Gallery and print store

Wednesday, 24 April 2013

They're reptiles Jim, but not as we know them

A couple of Therizinosaurus cheloniformis in a Cretaceous woodland,  one of whom is making the loudest 'coo coo' noise you've ever heard, and one of whom couldn't care less.
There is no 'saur like a therizinosaur. That should be a well-known saying among palaeontologists, but it's not. They're all too busy doing science to come up with silly puns, no doubt. This painting of these magnificent animals came as a response to my previous post concerning 'Feather Resistance' (the fact that some people just don't like the idea of feathered dinosaurs) which featured a photo of the feathered remains of the small Chinese therizinosaur Beipiaosaurus. The neck vertebrae of this specimen thread through an extensive, posteriorly-expanding wedge of neck feathers in a fashion very reminiscent of diagrams of pigeon anatomy that we've all seen in textbooks, and it's fairly easy to see where my mind went from there (makes a change from all the cassowary-inspired paleoart on the internet, a trope which I'm guilty of myself). The composition of the image is something of an homage to the brilliant Crystal Palace sculptures constructed by Benjamin Waterhouse Hawkins between 1852 and 1855. Many species reconstructed by  Hawkins were shown with individuals in both reclining and standing poses, often with the standing animals looking rather more spectacular than those resting alongside. You can see this arrangement in several dioramas, including his famous Iguanodon, as well as his pterosaurs, Anoplotherium, Palaeotherium and Megaloceros (see below). I've always thought this enhanced the believability of Hawkin's work, showing his reconstructions as individual animals with distinct attitudes and personalities rather than species which simply engaged in one activity. Their poses are also frequently enhanced with the local topography, giving the impression the more active animals have ventured to higher ground to survey their surroundings or intimidate their audience. There's a lot to like about Hawkins' work and I think we may overlook him somewhat when considering the Great Palaeoartists, but that's a story for another day.

A gypsy-russel of Hawkins Crystal Palace sculptures, showing his flair for juxtaposed reclining and standing animals. Clockwise from left, Anoplotherium; Megaloceros (including bonus imagery of pterosaur researcher Michael O'Sullivan); Pterodactylus(?), and Iguanodon. If you've not visited these sculptures but are reading a blog like this one, you owe yourself a visit to Crystal Palce. Photos by Witton.

Behind the lousy wordplay attempted above lies an element of truth: is there a group of dinosaurs that undermines the classic concept of dinosaurs more than therizinosaurs, the feathered, pot-bellied, herbivorous theropods? Perhaps more than any other group, therizinosaurs highlight how much our understanding of dinosaurs has progressed not only since Hawkin's day, but even within the last few decades. Dinosaur concepts of the  1980s and 1990s stressed the more reptilian aspects of their nature, even in light of the undeniable birdiness of dromaeosaurs and other maniraptorans. Back then, dinosaur books and television shows presented lizards and crocodiles as the best modern analogues for dinosaurs, and birds were only mentioned as distant dinosaur relatives. We were told that dinosaurs weren't just reptiles in the taxonomic sense, but were reptilian in terms of their appearance and lifestyles, which echoed sentiments expressed as  long ago as the early 1800s. As we all know, early concepts of dinosaurs saw them as little more than gigantic super-reptiles, an idea best embodied by the models discussed above (and shown again, below), which were clearly primarily generated by wrapping modern reptiles around a few dinosaur fossils.

What's in a name?
Despite the advances in our concept of dinosaur palaeobiology since Hawkins' superlizards first roamed south London, we still use reptiles as modern points of reference for dinosaurs. If asked to concisely summarise Dinosauria in a single sentence for a lay audience, most of us would use the word 'reptile' somewhere. But, strange as it may seem to say, it doesn't really make much sense to introduce dinosaurs as 'reptiles' any more, and I think we do out of habit rather than good reason. Taxonomically speaking, the word 'reptile' is somewhat nebulous, with different definitions depending on its use. If you're old fashioned, you may define Reptilia as an artificial group containing all amniotes which are neither mammals or birds, but that doesn't work for dinosaurs as it excludes birds. Others would take the word 'reptile' as indicating most members of 'Sauria', a natural group containing all amniotes except for the synapsids. That's fine, but Sauria is a big and diverse group, so labelling dinosaurs as 'reptiles' is an incredibly loose taxonomic address. It may have worked several decades ago, when concepts of reptile evolution were pretty murky, but not in today's world of cladograms and robust sauropsid phylogeny. Introducing dinosaurs as reptiles is correct, but not very informative. It's also inconsistent with the way we describe modern dinosaurs. We don't typically introduce birds as a group of reptiles: they're simply a type of animal. Why aren't other dinosaurs just described as 'animals' then? It seems that introducing dinosaurs as reptiles is either wrong, imprecise or inconsistent with the way we treat modern dinosaurs.

Hawkins' awesome model of a super-reptile Megalosaurus. Note the modern dinosaurs at the top of the photo, which give me an excuse to link to this. Photo by Witton.
Compounding this is the fact that, anatomically speaking, modern reptiles aren't a great match for dinosaurs, and there is an obviously superior alternative. Sure, many dinosaurs were scaly-skinned, terrestrial animals which laid eggs, so the 'reptile' reference still has some merit, but modern birds are far closer anatomical and behavioural analogues to Mesozoic dinosaurs. Birds also offer scaly skin, terrestrial habits and egg laying to dinosaurs, as well as filamentous integuments, rapid growth, extensive systems of air sacs in the body and neck, erect stances, long necks, and long limbs with digitigrade extremities (among many more detailed aspects of their anatomy). Because of this, we're all pretty happy that Mesozoic dinosaurs were a lot more like birds than lizards, snakes or crocodiles, but we still preferentially mention reptiles rather than birds when introducing the group.

So why is the word 'reptile' at the heart of our basic concept of Mesozoic dinosaurs, when using birds as an immediate point of reference would make more sense both anatomically and taxonomically? Maybe it's simply because the word 'reptile' got there first. The reptilian nature of dinosaur fossils was appreciated before their birdiness, so we labelled them 'fossil reptiles'. And it stuck. We probably would not be in this position if the first dinosaur discoveries had been of deinonychosaurs, oviraptorosaurs or another, obviously bird-like species, rather than fragments of megalosaurs or ornithischians. What if evidence of feathered non-avian dinosaurs had been available to early dinosaur workers? It could have happened that way. Feather impressions are now known to occur in multiple fossil sites around the world, including some deposits which may not, at first, seem likely to preserve them. Even in sites where feathers do not preserve, the forelimbs of some dromaeosaurs preserve feather-anchoring quill knobs. It's not inconceivable, therefore, that the birdiness of dinosaurs could have been revealed much earlier on in dinosaur research history, and that the concept of dinosaurs as fossil reptiles may never have been established. Perhaps, given how essential the reptilian nature of dinosaurs seems to their popularity, dinosaurs would never have become anywhere near as popular if this version of history had played out. Just think: we may never have had Jurassic Park.

So there we have it, then: introducing dinosaurs as reptiles is a bit silly but, like a lot of language, we stick with it because of its established nature and ease of use. With hundreds of years of momentum behind this idiosyncrasy, I doubt we'll ever see dinosaurs labelled as anything else. However, that's not to say that educators or science communicators wouldn't benefit from occasionally tweaking the way that they introduce Mesozoic dinosaurs to their audiences. Perhaps replacing the word 'reptiles' with 'bird-relatives' every now and then will jar a few minds awake, and especially if said educators are trying to put some distance between modern dinosaur concepts and those of the past.


  1. Good article ! ''Bird-relative'' can be applied to coelurosaurians,but in the case of Ornitischians/Sauropods/basal Tetanurids (they share some features in with birds but no avian-like integuments) ?


    1. Hi Oliver,

      Thanks for the comment. Feather-like integuments are now known from megalosaurs (Sciurumimus) and basal ornithischians (Tianyulong), so they could fit under the 'bird relative' umbrella as well. And yes, sauropods are not currently known to have any feathers, but this is perhaps due to the scarcity of sauropod skin impressions rather than their genuine absence, as Mike Taylor explains over at Walking with Dinosaurs.

    2. Is it OK if I play Devil's Advocate?

      "Feather-like integuments are now known from megalosaurs (Sciurumimus)"

      To be fair, that's not a sure thing ( ).

      "And yes, sauropods are not currently known to have any feathers, but this is perhaps due to the scarcity of sauropod skin impressions rather than their genuine absence,"

      Again, to be fair, "the embryonic titanosaurs from Auca Mahuevo have full-body skin casts" ( ). I can't help but wonder why Taylor didn't mention them in his article.

      Good post, BTW. When I talk to laypeople, I usually refer to reptiles as "4-legged backboned animals w/keratin scales", dinos as "land-living reptiles w/an erect posture", & birds as "flying (or secondarily flightless) feathered dinos". Does that help?

    3. Hi Hadiaz,

      "To be fair, that's not a sure thing"

      OK, but whether Sciurumimus is a carnosaur or not doesn't really change my point: feather-like filaments were present in parts of the dinosaur tree that are very distant to birds. Even if they aren't, they're only one characteristic among many avian-like features we see in all dinosaurs.

      "Again, to be fair, "the embryonic titanosaurs from Auca Mahuevo have full-body skin casts" ( ). I can't help but wonder why Taylor didn't mention them in his article."

      True. Sauropods skin impressions remain very rare though and, as discussed in this post, we may want to exercise caution when applying the integuments of one or two species to an entire group. I wouldn't be at all surprised if some sauropodomorphs had some feather-like filaments on them somewhere.

    4. Well, given that lots of birds hatch naked as well, the one about embryonic enteguement is not really an argument against anything, really.

  2. Thank you for this, I think, somewhat philosophical post. I too have at times wondered what we would have made of dinosaurs if our earliest discovered fossils had been different ones. For example, imagine that the Liaoning dinosaurs would have been the first to be discovered. Surely, 19th century naturalists would have classified *Sinosauropteryx* or even *Tianyulong* as birds rather than as reptiles? Had that happened, 'class Aves' might have been understood to mean and include all dinosaurs throughout the 20th century. What do you think those same naturalists would have made of Liaoning pterosaurs with their obvious wings and prominent integument? Would those too have been considered birds?

    1. "imagine that the Liaoning dinosaurs would have been the first to be discovered."

      That's more of less the thought that inspired this essay. And yes, who knows what our Victorian counterparts would have made of those specimens? Reptile, bird, or a whole new class of animal? If a maniraptoran, something like Sinosauropteryx was the first, perhaps they'd have labelled it the 'first bird' as they did with Archaeopteryx. But as you say, what if it had been Tianyulong or Yutyrannus? It's harder to fit them into the Aves mould, as it is with pterosuars. I think the latter at least would definitely remain classed outside of Aves, but the discovery of their integument would have shaken up some prominent Victorian scientists. Richard Owen, for instance, was convinced that pterosaurs would be scaly because of their reptilian heritage, and would not entertain contrary thoughts by Harry Seeley, who thought otherwise. If nothing else, early discovery of the Liaoning deposits would have blurred the concepts of Reptilia and Aves considerably. I wonder what sort of part these deposits would have played in discussions of natural selection and evolution?

  3. Therizinosaurus as a giant pigeon is awesome. Seeing dinosaurs as more birdlike than reptilian makes sense to me. It's pretty easy to visualize theropods as birdlike. What about sauropods or ornithischians? Are there specific characteristics that we could consider as birdlike?

    1. Hi jck,

      Thanks for the comment. As mentioned above, all dinosaurs - including ornithischians and sauropodomoprhs - share a suite of features with birds:

      "Birds also offer scaly skin, terrestrial habits and egg laying to dinosaurs, as well as filamentous integuments, rapid growth, extensive systems of air sacs in the body and neck, erect stances, long necks, and long limbs with digitigrade extremities (among many more detailed aspects of their anatomy)."

      Granted, some of these features may be secondarily lost in some lineages (no-one seems to know why ornithischians lack pneumatised postcranial skeletons for instance), but they were at least there in dinosaur ancestry.

  4. Pedant alert! Reptilia is actually not the same as Sauropsida. Sauropsida is a stem-based group, so will always be more inclusive than the node-based Reptilia. Mesosaurs may be non-reptilian sauropsids, for instance. Or if turtles are diapsids and thus there are no living proganosaurs, all non-saurian sauropsids would be outside Reptilia, such as pareisaurs, Youngina and Petrolacosaurus.

    Also, the new Dromiceiomimus remains don't actually have quill knobs on the ulna, just carbonized markings.

    Interesting thought experiment re: Yixian fossils being found first, but if I had to guess I'd say oviraptorosaurs and paravians would be interpreted as birds, but not Sinosauropteryx, Yutyrannus or Tianyulong. The complete skeleton of Compsognathus was known in 1859 after all, and the feathers of Sinosauropteryx aren't obvious enough to convince BANDits today. Also, my impression is that back in the 1800s, classification was more subjective and fluid, and groups were more poorly known of course. So they might not have had much of an issue with a bird-like feathered group of reptiles that was just another blob leading out of Eosuchia and unrelated to Aves.

    1. Thanks for the comment, Mickey. I erred with the 'sauropsid' comment: 'Sauria' fits my intentions much better, and I must have misremembered the quill knobs in Dromiceiomimus. Post updated accordingly.

      Interesting thoughts concerning the 'Yixian comes first' speculations. I don't know that using modern BANDits is a fair analogue of Victorian scientists, though, given that their thoughts on bird origins are such a minority. I could be wrong, but I like to think that our Victorian counterparts would see Yixian or equivlanet fossils for what they were. But yes, it would be interesting to see where they drew the line between Aves and Reptilia. At a superficial level at least, is there that much difference between Sinosauropteryx and Archaeopteryx? Would be very interesting to see what definitions they came up with for the different groups.

  5. protofeather are collagen fiber dead skin were you can find in shark dolphin sea snake. sinosauropteryx prove that dinosaurs are crocodilian the fossil show that it had a hepatic piston a crocodilian feature not found in birds allso found diaghragm not found in birds too the diaghragm almost same as a alligator both have complete different lifestyle won is a quadrupedal other is bipedal and much older oh the compsognathus group have no mandibular fenestra that so a bird feature and no crocodilian sphenosuchian protosuchian had them so compsognathus group is a protobird dispite all reptilian feature but that not true no mandibular fenestra is a trait in crocodilian prehistoric alligator allso has no mandibular fenestra another crocodilian link. archaeopteryx feature are very different from compsognathus group archaeopteryx is true bird with no prefrontal bone with special bone in jaw to make upper jaw move very well dinosaur have very poor upper jaw movement and poor arm movement compsognathus group does the same archaeopteryx allso a perching bird a bird feature dinosaurs have croc knee bird have bird knee a fix femur the thigh bone so can help them to fly if it is not fix they will die compsognathus have real ziphodont teeth like crocodilian not the fake kind like the maniraptoran compsognathus teeth replacement is like dinosaurs archaeopteryx is like a bird compsognathus group have small reptilian brain archaeopteryx have large bird brain microraptor is true bird all dromaeosauridae are true birds its not a protobirsd since microraptor is bfore most of dromaeosaur like velociraptor de------- so call proto bird in the past. microraptor have true flight feathers all bird with flight feathers are flying birds allso have flight sternum .its over they found the dinosaur allready its time to move on its tough to keep the scam going with ALL the IMFORMation on the net study crocodilian most of them almost gone and your fault

  6. This comment has been removed by a blog administrator.