Saturday 7 March 2015

How Ornithocheirus simus and other pterosaurs took to the air... from water?

Aquatically-adapted ornithocheiroid Ornithocheirus simus takes off using aquatic quad launch, as hypothesised by Habib and Cunningham (2010). Prints of this painting - which might be the first illustration of this launch strategy - are available from my shop.

Many pterosaur lineages seem to have close ties with marine environments, as evidenced by biases in their fossil and taphonomic records and indications of frequent interactions with marine fish. It stands to reason that these animals would find themselves in water on occasion, and trackways made by swimming pterosaurs indicate they may have been quite at home in this medium. Recent studies by Dave Hone and Donald Henderson have cast doubt on the swimming ability of pterosaurs because their floating postures seem rather awkward, which they suggest might have impeded breathing while swimming (Hone and Henderson 2014). Their studies found that, rather than sitting atop the water with arcing necks like birds, front-heavy pterosaur bodies collapse the head and necks into the water, bringing the mouth and nostrils close to the water surface. Although initially sceptical of this idea, I must admit to at least agreeing that avian-like postures may be difficult for pterosaurs. Our expectation that they floated in a bird-like fashion - which I've illustrated several times (Witton 2013 and elsewhere) - is actually quite silly given their proportions and differences with bird morphology. Pterosaurs lack the well-muscled hindlimbs which depress the back end of bird bodies into water, as well as the flexible necks and small heads required to attain duck or gull like floating postures. Indeed, birds seem unique in their floating posture, whereas pterosaurs seem to have floated in a manner more typical of other animals. Does the proximity of their nostrils or mouths to the water surface impede their swimming ability? Maybe not, given that virtually all non-avian animals I can think of  - including aquatic species like otters, crocodylians, swimming rodents, etc. - float and swim with their nostrils close to the water. As long as they have enough control over their swimming ability to clear their nostrils for respiration, they were probably fine. I see no clear reason to think pterosaurs were less competent in water than other animals, and maintain the view that some - for functional reasons - probably needed to swim to obtain the pelagic prey they did/likely did consume (Witton 2013).

But what did pterosaurs do when they needed to leave the water? Could they fly from the water surface or did they have to seek land to take off from? Anyone who knows anything about current pterosaur research knows that the leading hypothesis on pterosaur takeoff is quadrupedal launching, where the hindlimbs mostly serve to provide forward momentum, and vertical heft was provided by the forelimbs (Habib 2008) - many species of bats including vampires molossids and some mystacinids use a similar mechanism. It's worth stressing that this idea is not only supported by anatomical characteristics and positive results from biomechanical studies, but also by the fact that the hindlimbs were too weak to launch sensibly-massed pterosaurs into the air. All studies favouring bipedal launch have had to circumvent this somehow, typically by under-estimating pterosaur masses by, probably, as much as 60%. This makes bipedal launch a real no-go, whereas quad-launch has, to my knowledge, has meet all tests and predictions. We typically discuss quad-launch in terrestrial contexts, but can it work on water?

Schematic of water-hopping quad-launch strategy, from Witton 2013. The floating posture in panel 1 might be incorrect, but the general thrust of the image is OK.

It turns out, probably yes. According to work by Habib and Cunningham (2010), a version of quad-launch works just fine in aquatic settings. It seems that many folks have difficulty imagining how this works, but it's really not too dissimilar to terrestrial quad-launching. As with any takeoff mechanism, the the trick to water-launching is a leap providing sufficient height and speed to facilitate wing use (flapping alone doesn't really get you anywhere, and is especially ineffective at larger sizes). In this respect, it is just like terrestrial launch. An added complication of water launch is escaping surface tension, the cohesive force operating at air-water interfaces. This is where differences between terrestrial and aquatic launch become apparent, because it seems most pterosaurs were incapable of overcoming surface tension from a 'standing' (er... stationary floating) start. Their half-submerged posture and the fluid nature of the medium they are pushing against likely prohibited generation of sufficient energy for a standing launch. The solution to this is a series of hops across the water surface (above), each one providing further water clearance and velocity than the last. These were achieved, as on land, by the combined efforts of both the legs and arms. The wings are not fully deployed at this point, although the arm motion is probably showing some similarities to a flight stroke. The early stage of this takeoff might  look a little like swimming with a particularly powerful butterfly-stroke, albeit one where the swimmer is emphasising vertical motion rather than horizontal. Eventually, the pterosaur is leaping across - not through - the water and is clear enough to push off fully from the water surface. For this final push, the wing is opened and flapping can start. It seems that some pterosaurs - principally ornithocheiroids - were very well adapted for these manoeuvres, showing the shoulder reinforcement, upper-arm strength and distal-limb adaptations you'd predict for water-hopping quad launchers (Habib and Cunningham 2010, Witton 2013).

Some especially powerful pterosaurs, however, probably didn't need to worry about water hops. Giant azhdarchids, which experts predict were not only fuelled by muscle, but raw awesomeness sucked out of the universe itself, were probably powerful enough to water launch without hopping (Habib and Cunningham 2010). Some also ducks have sufficient power to do this - check out the launches in this video for cracking, slow-mo examples (the second is best, at 20 seconds in. Hat tip to Mike Habib for the link).

Recently, palaeoblogosphere regular Mike Traynor commissioned me to paint pterosaur water quad-launch - I think for the first time (if anyone knows different, please let me know). The results, a 6 m wingspan Ornithocheirus simus at the apex of the launch cycle, are above and, for fun, shown in progress from my Twitter feed below. If you'd like to own a copy of this image, you can: point your internet mobile at this page.

We're not done with aquatically-adapted animals just yet. Coming soon: the surprising aquatic adaptations of our own Mesozoic relatives.


  • Habib, M. B. (2008). Comparative evidence for quadrupedal launch in pterosaurs. Zitteliana, B28, 159-166.
  • Habib, M., & Cunningham, J., (2010). Capacity for water launch in Anhanguera and
  • Quetzalcoatlus. Acta Geosci. Sin. 31, 24–25.
  • Hone, D. W., & Henderson, D. M. (2014). The posture of floating pterosaurs: Ecological implications for inhabiting marine and freshwater habitats. Palaeogeography, Palaeoclimatology, Palaeoecology, 394, 89-98.
  • Witton, M. P. (2013). Pterosaurs: natural history, evolution, anatomy. Princeton University Press.


  1. what would the floating pose be like?
    I recall a paper that showed how top-heavy they are, but ... my brain.
    Would love to hear your opinion.

    1. The Hone and Henderson paper is available for free here:

      I expect the floating postures were pretty similar to what they show, although I must admit to wanting to see experiments with a few more poses and experiments with different densities.

  2. Beautiful painting! I have an odd question, though: was the moon substantially closer to the Earth in Mesozoic times? The crescent moon is gigantic.

    1. Er.... yeah. Got a bit carried away there! If I recall, the moon might have been a fraction closer to the Earth then, but only a smidgen. Nothing to make it a real 'Hollywood Moon' like this one.

    2. True, a slightly larger-than-normal sliver of crescent wouldn't look as awesome!

  3. An odd aside for you involving kites - specifically, the kites we use for kitesurfing, which consist of a single skin supported by an inflated (therefore floating) bladder at the leading edge. When we crash a kite into the water the standard relaunch method is to take the kite to the edge of the "window" (eg about 70 degrees either side of the line marking downwind), and the kite will then "sit up" and relaunch.

    We do this for two reasons - firstly, it's safer than a directly-downwind launch, which involves far higher forces, but mostly because this technique allows the kite to gradually break free of the surface tension, drain all that heavy water from the skin, and relaunch in a predictable manner.

    No idea if this is even potentially applicable, but it may be of interest. Footage of a kite relaunch:

    1. There may be some application here, or at least some insight offered by your observation: all marine-adapted pterosaurs are a bit skimpy in the headcrest department. That hadn't really occurred to me before: the biggest crests are typically found in species which likely inhabited inland settings, or were at least not adapted for aquatic life. I expect the conditions you report from kitesurfing may explain that: the effect of having a Tupandactylus-esque headcrest on the open sea might be quite traumatic!

  4. Would the body covering have served any buoyancy purpose and affected any aquatic posture?

    We all know how modern birds can float very effectively but even downy chicks are kept afloat and warm by their down.
    Probably less of a a factor as the animals get bigger and we cant know if their body covering was oily or not but I haven't seen the matter discussed at any length.

    1. We don't really know how well pycnofibres would trap air. Bird feathers do this very well, but mammal hair does seem to function in the same way (at least, as far as I know). My guess is that pterosaur fuzz is functionally closer to mammal hair than feathers, but who knows?

      On floating: there's no question about pterosaurs floating - they could do that without problem. It's the posture which has proven a point of discussion, although there's precious little experimentation and data on it. The Hone and Henderson paper, and work on swimming traces, is the current limit of investigation on pterosaur floating/swimming.

  5. Some overlooked issues here:
    1. The duck’s wings are dry and above the water until the moment they slap down on the surface where action/reaction pushes the bird skyward. In addition, the webbed feet are raised then extended creating a leap to elevate the duck and sufficient time for the first upstroke and downstroke of the wings, which is very rapid in short-winged ducks. The head and body do not rise until the third downstroke.

    2. At all times during these wing-beat cycles the duck wings are in their typical lateral flying configuration. So the large flight muscles are used to maximum effect.

    3. Now let’s move over to the pterosaur. Opposed to the duck, the pterosaur wings are submerged when floating on water. So the membrane is loaded with water like a pelican’s gular sac. As you have shown, due to the nature of the pterosaur wing-folding pulley, digit 4 is oriented posteriorly when folded.

    4. In ornithocheirds the folded wing is very long, extending beyond the rump a distance about equal to the snout/vent length, which is quite a big longer than the same dimensions in a duck. That issue is correctly shown in your launch series illustration, but foreshortened in your painting and oddly the wing fingers have assumed a more lateral orientation relative to the humerus and antebrachium than is possible (the plane of the wing IS the plane of the wing and cannot be bent or violated).

    So that’s our pre-launch starting point.

    5. As I mentioned earlier (pterosaurheresies/pterosaur-take-off-from-water) your lateral view series of figures does a fine job of showing how the pterosaur could have splashed in the water, but never gets to the point of freeing the wings from the water and elevating them for the first downstroke. We can’t rely on the hind limbs to provide decent propulsion or any sort of lift. The feet in ornithoheirids are quite reduced. You have drawn them twice to three times their actual size. You have omitted the wing membranes, which would have been loaded with water and the wing finger is oriented posteriorly, providing no lateral wing slap, as in ducks, or wing elevation, as in pelicans.

    6. Pelicans, like pterosaurs, have proportionately longer wings than ducks and a longer wing beat cycle. As illustrated in the pterosaurheresies blog post, pelicans raise their wings first, then start a series of half-flaps (not touching the water) which ultimately frees them from the water with the help of their paddling webbed feet. So in both ducks and pelicans the first step is to produce a dry downstroke and a flurry of paddling.

    7. As illustrated in the pterosaurheresies blogpost, the first effort of a floating ornithocheird is to elevate the folded wings dorsolaterally. Then they are free to extend dorsolaterally to the standard flight position, take time to shed their water load and create that first downstroke, then a second, and as many other wing beats as are necessary to break free of the water. I’m sure the hind limbs would be churning the water as well, but probably to little effect due to the tiny feet. Of course elevating the wings depresses the rest of the pterosaur into the water and introduces balance issues, but these are short-lived and minor if pelicans are analogous.

    8. The nostril and floating issues you raised were dealt with were more accurately illustrated (with bones, not fanciful silhouettes) in another pterosaurheresies blogpost: pterosaurs-were-unlikely-floaters-hone-and-henderson-2013.

    1. If I've learned one thing over the years, David, it's that discussing anything with you concerning pterosaur palaeobiology or scientific methodology is a complete waste of time, and I'm not about to attempt to engage with you again. As with all my pterosaur work, my painting and article are in full accord with published literature and contemporary data presented by reputed researchers on pterosaurs, biomechanics and animal flight. Of course you have different opinions on these matters - I would expect nothing less! But if even world-leading experts in specialist fields cannot sway your opinion once you have decided your preferred interpretation - and I've seen them try - what is the point of discussing anything with you?

  6. I thought whats this chap on about? Then realised what the intials DP stood for! He is similarly dismissive of how digital photography and files work

  7. As always I really appreciate your art. Looking at your Ornithocheirus pictures, just for a moment I felt like I could imagine how awesome it would be to watch one of these creatures actually taking off from the water on a bright summer day. Then I immediately was back to not being able to imagine it again, but thanks for the experience anyway!


  8. The post has really talks my interest. I am going to bookmark your website and keep checking for new information.

    Microsoft Office 2016