Pterosaurs as dinosaurs, or not

The origins of pterosaurs have of course come up on here a number of times but one hypothesis remains largely untreated and that is the idea that pterosaurs might have been descended from theropod dinosaurs. Normally I would not go into the details of such marginal ideas (it has only ever been mooted once and that was not in a formal reviewed scientific publication) but the fame of the source makes it worth revisiting. That source is the legendary ‘Dinosaur Heresies’ of Bob Bakker, probably the most famous and influential dinosaur book (both to the general public and the scientific community as a whole in as much as it summarised many years of research) ever written. That is the reason I assume this hypothesis has survived and occasionally re-surfaces in popular writing and the internet.

Amongst many things in the tome, Bakker suggests that pterosaurs might have been descended from basal theropods. (Actually he says ‘dinosaurs’ but based on the huge size of even the basal sauropodomoprhs compared to pterosaurs, and the herbivorous diet of ornithischians I rather assume he means theropods, especially given how much more we knew about them than the basal taxa of the other clades at the time, it hardly changes my points in any case). At the time (and indeed now) pterosaurs were widely considered to be derived archosaurs that derived from the dinosauromorphs (those archosaurs that are on the stem lineage to dinosaurs, but not quite dinosaurs, and he notes their importance). So in a sense this shift in position (from just outside the dinosaurs, to just inside them) was not a major one, but it worth discussing.

Obviously this was a book aimed at the general public primarily and thus I cannot be critical of the lack of scientific rigour in that there is no cladistic analysis attached to this (and hell, it was the mid 80’s), or a detailed list of anatomical characters. However, he does include a specific set of characters which he says unite the pterosaurs with the dinosaurs and these we can examine and assess.

He writes that the “S-shaped neck, the simplified shoulder, and the bird-type ankle are excellent clues to the ultimate ancestry of the dragons [pterosaurs] – the quick and agile early dinosaurs of the Triassic period”.

First off we can actually talk about the “quick and agile” part which could be part of the definition, as this implies a similarity between the two. Of course the actual status of basal pterosaurs as quick and agile or otherwise is unknown so we can rule that out, though even if they were of this kind, this hardly separates dinosaurs from dinosauromorphs either so it hardly favours Bakker’s dinosaur hypothesis. (Of course he could have meant ‘quick and agile’ simply meaning not sauropodomorphs).

Secondly let’s take the shoulder. While I said I would not be critical of a lack of detail, it does make it harder to assess the character – what is a simplified shoulder? Compared to what? What does it look like, or not look like? In any case, again the dinosauromorphs do not have very different shoulders to those of theropods or other basal dinosaurs so assuming the pterosaurs do have a simple shoulder (and I am unconvinced, the fossil record of basal pterosaurs is problematic) again this does not differentiate between the two groups.

Next up is the “bird-type ankle”. I definitely don’t want to delve into archosaurian ankles in detail anytime soon but the same issues are present again. Dinosauromorphs and basal dinosaurs share similar ankles, and we are not really certain what kind of ankle basal pterosaurs had, though actually it was probably similar to those of these two clades. Once more this does not distinguish between the dinosaurs and dinosauromorphs.

IMGP2966Finally we get to the shape of the neck. Now once again there are few differences between the dinosaurs and dinosauromorphs in this regard, and given the lack of detail it’s hard to get beyond the fact that both have flexible necks that are quite long. However in the pterosaurs this is certainly not the case. While again the exact state of the neck in basal pterosaurs is again hard to determine, those animals for which we do have good 3-D necks preserved suggest that (in the words of Dave Unwin) pterosaurs were a “stiff-necked breed”. The articulations in the neck would not have allowed much flexion in the dorso-ventral plane, let alone the kind of S-shape retraction (something akin to that of pelicans say or the illustrated peacock) that Bakker (and for some unfathomable reason a whole load of palaeoartists) favours.

In summary, none of the characters can really be used to differentiate between the dinosauromophs and the dinosaurs, and in the neck at least the character seems positively against such an association with pterosaurs. Of course the book does not provide the details that might make these characters far more explicit and easy to assess, but certainly the idea has never appeared in any peer-reviewed literature (to my knowledge) and no phylogenetic analysis has ever recovered a direct association between the pterosaurs and dinosaurs. It’s pretty safe to lay this concept to rest, interesting and appealing though it might be: pterosaurs and dinosaurs just do not share any deatailed characters in common that are not already present in the dinosauromorphs, or might not also be present in them (like body fibres). I’ve said it many times before in terms of incorrect media reports but its worth saying once more in this context, pterosaurs are not dinosaurs.

12 Responses to “Pterosaurs as dinosaurs, or not”

  1. 1 Christopher Taylor 18/05/2009 at 11:41 am

    I’d always read Bakker as simply expanding the coverage of the term “dinosaur”, rather than any argument for a different phylogenetic position for pterosaurs as such. Or, to put words in his mouth, dinosauromorphs are phenetically pretty much ‘dinosaurs’, so why not just call them dinosaurs?

    I hasten to note that I’m not personally supporting such an argument – to be honest, I was always rather underwhelmed by the ‘Heresies’.

  2. 2 David Hone 18/05/2009 at 12:23 pm

    That had been my reading until I went over it again. Given that he specifically discusses Lagosuchus and Scleromochlus as being dinosauromorphs and NOT dinosaurs though, it’s hard to see how a possible ‘extended’ Dinosauria could include pterosaurs, but not the dinosauromorphs.

  3. 3 Mickey Mortimer 18/05/2009 at 1:48 pm

    Gauthier (1984) grouped pterosaurs and dinosaurs together as Ornithotarsi, with Lagosuchus just outside this clade. This was based on several characters, including a “birdlike tarsus” and “interclavicle absent and clavicles reduced”. The “s-shaped neck” is covered in his diagnosis of Ornithodira (which included Ornithotarsi and Lagosuchus). I think this arrangement is what Bakker had in mind. He wouldn’t have had pterosaurs descending from theropods alone based on those characters, since his 1974 paper with Galton has the birdlike tarsus and simplified pectoral girdle as dinosaurian characters, and theropods share these and s-shaped necks with ornithischians and sauropodomorphs, as you note. You are correct insofar as the basal dinosaurs hypothesized by Bakker probably resembled theropods, which has been confirmed by new discoveries.

    If you want theropod pterosaurs, how about…
    – enlarged skull (reversed in Ornitholestes, most maniraptoriforms, etc.).
    – quadratojugal squamosal contact absent (reversed in tetanurines)
    – pneumatic cervical centra.
    – at least four sacral vertebrae.
    – elongate distal caudal vertebrae.
    – strap-like scapula (reversed in coelophysoids).
    – thin walled limb bones.
    – penultimate manual phalanges most elongate.
    – trenchent, recurved manual unguals II and III.
    – metacarpal V absent.
    – preacetabular process elongate.


  4. 4 David Hone 18/05/2009 at 2:10 pm

    Well I’d disagree with a ton of those purely in terms of being able to demonstrate that they actually occur in basal pterosaurs (the ‘4 sacrals’ one is incredibly problematic for example) or for that matter quite where you want to draw the line for ‘theropods’. Are we talking Herrerasaurus or Coelophysis, or ceratosaurs or what? Which taxa can we include and thus which of those characters are or are not present in whatever one we consider – metacarpal V is present in at least some basal saurischians for starters, plus of course some other obvious convergence. (I know this is just for fun, I’m just making a point).
    Obviously they do share some characters in common (pneumaticy being a potentially big issue depending on what you can ‘pneumaticy’) but in terms of recovering anything like a really close tie between actual Dinosauria and Pterosauria, the phyloghenetic analyses don’t provide any measure of support. I’m not trying to hammer Bakker for his views (it’s a popular book, written for a non-technical audience and in an age where many of the taxa and methods we take for granted were not avialbale), merely point out to at least a few readers that while yes, there is one technically published hypothesis for linking dinosaurs and pterosaurs, there’s really no support for it now, and frankly wasn’t much then.

  5. 5 Mickey Mortimer 18/05/2009 at 2:52 pm

    I was thinking of Avepoda for that list, with Herrerasaurus as a possible next closest relative since it has characters 1-2 and 5-9. Admittedly pterosaurs aren’t my area of expertise, but Preondactlus and/or Eudimorphodon have characters 1, 3 and 5-10 based on Dalla Vecchia (1998), Wellnhofer (2003) and Butler et al. (2009). Are there basal pterosaurs that are known to lack characters 2 and 4?

    I’m sure pterosaurs wouldn’t end up as theropods in any well done analysis (you’d have to get past the missing dinosauromorph, dinosauriform, dinosaurian and saurischian characters for instance), but Novas (1993) had a similar amount of characters supporting a theropod placement for Eoraptor and Herrerasaurus as I have above. Just goes to show the caution needed when conclusions are from small character matrices or matrices that only include characters which support the author’s phylogeny.

  6. 6 David Hone 18/05/2009 at 4:06 pm

    That’s really the point I wanted to get across. If you start to look at small subsets of data, pick only general characters, ignore other synapomorphies, obvious lines of convergence, intermediate taxa etc. then you can get a misleading picture. You can find plenty of synapomorphies between Aves and Mammalia if you look for them, but it doesn’t make them sister taxa.

    Some basal pterosaurs definately do not have 4 sacrals, and indeed as a general point, people need to stop using Eudimorphodon as a ‘basal’ pterosaur / outgroup just because it’s Triassic, quite complete and a rhamphorhynchoid. I don’t want to get into details here (well, not right now, especially not specific characters) as I have a huge piece going through slowly covering all of this, just making a couple of points.

  7. 7 Zach Miller 19/05/2009 at 2:00 am

    Yes, I thought Dimorphodon was the “basalmost” known pterosaur. Can’t quite take the theropod-pterosaur connection seriously, though (as, it seems, I shouldn’t). Even though pterosaurs are pretty firmly entrenched in Ornithidira, it would be nice to find an animal more basal than Dimorphodon if only to cement that placement.

    A girl can dream! 🙂

  8. 8 David Hone 19/05/2009 at 9:02 am

    Not quite, Kellner and Bennett both recover the anurognathids as the most basal clade in their analyses (and I favour this interpretation) and Dave Unwin get’s Preondactylus, followed by the dimoprhodontids (of which the only really complete one is Dimoprhodon). Given the terrible preservation of Preondactylus and the fact that it’s a juvenile, we really should probably use a combination of an anurognathid, Dimoprhodon, Preondactylus and perhaps anopther taxon or two to act as a ‘basal pterosaur’ for analysis, but as far as I’m concerned there is little place for Eudimorphodon. It was understandbaly used a lot in the past when it did appear to be basal, complete and Triassic, but the ‘basal’ part now appears to be quite wrong.

  9. 9 Zach Miller 20/05/2009 at 2:36 am

    Anurognathids, really? Awesome! So pterosaurs were ancestrally bloodthirsty? *LOL*

  10. 10 David Hone 20/05/2009 at 8:43 am

    Well, that is my take on the character data and the analyses available, though Dave Unwin and Brian Andres would disagree strongly.

  11. 11 monado 31/05/2009 at 2:07 pm

    Now if he’d just said that both basal dinosauromorphs and pterosaurs probably had protofeathers, I could agree.

  12. 12 David Hone 31/05/2009 at 4:28 pm

    Well there is still a huge question over the homolog of those structres and understandably most people are all but refusing to comment becuase we just don’t know enough about either type (the ‘fur’ in pterosaurs is definately different in at least some respects to the basal ‘protofuzz’ in maniraptorans. So wile they both have structures of soem kind, the idea that they both have the same structure, or even one with a shared evolutionay hisotry is very far from certain right now.

Comments are currently closed.

@Dave_Hone on Twitter

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 553 other followers

%d bloggers like this: