Friday, 4 May 2007

The Evolution of Wings

Image of Archaeopteryx and Microraptor by Jennifer (Jenna) Bolen, UC Santa Cruz.

This is a sequel to my earlier blog entry on an exhibition of Chinese dinosaurs, with a PhD student in attendance giving visitors some free lessons in paleontology and dinosaur evolution.

The thing I found most interesting was his discussion of the probable evolutionary development of wings.

Background: what use is half a wing?

It's a common argument from anti-evolutionists that evolution of specialized structures like a wing is impossible, because the initial stages would serve no useful function and so could not possibly be selected. As creationists sometimes ask: "What use is half a wing?". This is number CB921.2 in the Index to Creationist Claims.

The index provides answers -- there are many things that a "half wing" could be good for.

The question incorporates a misconception. Individual organs don't evolve in isolation. They are part of a complete living organism, and the whole organism evolves together. At any point along a successful lineage, you are bound to have effective fully functional organisms, well adapted to their particular lifestyle. There may be positive selective pressures giving some direction to the changes that accumulate along the lineage, but successful variants are those that incorporate changes with an immediate benefit. Just grafting a wing that is complete for an eagle onto a Velociraptor dinosaur would not work; because the rest of the Velociraptor's body is adapted for a different lifestyle to that of the soaring eagle. If a particular organism has "half a wing", than that is most likely pretty much the right kind of wing for that organism.

Another consideration is functional shift. The ancestral forms of any organ may have been adapted for quite different purposes, and over time their primary function shifts from one role to another. If you could replace "half a wing" with a "full wing", then you may actually make matters worse, since the change removed some other other important function that is vital to the ancestral form.

A common argument from the evolutionist side, in attempting to fill out an entire hypothetical picture of evolution, is that even the tiniest fraction of a wing might give at least SOME benefit, perhaps in breaking a fall, and so this is leverage for evolution. This response runs the risk of contributing to the misunderstanding of thinking that the original form is properly understood as a half-wing.

Natural selection is, for the most part, a conservative force, acting to weed out damaging mutations and to retain existing function. You can have positive selection for a gradual change, but this really only works if there is also a strong selection against losing change in the reverse direction.

Functional shift

The relevance of this to wings is that if an organism has some proto-wing which has only the tiniest contribution as a wing, then this is not likely to be a sufficient explanation for how a wing can evolve. You might think that even 1% of a wing still gives some leverage to evolution. Actually, it quite probably does not.

This point is well argued in the context of insects by Kingsolver and Koehl, Selective Factors in the Evolution of Insect Wings (Annu. Rev. Entomol. 1994.39:425-5). It was found that the tiny contribution of the earliest proto-wings to flight was not able to account for the kinds of selective pressures that could plausibly explain wings. The solution is found in functional shift: wings originally were selected for a different function – thermoregulation. In this case, the initial thermo-regulators could still give a substantial benefit to a well-adapted organism, despite being too small for aerodynamic qualities to be of much importance. This allowed leverage for evolution to drive to greater surface areas for more effective thermoregulation; and co-incidentally this was just the kind of change that gave them also the quality of aerodynamic function.

Furthermore: at small sizes, an increase in size of wings had a significant effect on thermoregulation but little effect on aerodynamics. But as the wing size became larger, further increases in size had little further impact on thermoregulation but a substantial impact on aerodynamics. This is illustrated in the following figure.

"The thermoregulatory (upper curve) and aerodynamic (lower curve) Advantages for increasing wing length in insects. Note that thermodynamic benefits accrue rapidly when the wing is very small (too small for flight), but scarcely increase at all for wings of larger size. Aerodynamic advantages, on the other hand, are insignificant for small size, but increase rapidly at larger wing dimensions, just as the thermodynamic benefits cease."

(Reproduced from Not Necessarily a Wing, by Stephen Jay Gould.)

Back to dinosaurs: Sinornithosaurys millenii

One of the fossils recently found in China is Sinornithosaurus milleni. Sinornithosaurus is a dromaeosaur, part of a family of agile meat eating dinosaurs. Another famous dromaeosaur is the Velociraptor, which had a starring role in Jurassic Park. In fact the models in the movie were based on Deinonychus; and since then we have learned that these animals had feathers. A fossil Deinonychus is illustrated below, and an artist's impression for Sinornithosaurus graced my earlier blog article on the subject.

Sinornithosaurus was not a bird, and it could not fly. Its forearms are not wings; but the fossil shows two crucial "preadaptions" for bird wings. Sinornithosaurus has feathers, or proto-feathers. It seems probable that these structures developed mainly for insulation. The feathers were not adapted for flight, but would have been more like a kind of soft coat of down. Such a coat can help trap a layer of air next to the skin. We can't prove definitively the various stages, but it is plain that selection for a capacity to give insulation will lead to structures that trap air, and this in turn gives an increased capacity to push on air, or develop an aerodynamic effect.

The other significant preadaption are the long forearms, and their range of movement. Many dinosaurs had very small forearms; Tyrannosaurus Rex being a classic example. Such forearms could not possibly give rise to wings, and small changes in a stunted forearm will have no selective advantage for wings.

However, there may be selective advantage for a longer forearm in dinosaurs that use their arm for an attack. Sinornithosaurus had very long arms, and furthermore they had an extended range of movement. This was a matter of special interest to the PhD student at the Chinese Dinosaurs exhibition. The skeleton indicates that this dinosaur was able to move its arms above the head, and would have been able to bring them down for a slashing attack on their prey. This movement needs to be in place before there can be selection for wings; but the selection force for developing this movement is a capacity to attack. They had long legs, and were probably well suited to run down and leap on their victims.

Dinosaur dreaming

Fossils of Buitreraptor (foreground) and Deinonychus (background) at the Field Museum of Natural History in Chicago, Illinois. Photo by C. Horwitz, available at wikipedia.

We don't know all the stages in the ancestry of birds, but the transitional fossils we see indicate how the initial dinosaur form merges into that of birds. Sinornithosaurus, and many other fossils from the Yixian formations in Northern China and elsewhere, give an insight into now extinct forms of life. It's not possible to prove how selection worked on these creatures; but that they were as subject to selection as any other living thing is plain.

The relationships between these dinosaurs and birds have been emerging as we continue to find fossils of bird-like dinosaurs, and dinosaur-like birds; all fully functional living organisms in their own right, but with a mix of characters that make them splendid examples of "transitional" fossils. Precisely how the various processes of evolution worked out, and the full details of lineage, will never be given as a complete finished picture.

The glimpses we have do show the kinds of forms that gave rise to modern birds, and they show ample scope for selection to apply by small changes in form as an arm becomes longer, and with a wider range of movement, and with a coat of feathers, and with a benefit to a sudden spurt of speed to jump and leap out at prey. It may not have been exactly like that; and maybe other processes and effects will be found; and quite possibly there are significant contributory factors that will never be revealed.

Imagination and speculation are important qualities for a scientist. These ancient forms are thrilling and fire the imagination all the more for being based in reality. Children love dinosaurs, and the best scientists (IMO) continue to be driven by a child-like wonder at the ancient world they explore, combining an active imagination with detailed and rigourous study of all evidence available.

Update. This post is part of the Tangled Bank blog carnival #79. Check out all the other posts available.


  1. Good post. You should submit it to one of these new fangled carnival thingies. Get a bit of traffic going.

  2. If you like this, check out also Another Booboo From Babu at "Thoughts in a Haystack". John Pieret holds up Babu G. Ranganathan of the Bob Jones University for some well deserved ridicule. Babu makes precisely the kind of shallow and ignorant comments about half wings that is so utterly demolished by the fossil evidence I have described here.

  3. Hey Sylas,

    You might like to check out a thread I just started over at IIDB on the evolution of flight.

    It was motivated by this recent paper which is very relevant to your blog post:



  4. Thanks, Steve. I had not seen that paper... but it is also relevant to another paper I have been looking at.

    Your paper is "A new Chinese specimen indicates that ‘protofeathers’ in the Early Cretaceous theropod dinosaur Sinosauropteryx are degraded collagen fibres", by T. Lingham-Soliar1, Alan Feduccia and Xiaolin Wang, in Proc. R. Soc. B; (2007) published on-line doi:10.1098/rspb.2007.0352.

    This paper was evicerated by the folks at the dino mailing list. See the message on May 23 and following. If their criticisms have any merit then the paper is badly flawed. Alan Feduccia is somewhat notorious in this whole area, as a longstanding skeptic about the association between birds and theropods, in spite of all evidence to the contrary. He is in danger of crossing the line from maverick to crank. A bit presumptuous of me, as paleontology is an area in which I am particularly weak. I'm relying very much on interpretations by others. But I have been contemplating a blog post speculating that Feduccia might be softening his stance somewhat. This latest paper suggests maybe not so much. But basically, don't go betting the farm on the paper's conclusions.

  5. Yes, there was also some scepticism on the IIDB thread with Dlx2 (who is some sort of palaeontologist, quite possibly dino) making some interesting arguments. However, the best point (IMO) was made by Martin Brazeau:

    "Yeah, I think therein lies the key point. Dinosaurs aren't the only things found in the Liaoning beds. There are heaps of other reptiles, amphibians, birds, fishes, etc. If their argument follows, then there should be plenty of apparently feathered non-dinosaurs. There are none."

    Simple and elegant and bloody annoying that I didn't think of it!

    As for Feduccia softening his approach, actually I was wondering a similar thing. In the press releases accompanying the paper, one quote was:

    "Lingham-Soliar's team do not take issue with the theory itself"

    Given that Feduccia was part of the team, perhaps this could be taken to imply a bit of softening up? Or more likely just crappy journalism.

  6. Larry Moran beat me to it. See Did Dinosaurs Have Feathers?. This is about the paper that SteveF mentions. Quoting Kevin Padian: "It is appalling that Proceedings B chose to publish this nonsense."

    That was pretty much the reaction I had on looking at the paper, though not quite so strong. They seemed to focus on just one fossil, and trying to resurrect an unusual interpretation, and ignoring a mountain of other evidence across a range of fossils where feather impressions are far less ambiguous.