.

Almost no one in 1986 knew that, in little more than a decade, the origin of birds would be so clearly fixed in the Theropoda that there could essentially be no rational debate. Accordingly, it was appropriate to do as Gauthier did even if he was one of the few who did know better) and simply define Aves by reference to the living crown group: as the last common ancestor of all living birds and its descendants, thereby avoiding a pointless debate about the appropriate taxonomic anchor for extinct forms.

Unfortunately, this makes Aves and Neornithes completely redundant. Worse, it is now the placement of this neornithine node that is vague and difficult to find, either in time or among the available bones. The timing and source of bird origins, and the first well-known remains too avian to dispute, are firmly fixed in Archaeopteryx. Certainly, other bird-like fossils will be found -- either earlier or more primitive. But no fossil will ever replace Archaeopteryx, the Urvogel, as a historical and cultural icon of "birdness," and this is the sort of stuff that ought to inform our definitions.

In 1992, Luis Chiappe attempted to introduce some reason to the area, by redefining Aves as Archaeopteryx + Neornithes. This would have been appropriate. For taxa beyond the genus level, there is no iron rule of priority. But madness had been loosed in the world. Rather than adopting Chiappe's useful suggestion, many workers have taken up an even less salubrious proposal of Gauthier: the taxon Avialae = Passer [sparrow] > (Dromaeosaurus or Troodon). All scientific matters aside, what possible demon could have inspired Gauthier to adopt two anchor taxa more likely to cause international spelling confusion or search and indexing meltdowns? Consider: Dromaeosaurus, Dromæosaurus or Dromeosaurus? -- all depending on one's relationship to the Atlantic Ocean and routinely available font. Is it Troodon, Tröodon, or Tröödon? Only an ancient Greek (or perhaps Laöcoön, a Trojan) knows for sure. Finally, just to add to the fun, yet a third group of workers uses Avialae, but in a sense identical to Chiappe's Aves. See, e.g., Novas (1997).

Although we confess to be fighting a rearguard action, we will use Chiappe's terminology until, perhaps inevitably, our cowardice and spineless conformity smother what lingering sparks of principled rationality may remain. ATW020813.

Characters: $ ventral margin of nares dorsal to maxilla [M+02]; $ quadratojugal (posterior) process of jugal tapered [M+02]; $? posteroventral process of dentary long & shallow, reachong posterior margin of external mandibular fenestra [M+02]; $ Teeth not serrated [C95]; $ teeth with constrictions between crown & base [C95]; $ <26 caudal vertbrae [C95]; $ caudals with short prezygapophyses [C95]; arms long; coracoid not greatly expanded; $ reversed hallux [C95]; $ asymmetrical remiges (!?).

Pigeon-sized extremely primitive bird, probably capable of flight. Skull moderately long and pointed; maxilla, premaxilla and dentary toothed; teeth sparse, small, with constriction between crown & roots and not laterally flattened or serrated; interdental plates present; nasals are primary rostral bones; nares large; antorbital fossa with 2 fenestrae(?), very large; orbit very large; jugal reduced almost to bar connecting maxilla and quadrate; frontal expanded laterally; probable very large brain; parietal markedly narrower; quadrate may be moveable; large surangular forming upper surface of posterior portion of lower jaw; angular forms lower surface; most other dermal skull components present but strongly reduced; small, posteriorly-directed cervical ribs; notarium absent; 4 sacral vertebrae; tail longer than dorsal spine; sternum absent; coracoid short; scapula long and narrow; furcula large; tridactyl wings with short, sharp, curved claws; radius & ulna unfused; flexible carpus; pelvis long and fairly shallow; angle of pubes disputed, but certainly not propubic; knee hinge-like; femur < tibia; fibula reduced; distal tarsals fused to metatarsals, which are partially fused proximally (i.e. tarsometatarsus present); digits anisodactyl.

Note: Note how few of the Neornithine bird adaptations are found in Archaeopteryx. See also image at entry for Ornithurae, comparing skulls of Columba and Archaeopteryx.

Aves

Taxa on This Page

The First Birds

The Sequence of Flight Adaptations

The Bird is the Word