A remarkable “flying” snake

An article in this week’s New York Times Science Observer column highlights the paradise tree snake (Chrysopelea paradisi) of Asia, long known to escape predators by hurling itself from a tree and sailing through the canopy to alight on a new tree.  Studies suggest it can travel as far as 300 feet in this way.  A new study published in the oddly-titled and hard-to-find journal Bioinspiration and Biomimetics analyzes the mechanics of this “flight”:

. . . a study in which scientists threw the snakes from a 50-foot tower and recorded their descent on video suggests that the snakes are active fliers, manipulating their bodies to aerodynamic effect.

“It essentially looks like they are slithering in the air, like a whip moving left and right,” said Jake Socha, the study’s lead author and a biomechanist at Virginia Tech. “The body itself moves up and down as well.”

Dr. Socha and his colleagues found that the paradise tree snake tilts its body about 25 to 30 degrees relative to the airflow to stay as aerodynamic as possible. The farthest a snake was able to travel from the tower was about 79 feet.

This is far better seen than described; here’s a nice video from PBS:

And Wikipedia says a bit more:

Upon reaching the end of a tree’s branch, the snake continues moving until its tail dangles from the branch’s end. It then makes a J-shape bend,[7] leans forward to select the level of inclination it wishes to travel to control its flight path, as well as selecting a desired landing area. Once it decides on a destination, it propels itself by thrusting its body up and away from the tree, sucking in its stomach, flaring out its ribs to turn its body in a “pseudo concave wing”[8] all the while making a continual serpentine motion of lateral undulation[9] parallel to the ground[10] to stabilise its direction in midair in order to land safely.[11]

The combination of sucking in its stomach and making a motion of lateral undulation in the air makes it possible for the snake to glide in the air, where it also manages to save energy compared to travel on the ground and dodge terrestrial bounded predators.[7] The concave wing that a snake creates in sucking its stomach, flattens its body to up to twice its width from back of the head to the anal vent, which is close to the end of the snake’s tail, causes the cross section of the snake’s body to resemble the cross section of a frisbee or flying disc.[10] When a flying disc spins in the air, the designed cross sectional concavity causes increased air pressure under the centre of the disc, causing lift for the disc to fly.[12] A snake continuously moves in lateral undulation to create the same effect of increased air pressure underneath its arched body to glide.[10] Flying snakes are able to glide better than flying squirrels and other gliding animals, despite the lack of limbs, wings, or any other wing-like projections, gliding through the forest and jungle it inhabits with the distance being as great as 100 m.[10][13] Their destination is mostly predicted by ballistics; however, they can exercise some in-flight attitude control by “slithering” in the air.[1]

34 Comments

  1. Greg Esres
    Posted December 5, 2010 at 1:16 pm | Permalink

    “25 to 30 degrees”

    In aerodynamics, we refer to that as “angle of attack”, and our goal in long glides is maintaining “L/Dmax”, which is the highest ratio of lift to drag. The angle of attack combined with an airflow is what generates lift.

    For real airplanes, a 25 degree angle of attack would be larger than our wings can actually achieve without “stalling”. I suspect that the snakes are able to achieve higher because the wings are what we call “low aspect ratio” meaning that the are vary narrow compared to their length; this sort of wing can achieve higher angle of attacks without stalling, but tend to be very draggy.

  2. Greg Esres
    Posted December 5, 2010 at 1:24 pm | Permalink

    BTW, I’m skeptical of this assertion that this mode of travel “saves energy”. The snake is merely cashing in on the potential energy it stored by climbing the tree in the first place. I would bet heavily that climbing the tree and gliding a distance X uses more energy than just slithering along the ground distance X.

    • Posted December 5, 2010 at 3:23 pm | Permalink

      And I’m sure it’s more fun. Same-old same-old slithering vs. WHEEEEEEEE!!

    • Patrick Q
      Posted December 5, 2010 at 3:58 pm | Permalink

      If the snake travels all the way down the first tree and all the way up the second tree, it first gives up all its hard won gravitational potential energy and then has to put it all back.

      If it leaps instead and lands halfway up the second tree, it only loses half the potential energy and only has to reinvest that half to achieve the same height in the new tree. Thus it expends less energy as it moves from one hunting zone to another.

      • Greg Esres
        Posted December 5, 2010 at 6:58 pm | Permalink

        Well, yes, if the goal is to get to the top of the next tree, then I suppose it does work out.

        • Notagod
          Posted December 5, 2010 at 10:27 pm | Permalink

          That would be a reasonable hunting strategy for a tree dweller.

          And agreeing with Aratina I hope the snake is realizing the WHEEE potential.

      • Torbjörn Larsson, OM
        Posted December 6, 2010 at 3:25 pm | Permalink

        Yeah.

  3. Dunbar
    Posted December 5, 2010 at 2:39 pm | Permalink

    I wonder, can the snake glide after ingesting prey?

  4. Torbjörn Larsson, OM
    Posted December 5, 2010 at 2:40 pm | Permalink

    I don’t see how that energy saving works either.

    The real cool thing to me is that they have seen momentarily positive upward force with reference to force summation: that snake can indeed in principle fly.

  5. Richard Thomas
    Posted December 5, 2010 at 2:42 pm | Permalink

    The vent of Chrysopelia is not close to the end of the tail! The tail, I wager, is at least a third of the body length.

  6. Richard Thomas
    Posted December 5, 2010 at 2:44 pm | Permalink

    I should have said that the vent defines the beginning of the tail, so probably what was meant is that the vent is at the end of the body.

    • Diane G.
      Posted December 6, 2010 at 4:54 pm | Permalink

      And perhaps a tail that doesn’t flatten serves better as a rudder in flight.

  7. Sili
    Posted December 5, 2010 at 3:37 pm | Permalink

    But what use is half a wing?

    • Torbjörn Larsson, OM
      Posted December 6, 2010 at 3:27 pm | Permalink

      Half the glide lift as a “whole wing”, whatever that is?

      More pertinently, what use has a creationist for his half brain?

  8. J.J.E.
    Posted December 5, 2010 at 4:48 pm | Permalink

    So that paper was written by Jake Socha, it seems. He got his PhD in OBA at University of Chicago. Congrats to him for getting such a nice PBS spread associated with his work! I wonder if that footage is his or not?

  9. Insightful Ape
    Posted December 5, 2010 at 5:34 pm | Permalink

    Clearly the snake is intelligently designed to enable it to perform this feat.
    Equally obvious is the fact that god wants the lizard to go hungry.

  10. Posted December 5, 2010 at 6:29 pm | Permalink

    Wow – they’re all Sully Sullenbergers.

  11. chriss
    Posted December 5, 2010 at 7:50 pm | Permalink

    That has to be one of the most remarkable and beautiful things I have ever seen. I couldn’t help but wonder what life would be like for snakes if we exposed our children to those images before we dragged the serpent of Eden into their dreams.

  12. strangebeasty
    Posted December 6, 2010 at 3:01 am | Permalink

    That is amazing! I’d love to find out about how those might have evolved!

  13. Noam Chomsky
    Posted December 6, 2010 at 6:45 am | Permalink

    Let me quote Sean B Carroll from his book the Making of the Fittest: “Because decaying genes generally accumulate multiple defects, their inactivation cannot be easily reversed. This means that the loss of gene functions is generally a one-way street. Once gone, these functions will not return.” If Natural Selection through random mutation cannot repair a gene what makes you think it can build a gene?

    • J.J.E
      Posted December 6, 2010 at 8:42 am | Permalink

      Try gene duplication, exon shuffling, or even de novo creation of genes from non coding sequence. These are all well attested mechanisms of creating the raw material for evolutionary novelty. They are active areas of research. In face, duplication has been a hot field for quite some time. More recently, copy number polymorphism has been “hot”. Duplication is so common that thee are thousands of new segments of DNA segregating in every organism that had so far been examined thoroughly. Next question.

      • Noam Chomsky
        Posted December 6, 2010 at 9:00 am | Permalink

        Duplicating genes, shuffling genes around, deleting genes, mutating an already-built gene is not the same as building a brand new gene. Read what Sean said: NS cannot repair genes. If NS cannot repair genes then it is reasonable to conclude that NS cannot build genes.

        • Tim Martin
          Posted December 6, 2010 at 10:05 am | Permalink

          First of all, what does this have to do with flying snakes?

          Second, what does it matter if a gene is partially made from some old DNA, or is built (over a very long time and after a large number of insertions, I suppose) from scratch? Either way, natural selection will operate on the gene. For a list of “new” genes that are known to have been created through different types of mutations, please consult this article. If you can’t access it I’ll be happy to email you the .pdf.

          Third, “If NS cannot repair genes then it is reasonable to conclude that NS cannot build genes” is wrong. You misunderstand statistics. Once a gene has been mutated into deactivation, it is incredibly unlikely for the exact mutation that would be necessary to undo this to occur. Gene repair isn’t impossible – it’s just incredibly improbable. But this is only true because you have a specific goal (a specific mutation that has to occur). Evolution, however, has no goal. While the odds of getting some specific mutation are incredibly low, the odds of getting a mutation are basically 100% (for example, there are on average 128 mutations per human zygote). Once mutations have occurred, natural selection, genetic drift, etc. will operate on them.

          • Noam Chomsky
            Posted December 6, 2010 at 10:35 am | Permalink

            “what does it matter if a gene is partially made from some old DNA, or is built (over a very long time and after a large number of insertions, I suppose) from scratch? Either way, natural selection will operate on the gene.”

            Yes, Natural Selection will operate on an already built gene, but only intelligence can build a new gene.

            “For a list of “new” genes that are known to have been created through different types of mutations, please consult this article.”

            Mutating an already built gene and getting a new function, yes, Natural Selection can do that, although it is very rare. Building one of the original 500 immortal genes, NS can not do that.

            [“If NS cannot repair genes then it is reasonable to conclude that NS cannot build genes” is wrong. Once a gene has been mutated into deactivation, it is incredibly unlikely for the exact mutation that would be necessary to undo this to occur. Gene repair isn’t impossible – it’s just incredibly improbable.]

            Correct, and if it is extremely improbable to repair a broken gene, think of how improbable it is to build a completely new gene.

            [But this is only true because you have a specific goal (a specific mutation that has to occur). Evolution, however, has no goal. While the odds of getting some specific mutation are incredibly low, the odds of getting a mutation are basically 100% (for example, there are on average 128 mutations per human zygote). Once mutations have occurred, natural selection, genetic drift, etc. will operate on them.]

            You’re assuming that just any mutation will do some good. This is wrong. Proteins are very precise and not just any sequence of amino acids will get you a functional protein. Just analyze how similar the so-called 500 immortal genes are and you will find that on average 60% of the amino acids sites are identical. This tells us that 60% of those sites cannot be tinkered with.

            • J.J.E
              Posted December 6, 2010 at 11:15 am | Permalink

              You are wrong. But enough wasting time with you. You can read about de novo genes here: http://blogs.nature.com/nature/journalclub/2007/10/manyuan_long.html

            • Tim Martin
              Posted December 6, 2010 at 12:50 pm | Permalink

              only intelligence can build a new gene.

              Unsupported assertion.

              Building one of the original 500 immortal genes, NS can not do that.

              Unsupported assertion.

              Correct, and if it is extremely improbable to repair a broken gene, think of how improbable it is to build a completely new gene.

              I’m aware of how improbable it is to build a new gene. Are you aware of how improbable it is that you were born? What were the chances that your parents would be born and their lives would play themselves out in the exact way that would allow them to meet and conceive you? And what about the chances of your grandparents being born, meeting, and conceiving your parents? Of all the things that could happen in a person’s life, the chances of creating a specific child with a specific person are vanishingly small, and yet this had to happen an incredible number of times back through the generations of your lineage in order for you to be born.

              And yet you are not surprised that you exist. Your mistake regarding NS, as I already explained, is that you’re looking at the probability of getting a specific outcome, as opposed to getting any outcome. You can learn more about this type of perspective error here, though I doubt you will.

              just any mutation will do some good. This is wrong.

              You’re absolutely correct. Do you think this is some big revelation that scientists have managed to miss for 150 years? Do you know anything about natural selection? Of course mutations can be deleterious as well as advantageous. Most often, they have no significant effect whatsoever. Organisms that are inviable or disadvantaged due to their mutations will die. Organisms with an advantage will increase the proportion of their genes in the gene pool. That’s the entire idea of selection.

              That’s all the time I’m willing to spend on you, sir. Good day.

            • Torbjörn Larsson, OM
              Posted December 6, 2010 at 3:35 pm | Permalink

              What is an “original” and an “immortal” gene? No one has found such in biology.

              Now I’m just a layman, but even I know that you can’t go around making assertions out of thin air. Mutations means that no gene is immortal more than species are, and selection on top of that means no gene is “original”.

              What people have seen is AFAIU that the LUCA had more gene families (~ 1000) than the smallest genomes today. Out go ~ 500 of the so called “immortal” genes.

              Also, I’ve seen papers where horizontal gene transfer is at least 1/gene today for _all_ genes. *Swap*, out goes “original” genes, no one is seen to be safeguarded by mystical unobserved religious forces.

            • Noam Chomsky
              Posted December 6, 2010 at 5:15 pm | Permalink

              @Tim Martin

              “Of course mutations can be deleterious as well as advantageous. Most often, they have no significant effect whatsoever.”

              Do the math. If you look at the immortal genes they are on average 60% identical. That’s 60 amino acids that are exactly the same in two genes. This tells us that those amino acid positions are necessary and that any divergence from them will result in death. 20 raised to the 60th power is 10^78. That’s the odds of making one of the 500 immortal genes which are most likely necessary for the simplest life forms.

        • J.J.E
          Posted December 7, 2010 at 2:02 am | Permalink

          And what about de novo genes? Or do you always ignore the inconvenient?

          http://blogs.nature.com/nature/journalclub/2007/10/manyuan_long.html

  14. Quidam
    Posted December 6, 2010 at 10:27 am | Permalink

    “When a flying disc spins in the air, the designed cross sectional concavity causes increased air pressure under the centre of the disc”

    No it doesn’t. Spinning a Frisbee does not create lift. This can be confirmed by putting a dimple in the center and spinning it balanced on a stick. The frisbee stays on the stick.

    It is the velocity through the air coupled with the angle of attack that provides lift, the spin provides stability to maintain the angle of attack through the gyroscopic effect.

    The concave aerofoil shape of the snake allows air to travel past the body with a minimum of drag and eddying (and no, aerofoils don’t produce lift because the air on top has to move faster) I suspect the ‘swimming’ motion is to maintain balance and orientation.

    • Torbjörn Larsson, OM
      Posted December 6, 2010 at 3:47 pm | Permalink

      We may need to hunt down the paper, but I don’t see what would be wrong with the sentence. It is indeed an increased air pressure, a flying disc will have an angle of attack, et cetera.

      At worst it is misleading for the reasons you mention, because someone (like the author probably did) is thinking of other aerodynamic forces than the angle of attack gives.

      Which btw I don’t think can be discounted totally, even if the reason for keeping the air stream on top of the wing is mostly to prevent excessive drag IIRC.

      • Quidam
        Posted December 7, 2010 at 10:56 am | Permalink

        By suggesting that the spin provides lift, the wikipedia article is mis-stating the original author. The wiki link fails but it linked to Sarah Hummel’s M.Sc. dissertation. PDF file

        Quote:
        “How does spin affect the flight? Spin has been shown to have little affect on the aerodynamics of a Frisbee but it does have a critical role in the flight dynamics.
        Although the lift and drag coefficients are unaffected by spin and the pitch and roll moments are affected but only by a very small amount (Potts and Crowther, 2002), the critical aspect of the spin is the stability it provides in flight due to angular momentum and gyroscopic precession.”

  15. Marvol19
    Posted December 6, 2010 at 1:28 pm | Permalink

    What to me is the coolest is to imagine that early feathered dinosaurs must have started this way… and look what they became.
    Give natural selection on these snakes another what, 10 My? Then imagine the sky filled with flying snakes… awesome :O!


2 Trackbacks/Pingbacks

  1. […] A remarkable “flying” snake An article in this week’s New York Times Science Observer column highlights the paradise tree snake (Chrysopelea […] […]

  2. […] Via Why evolution is true […]

%d bloggers like this: