Wednesday, January 2, 2013

Guest Post. Dragon Tails: What Pterosaurs Teach Us about Velociraptor

As part of the flurry of new papers in the Flugsaurier 2012 paper pseudo-volume, Scott Persons has a paper out looking at the remarkable convergences between various dromaeosaurs and a number of rhamphorhynchoid pterosaurs. Here in a guest piece he takes us over this project. - Dave.

An Unexpected Tail
Last week, like a growing 149 million dollars’ worth of other holiday movie goers, I took three hours out of my yuletide respite to make an epic trek across sidewalks long and public transportation foul to my local cinema and saw “The Hobbit: An Unexpected Journey”. I was particularly keen to see how the special effects wizards of Weta Workshop would depict the story’s big bad: Smaug, the dragon. Naturally, since The Hobbit has been broken up into a trilogy, Peter Jackson decided not to fully unveil Smaug, and I’ll have to wait until the sequel. But there was a surprise appearance by a Megaloceros (the “broad antlered” or “Irish Elk”), and Jackson did give the audience a dragon teaser. We got to see a smoke-obscured silhouette here, a clawed foot there, and . . . the tail. 
I’m a PhD student of paleontology at the University of Alberta and, with the help of my supervisor Dr. Phil Currie, I’ve dissected, excavated, measured, photographed, and digitally sculpted the tails of many creatures, ancient and modern. To me, Smaug has an interesting tail. Unlike a crocodile or a dragon from the isle of Komodo, he did not drag his tail behind him. Rather, the Tolkienian drake carried it raised above the ground -- a caudal posture that I suspect shows the influence of dinosaur paleontology on the animators. The tail was very flexible and muscular enough to casually toss aside a group of armored castle-defenders and to do collateral damage to the medieval architecture.
Swooping out of the sky, Smaug’s fiery breath and wrecking-ball tail deal a one-two punch to a stone tower.
That brings me to a fun, blatantly whimsical, and entirely inconsequential thought problem: if the monsters of Middle-earth were not spontaneously generated from Uruk-hai pits and human imaginations, but instead were products of biological evolution, what sort of tail would a dragon most likely have? It’s a silly question, but at least twice in the history of our own planet, reptilian beasts have taken on dragon-like form, and both times their tails have followed remarkably convergent evolutionary paths.

Tail of a Revolution
It can be argued that, of all the dinosaurs ever dug, none have been more scientifically important than Deinonychus. Deinonychus is a kind of dromaeosaurid (meaning that it belongs to a group of carnivorous dinosaurs commonly known, because the group includes the iconic Velociraptor, as “raptors”).  The 1969 study of Deinonychus by Professor John Ostrom documented a wealth of anatomical features previously unseen, or at least unrecognized, in any other dinosaur. Many of these features seemed to imply a high metabolic rate, an active lifestyle, and an ancestral relationship with modern birds. Deinonychus was the catalyst that began the “dinosaur renaissance” or “dinosaur revolution” (a paleontological paradigm shift that has affected how we think about, and approach learning about, a great deal more of prehistory than just dinosaurs -- including pterosaurs). Among the unusual anatomy of Deinonychus was its tail.
The skeleton of a tail is an extension of the spinal column and, in a dinosaur, the tail skeleton is composed of three major kinds of bones. First and foremost, are the interlocking vertebrae, which protect the spinal nerves and sport upwards-projecting neural spines, to which epaxial muscles attach. Then, there are the caudal ribs (or simply “transverse processes” – there is controversy over the proper terminology), which fuse to the vertebrae and project outwards (perpendicular to the long axis of the tail) and also provided anchorage for tail muscles. Last, but certainly not least, are the chevrons. Despite the fact, or perhaps a little bit because of the fact, that these tail bones have received little descriptive or collecting priority from other researchers, I positively love chevrons. They are elegantly shaped, usually resembling capital Y’s, are important for tail muscle function, and (contrary to historic assumption) are useful in identifying taxonomic and evolutionary relationships.  In the tails of both dinosaurs and pterosaurs, chevrons are positioned in between sequential pairs of vertebrae and project downwards.
The basic parts of the tail skeleton shown on a duckbilled dinosaur.
At the start, the tail skeleton of Deinonychus appears normal. Just past the hips, the neural spines, caudal ribs, and chevrons all have a typical shape and they all project to a respectable extent. But, as the tail progresses towards the tip (and it doesn’t take long) things start to get weird. The neural spines, caudal ribs, and chevrons all shrink in, with the former two disappearing entirely . . . and then come the caudal rods. Both the vertebrae and chevrons abruptly develop pairs of elongated rods of bone that project towards the hips. These rods are slender, but very long (the longest easily overlap seven other sequential vertebrae), and they split, each becoming two still thinner rods. Together, rods of the vertebrae form a quiver that virtually encapsulates the dorsal (upper) portion of the tail, and together the rods of the chevrons do the same to the ventral (lower) portion.
Raptor tails are strange! Elongated vertebral rods form an upper quiver and elongated chevron rods form a lower quiver. Skeletal image of Velociraptor courtesy of Scott Hartman (
  Professor Ostrom was rightfully impressed by the caudal rods of Deinonychus, and he realized that such unusual structures must have evolved to serve some sort of unusual function. Ostrom’s best specimens of Deinonychus were preserved lying on their sides and their tails were straight as boards. Ostrom was also much taken with what he thought was the overall highly-athletic nature of the hindlimbs of Deinonychus, so he speculated that the caudal rods were adapted to aid in high-speed pursuit. His idea was that the rods must have stiffened the tail and allowed it to function like the balancing pole of a tightrope walker. This balancing tail, he reasoned, would have come in handy when turning while running or when leaping onto the back of some poor dinosaurian herbivore.

Enter the Sky Dragons
The tail of Deinonychus and its raptor relatives is bizarre, but it is not (as Professor Ostrom himself realized) unique. Among all known vertebrates, a similar tail anatomy has evolved in one other group. . . and now we come to why I have been allowed to spend so much time discussing dinosaurs on what is supposed to be a blog about pterosaurs.
While later and more advanced pterosaurs (like Pterodactylus) only had short, stubby tails, early pterosaurs had long ones. The caudal skeletons of these long-tailed pterosaurs (with the exceptions of the dimorphodontids and very primitive forms) are strikingly similar to that of Deinonychus. In the case of long-tailed pterosaurs, the function of the caudal rods has always seemed obvious. As flying animals, increased rigidity would have helped a tail to serve as a stabilizer or as a rudder.

Look familiar?  The tails of pterosaurs and dromaeosaurids are so similar that, in the fossil-forging black-markets of China, the tail of one is often used to “complete” a partial skeleton of the other. Skeletal image of Rhamphorhynchus courtesy of Scott Hartman (
In Professor Ostrum’s description of Deinonychus, he expressed his interest in considering this striking example of convergent evolution in a later study. Regrettably, however, he never got around to it -- after all, he soon had a revolution on his hands.

Fleshing Out the Evidence
            A lot has changed since the dinosaur renaissance. For instance, we now know that caudal rods are characteristics of all dromaeosaurids (except the South American unenlagiine dromaeosaurids, which are odd-ducks in many regards). We know a lot more about the evolutionary history of caudal rods, in both dromaeosaurids and pterosaurs. We also know that the rods were not as stiff as Ostrom had thought. Consider the below images of a tail of a Bambiraptor and of a Velociraptor. Both are dromaeosaurids with caudal-rod bearing tails and both are fully articulated. Yet, both are preserved in a sinuous curve (or in the case of the Velociraptor, many sinuous curves).
The strongly curved tail of Bambiraptor.

Despite its caudal rods, this Velociraptor tail is preserved in a graceful S-shaped curve.
So, it turns out that caudal rods are flexible (not surprising when you think about how thin each rod was). But, if caudal rods permitted their tails to curve to such a degree, what good were they? Well, I suspect that the rods were very helpful in keeping the tails rigid. No, I am not talking in circles, and I don’t think that I am talking nonsense. The rigidity provided by the rods of dromaeosaurids was one-dimensional. I have been able to see a lot of dromaeosaurid tail fossils, and many specimens, like the two above, are curved laterally, but I have never seen one that shows articulated caudal rods bending strongly dorsoventrally (that is, up or down).
            That caudal rods provide mostly dorsoventral rigidity makes sense, if you consider the way the rods are arranged in their quivers. The rods are not haphazardly piled on top of one another; rather, they are tightly pressed against the vertebrae and chevrons and are neatly stacked vertically. In other words, the quivers were arranged to be thicker dorsoventrally than they were laterally.
Cross-section through the tail of Deinonychus showing the arrangement of the caudal rods. The arrangement of the rods made the tail harder to bend up-and-down than side-to-side.
A cylindrical tight-rope walker’s pole is the wrong analogue. Instead think of a meter stick, which may be bent with moderate force, but only perpendicular to its broadest plane.
It is now also possible to think a step further and consider the muscles of the tail. Let’s first try to do that in very general qualitative terms. Remember the quickly reduced neural spines, caudal ribs, and chevrons? Those all indicate that the caudal muscles of both dromaeosaurids and pterosaurs were substantially reduced.
To help consider the problem quantitatively, a technique I used was to create digital models of the tail skeleton of a Velociraptor and a Rhamphorhynchus (a pterosaur) and to sculpt the corresponding muscles over the skeletal models. The results of this modeling concur with the qualitative inference. In particular, raptors and pterosaurs were found to have very weak caudofemoral muscles (indeed, some pterosaurs may not have had caudofemoral muscles at all).

Digital reconstructions of the tail of Velociraptor, showing the tail and hip skeleton (A), the caudofemoral muscles (B), and the full muscle reconstruction (C).
Digital reconstructions of the tail of Rhamphorhynchus, showing the tail and hip skeleton (A), the caudofemoral muscles (B), and the full muscle reconstruction (C)
 These caudofemoral muscles merit special explanation. They are muscles found in the tails of reptiles and dinosaurs that are actually part of the hind limb system. They are the primary limb retractors and provide a major power boost when walking and running.
Again, these adaptations seem easy to explain in the tail of pterosaurs. If a tail is to serve as an inflight rudder, some lateral flexibility would be needed, but strong dorsoventral stiffness would have prevented the constant pull of gravity from deflecting the tail downwards and disrupting the body’s aerodynamic form, and all the better if this stiffness was obtained by the passive rigidity of bone, rather than the hard work of muscle. Reducing weight is always a benefit for flight, and, perhaps, particularly reducing posterior weight. What better posterior weight could a flying animal lose than that of a tail muscle whose primary function is in land-bound locomotion?
Thanks in large part to the work of pterosaur researcher Dr. Dalla Vecchia, we know that among the most primitive of pterosaurs (animals like Austriadactylus and Eudimorphodon) caudal rods had not yet evolved (though there are some anatomical signs that they were in the works). However, all these primitive forms already had clear possession of the power of flight. We can be certain (or about as certain as the fossil record ever permits) that, when the caudal rods of pterosaurs evolved, it was in the context of an aerial lifestyle.
To me the remarkable similarity in form between the tail skeletons and muscles of pterosaurs and dromaeosaurids indicates an equally strong similarity in function.  Is it possible then, that the tails of dromaeosaurids also evolved on the wing?

Feathered Dragons from the Orient
The little dromaeosaurid Microraptor has sparked something of its own dinosaur revolution. Microraptor is among the oldest and most primitive of the known dromaeosaurids. Buried in fine volcanic ash, specimens from the famous fossil beds of China have revealed that Microraptor had feathered wings on the fore and (oddly enough) on the hind limbs. It also had caudal rods. Whether or not Microraptor could truly fly or simply glide is a matter of current scientific debate, but the winged-raptor was clearly not a land-bound creature.

Microraptor has wings and caudal rods.
An older Chinese dinosaur, Anchiornis is a primitive member of the Deinonychosauria. Named in honor of Deinonychus, the group Deinonychosauria includes the dromaeosaurid and their close relatives the troodontids. Like Microraptor, Anchiornis had wings, but it lacked caudal rods. Thus, as with pterosaurs, it appears possible to bracket the evolution of the bizarre dromaeosaurid tail between two aerial genera.
Anchiornis has wings but not caudal rods.
Does all this mean that the tails of Deinonychus and Velociraptor indicate that these dinosaurs could fly? Certainty not. However, I do think it means that we should think of Deinonychus, Velociraptor, and other dromaeosaurids like Cretaceous ostriches. They are animals without the ability to fly or glide but who have inherited a few telltale anatomical feature that attest to their ancestor’s aerial life. (I am certainty not the first person to suggest this. Based on various other lines of anatomical evidence, many paleontologist, most prominently Greg Paul, have argued that dromaeosaurids were secondarily flightless.)
I also think it means that, when you imagine a flying dragon, be it a dinosaur, pterosaur, or thought problem fantasy, you should not envision it with a tail that limply streams behind it. The tail should be, most probably, held up; be capable of lateral, but not vertical, swishes; be muscularly reduced and light weight; be quite elegant and graceful; and, thus, not (I am afraid, Mr. Jackson) be of much use as a siege weapon.


  1. Scott, great post!

    But Smaug's tail as a weapon of destruction is not a device of Jackson's, but is canonical JRRT. In his great boast, Smaug describes his armaments: "My armour is like tenfold shields, my teeth are swords, my claws spears, the shock of my tail is a thunderbolt, my wings a hurricane, and my breath death!"

  2. Perhaps Smaug evolved from ancestors that had rather flimsy, dromaeosaur-like tails, but later developed a large, club-like tail independently as the tail no longer played a large role in flight (think of it kind of like the transition between "rhamphorhynchoid" and pterodactyloid pterosaurs, except in this case the tail is co-opted instead of reduced). The stiffening rods in the tail could function in the support of the tail club, as they do for some ankylosaurs.

  3. The Rhamphorhynchus skeletal is credited to the wrong artist. Its not Scott Hartman's but Mike Hanson's. He has a deviantart page, though he's removed the original skeletal, so you'll just have to take my word for it.

    1. Chris is correct, that is my skeletal reconstruction. I removed it from DeviantArt due to some concerns about its accuracy following some recent discoveries. Nevertheless, I would appreciate it if credit were given to the proper person.

  4. Are caudal transverse processes actually thought to be homologous to ribs, and thus be caudal ribs? It was my impression they were homologous to diapophyses, and the ribs were merely lost. They are never(?) found disarticulated from the neural arch in dinosaurs AFAIK, and they seem to grade into sacral diapophyses, not sacral ribs.

    Also, do you have the specimen number for that Anchiornis? I have a larger photo that shows the plaque says IVPP V_____, but I can't quite make it out. Maybe V13065?

  5. Is Persons' paper available somewhere?

    1. It's hear Andrea:

      Click the left hand set of Chinese characters of the three at the bottom of the page and then add a .pdf suffix to it and it should work.

  6. This blog is definitely entertaining and also factual. I have picked a bunch of helpful advices out of this source. I ad love to come back again and again. Thanks! IDN POKER APK

  7. Thanks again for the blog.Thanks Again. Much obliged. LINK SBOBET88

  8. Thanks for sharing, this is a fantastic article.Really thank you! Much obliged.

    idn poker

  9. Great article! Thanks for sharing an informative blog with us. I learn many new things from your article. I really like your post, It is very helpful for us. Please keep sharing these types of articles.

    Fantasy Sports App Development

  10. I have been browsing on-line more than three hours these days, yet I by no means discovered any attention-grabbing article like yours. It is beautiful worth enough for me. In my opinion, if all website owners สล็อตjoker and bloggers made just right content as you probably did, the internet will likely be a lot more helpful than ever before.

  11. Hi, i would like to show you a vertebra fossil, for which i hesitate between dromaeosaur and pterosaur. Is there an email where i could reach you ?