Thursday 19 March 2015

Short-necked azhdarchid pterosaurs - say what?

LPV (FGGUB) R.2395, our unnamed short-necked azhdarchid from Maastrichtian deposits of the Hațeg basin. Prints of this chap are available.

Odds are that most regular readers of this blog are familiar with azhdarchid pterosaurs, the toothless, often gigantic flying reptiles which increasingly dominated pterosaur evolution in the Cretaceous. With a better fossil record than most pterosaur lineages, they are among the best understood and most researched of all flying reptiles. In recent years our understanding of their anatomy, distribution and palaeobiology has advanced considerably.

Since azhdarchids were recognised as a group in the 1980s it has been realised that their most striking and characteristic features pertain to their neck anatomy. In both relative and absolute terms they have the longest necks of any pterosaurs, stretching their mid-series cervical vertebrae to long tubes with reduced features. Cervicals IV - VI are especially long and tubular, with the mid-portions of the neural spines reduced to such a degree that they are effectively split into anterior and posterior sections. The only significantly developed features of these cervicals are bulbous zygopophyses and condyles, which nestle together so snugly that neck articulation seems limited. Regular readers will know that this has considerable bearing on the likely habits of these animals, probably precluding strenuous lifestyles such as skim-feeding or pelican-like scooping. The extremes of their neck bones - cervicals I - III, and the somewhat 'dorsalised' VIII - IX - are less modified, although some azhdarchid weirdness infects these too in terms of length, neural spine shape, or both.

With azhdarchid cervicals being so diagnostic, they can be identified when found in isolation and even when only poorly preserved. Indeed, many azhdarchid occurrences are represented by isolated cervicals - they compete with jaw tips for the most common type of azhdarchid fossil. In recent years, the complete neck osteology of the Santonian, central Asian azhdarchid Azhdarcho lancicollis has been documented in some detail (Averianov 2010), permitting some insight into which specific part of the neck sequence isolated neck bones represent. It can be a little tough to tell a cervical IV and V apart - they seem very similar, except that V is invariably the longer of the two (indeed, the longest of the entire neck) - but we can now at least tentatively identify bones from the rest of the sequence. This is a major breakthrough, meaning that isolated cervicals can tell us a lot more than just where an azhdarchid was preserved: their potential for taxonomic and biomechanical studies has been increased considerably.

Eyes to Romania

A lot of new azhdarchid material, including said isolated cervicals, has recently been emerging from various Maastrichtian deposits of Romania. These sediments represent the rivers and lakes which once ran through Hațeg island, an ancient setting famous for its dwarf dinosaurs and the enormous azhdarchid Hatzegopteryx thambema. The Hațeg pterosaurs, and their neighbours from Transylvania, have been the focus of a number of recent pterosaur papers and, this week, a team of researchers (including Mátyás Vremir, myself, Darren Naish, Gareth Dyke, Stephen Brusatte, Mark Norell, and Radu Totoianu) published another in American Museum Novitates (Vremir et al. 2015). It reports the discovery of just one bone* - the near-complete azhdarchid cervical LPV (FGGUB) R.2395, from Hațeg Basin red beds - but it's enough to cast new light on these otherwise familiar azhdarchid fossils, as well as the general evolution of azhdarchids themselves.

*So yes, the picture at the top is 1% fossil data, 99% palaeoart polyfiller.


Line drawing of LPV (FGGUB) R.2395, isolated azhdarchid cervical IV or V. It don't look like much, but it's got it where it counts. From Vremir et al. 2015.
R.2395 (above) is not a large vertebra, being an estimated 100 mm long and 44 mm across the prezygapophyses when complete (it's missing the posterior portion, including most of the condyle and postzygagpophyses, as well as the left prezygapophysis). Our paper provides a long discussion about the likely portion of the neck represented by R.2395, concluding that it is probably a cervical IV or V. In azhdarchids, these vertebrae are among the longest of all, sometimes being eight times longer than wide. R.2395 was clearly a lot stouter than this however, barely being twice longer than broad. This proportion is unique among azhdarchid CIV and Vs, and there was some discussion among our team as to whether or not this was sufficient to erect a new taxon. In the end we decided the material was too scant to support a name of its own, but it is almost certainly a new species.

The width of R.2395 suggests the neck owner was not a tiny animal. It's difficult to get a size estimate from a single neck bone, especially with cervical length varying so much taxonomically across Pterosauria, as well as ontogenetically. However, vertebral width is a little more stable with respect to overall body size, and that of R.2395 indicates an animal on the small size for an azhdarchid - an arm-wavy wingspan estimate of 3 m seems about right. When we plugged the length of R.2395 into a database of near-complete azhdarchid necks, we found the estimated CIII - CVIII length was a paltry 352–419 mm (the range depends on whether it represents a CIV or CV): 23-41% shorter than the estimated neck length of the similarly-sized Transylvanian azhdarchid Eurazhdarcho, langendorfensis and notably shorter than neck of the smaller (2.5 m wingspan) Chinese azhdarchid Zhejiangopterus linhaiensis  (measured neck length 502 mm). It seems R.2395 did indeed have a short, robust neck for an azhdarchid of its size.

The possibility that R.2395 represents a short-necked juvenile of a long-necked species was something we looked into. After all, pterosaur necks seem to increase in length disproportionately to body size, and a 3 m wingspan would leave a lot of growing room for several azhdarchid species. However, the bone texture of R.2395 is characteristically 'polished' and avascular where well preserved, which has been suggested for some pterosaurs as an indicator of skeletal maturity. This observation is bolstered by the sharply ossified anatomy of R.2395: young pterosaur skeleltons tend to have rounded, poorly defined features, but our vertebra shows a surprising amount of sharply-defined detail in its superficially simple form. We don't know how old R.2395 was when it died, but it did seem to have a very well-ossified skeleton: it was likely at, or very near to, full size, and further neck growth seems unlikely. 

What does it mean to find short necks in a clade thought to be defined by long necks? We have no idea how this plugs into azhdarchid evolution, although the retention of all azhdarchid cervical features apart from the enhanced length is of interest there. We might also conclude that suggestions azhdarchids were all anatomically similar (e.g. Witton and Naish 2008) are questionable. There are doubtless some functional and palaeoecological implications of this discovery as well - we speculate that neck mechanics and strength likely differed between long- and short-necked azhdarchids - but further remains are clearly needed to say anything substantial about specific functionality. And... in truth, I'm biting my tongue here: there's tons to say about this, but I don't want to scoop other papers which are in prep and review. Both myself and Darren Naish have been hinting at some of this short-necked azhdarchid stuff for a while now, and heavily implying that we have things to say about Haztegopteryx as well as this much smaller animal. Some readers may, therefore, be wondering why there's no discussion of giant pterosaurs here. We were both hoping that this would be out by now as well, but it turns out that this parallel-running project has been published first - that's just how these things work out sometimes! Further details on these finds is coming, with other publications in states of progress which will add to this picture. Watch this space, in other words.

To end on a less cagey note, it is worth briefly mentioning why the seemingly maximum size of R.2395 is rather interesting. It is often said that small pterosaurs are absent from the upper Cretaceous, a fact sometimes controversially attributed to competition with birds. If R.2395 really is an adult, it joins Eurazhdarcho and the 2.5 m wingspan Montanazhdarcho in suggesting that smaller pterosaurs were not absent, and perhaps even not uncommon, in the latest Cretaceous. Granted, these species are not as small as some pre-Cretaceous pterosaurs, but they are certainly size-consistent with taxa found in Lower Cretaceous Lagerstätten of China and Brazil. I do wonder if the lack of sites of exceptional preservation in the Late Cretaceous has resulted in under-sampling of smaller pterosaur species in Late Cretaceous rocks, and that drawing conclusions about the Cretaceous decline of diminutive pterosaurs, and associated competition with birds, is premature. An elephant in the room here is the scarcity of small, Late Cretaceous juvenile pterosaurs: we know they had to exist, and yet they are exceptionally rare fossils. There is clearly a preservation bias against these small individuals - can we rule out that the same bias was not acting against small adults, too? That might be nonsense, but I do wonder if  we sometimes take the - knowingly poor - fossil record of pterosaurs a bit too literally.

A quick plug for a good cause

Finally, some readers will know that I like to bang drums about Supporting Original Palaeoart, and that it's often independent artists who're providing some of the more interesting and creative palaeoart projects out there. One of these comes from David and Jennie Orr (David is perhaps best known around these parts for founding Love in the Time of Chasmosaurs, as well as a his unique, stylish artwork), who have launched an Indiegogo campaign to fund their book Mammoth is Mopey. It's a wonderfully illustrated book for younger readers showing a different prehistoric animal for each letter of the alphabet. The animals in question aren't the same tried-and-tested taxa we see in every kids book however: they're the likes of Brontomerus, Jeholopterus, Gorgonops and so on, and each is wonderfully illustrated in David's Bézier-curve-loving style. It looks great, and I've already pledged enough for two copies: one for my nephew, and another for me (hey, you're never too old to have an alphabet refresher, right?). An example image from the book is below - I can't wait to see the rest.

Artistic Ankylosaurus is artistic. From Mammoth is Mopey, which you can support here. Illustration by David Orr.

References

  • Averianov, A. O. (2010). The osteology of Azhdarcho lancicollis Nessov, 1984 (Pterosauria, Azhdarchidae) from the late Cretaceous of Uzbekistan. Proceedings of the Zoological Institute RAS, 314(3), 264-317.
  • Witton, M. P., & Naish, D. (2008). A reappraisal of azhdarchid pterosaur functional morphology and paleoecology. PLoS One, 3(5), e2271.
  • Vremir, M., Witton, M., Naish, D., Dyke, G., Brusatte, S. L., Norell, M. & Totoianu, R. 2015. A medium-sized robust-necked azhdarchid pterosaur (Pterodactyloidea: Azhdarchidae) from the Maastrichtian of Pui (Haţeg Basin, Transylvania, Romania). American Museum Novitates 3827, 1-16. 

38 comments:

  1. Looking at the illustrations, I get the strong impression that this is just the front half of a vertebra, with the undeniable distortion creating the illusion of a short cervical. And even if it were short, it offers only weak support for the existence of a new taxon. The correct inference is not: "The vertebra is short. But its morphology is that of a fourth cervical. Therefore we have discovered a new species" but "The morphology of this heavily damaged and distorted pterosaur fragment resembles that of a fourth cervical. But such short fourth cervicals are unknown in the Azhdarchidae. Therefore it is either not short or not a fourth cervical - or not an azhdarchid in the first place".

    Still a very interesting and suggestive specimen, though.

    Mark Konings

    ReplyDelete
    Replies
    1. Thanks for the comment Mark. I disagree with the idea this we are dealing with just the front half of a vertebra, though. As can be seen in the illustration, the posterior end of the neural spine is clearly present, and the specimen flares posteriorly from the corporeal 'waist' that typically sits towards the back end of azhdarchid mid-series cervicals. Neither a 'double peak' neural spine or corporal waist are recorded from the front portion of an azhdarchid cervical - and we have a good specimen bank to back that up. Plus, while I grant you that the posterior end of the vert is not in great shape, a remnant of the left postzyg is definitely there. If all these features are the result of crushing and distortion, they have converged on a morphology exacting that of the posterior end of an azhdarchid cervical. I think the interpretation that most of the vert length is represented is more parsimonious, and especially so given indications from other Romanian pterosaurs that short-necked azhdarchids are a genuine phenomenon.

      Have you looked at the specimen photos? These features, and their interpretation, are pretty clear in those. A link to the paper is provided in the text.

      Delete
  2. Mark K - interesting that a vertebrae which possesses a postzygapophysis and other features of the posterior articular end might be considered "just the front half of a vertebra". Also interesting that you seem to know more about the specimen that those of us who examined it and described it. It is somewhat crushed and incomplete, yes, but it's a near-complete vertebra, identified with confidence to Azhdarchidae, corresponding most closely to CIV, and which has unusually robust proportions distinct from those of other azhdarchid taxa. In other words, your scepticism (and arrogance, sorry) seem misplaced.

    ReplyDelete
  3. This is a wild idea, but is it possible the spread of angiosperms as dominant plants in the late Cretaceous, rather than competition with birds, could be somehow linked with the decline in small pterosaur fossils? It could mean a loss of habitats, or nesting grounds, of pterosaur clades traditionally adapted to other environments (resulting in decline or restricted distribution, but not necessarily a complete extinction).

    My question is inspired by the fact that you clearly put your short-necked azhdarchid in a foresty-looking area... any particular reason behind it? A Wikipedia summary of Hateg island conditions in the Cretaceous indicates lots of rivers and lakes and dry climate paired with tropical plants.

    http://en.wikipedia.org/wiki/Ha%C8%9Beg_Island#Climate_and_ecology

    ReplyDelete
    Replies
    1. Funnily enough, there are published contrary speculations about pterosaurs and plants getting along very well, and that pterosaurs were Cretaceous angiosperm seed dispersers. I'm not sure how well founded that is, but pterosaurs certainly co-existed alongside flowering plants for a long time before their Late Cretaceous decline - I'm not sure there's any reason to think the spread of angiosperms killed pterosaurs. Maybe there is something to the idea of changing habitats and climates, but I imagine there's a suite of factors contributing to pterosaur decline.

      Delete
    2. Interesting. So I guess this could have led up to an exact opposite situation (some dying out because of a potential pterosaur decline?).

      I may rely a bit too much on Wikipedia, but it includes the following line on angiosperms: "They diversified enormously during the Lower Cretaceous and became widespread around 120 million years ago, but replaced conifers as the dominant trees only around 60–100 million years ago." Unless I'm mistaken, non-azhdarchid pterosaur fossils seem to decline somewhere around the 80 million-years-ago boundary.

      While the reason I mentioned 'nesting grounds' is because there were probably stark differences between pterosaurs and birds when it comes to reproduction. Birds (and related dinosaurs) could build nests in foresty areas without issue, but I don't imagine pterosaurs doing the same, with current evidence being towards shabby mud nests in isolated areas and no traces of parental care.

      Delete
  4. Since Hatzegopteryx also had shorter, stouter neck vertebrae, do you think this might correlate with more macropredatory tendencies in this new specimen? If that'd be the case, then this would join Hatzegopteryx and Thalassodromeus as "pterosaurs of prey".

    Additionally, in your book you suggest that Navajodactylus is not an azhdarchid, and that there's some doubts about its "validity". Does it mean that it isn't a pterosaur?

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. I'm a little confused by the definition. By 'macropredatory' do you mean 'a big predator' (which might include a lot of other pterosaurs), 'a really big predator' (does this new species count?), or 'a predator of big prey'? (which, considering the most relevant google results for 'macropredator' turn up a TetZoo article about lions predating on elephants, might even exclude Hatzegopteryx...)

      Delete
    3. I'm going to refrain from saying anything about Hatzegopteryx for the time being - we'll have a lot to say about it when the time comes.

      I will agree with warren that 'macropredation' is probably unlikely for any known pterosaur. As with modern, non-raptorial birds, they seem best adapted for subduing prey smaller than themselves. The question we're asking about these new, robust specimens is 'how much smaller?'.

      On Navajodactylus: There's no doubt it's a pterosaur, but I question its validity as a diagnosable genus. The holotype (distal wing metacarpal and proximal wing phalanx, if memory serves) is not a portion of pterosaur anatomy typically considered diagnostic to species level, and the descriptive paper did not provide convincing evidence to the contrary.

      Delete
    4. Hategopteryx might be an exception to the small prey thing. I mean, what else is feeding on such big stuff, and why else would it deviate so much from it's relatives? It wouldn't be out of the question to guess that it might be going after 10-15ft Telmatosaurs and Rhabdonts.

      Delete
    5. To clarify, "macropredatory" in regards to proportionally large prey. According to your book, Thalassodromeus might have been exceptional among pterosaurs in this regard, and Hatzegopteryx was adapted to tackle larger prey than normal for azhdarchids.

      Of course, I'm not suggesting that these animals were the eagles of the Mesozoic, but even species not specialised for macropredatory habits, like gryphon vultures, are known to kill large animals on occasion.

      Delete
    6. I have the idea of azhdarchid super-storks so firmly in my head that for a while I imagined Hatzegopteryx swallowing rhabdodonts whole... How did you imagine it, Sean?

      If guarding and dismembering a carcass, vultures seem a bit better equipped for the job. Marabou storks, a slightly closer analogy maybe, often need vulture help to break into them. How do the storks take apart the carcass when they predate on flamingos? How applicable is that to azhdarchids?

      I should also take the opportunity to compliment the artwork in this blog entry too. ;) Very natural. Looks like it's quietly watching something (the viewer?) about cock it's head in curiosity.

      Delete
    7. It's worth to note that Thalassodromeus possess scizzor-like jaw tips, while several azhdarchids have distinctive "short profiles", so a stork analogy might not be the best in terms of function. For the former at least one can assume a slicing/ripping potential.

      Delete
    8. Haha! Someone must now draw that...

      I was more thinking killing it with the beak (probably pecking, and just general overpowering) and then eating it like most predators would; bit by bit. Thought swallowing Zalmoxes whole might not have been entirely impossible... if potentially fatal. Another thing I now want someone to draw...

      Delete
    9. To clarify, they'd probably pick off small bits of meat at a time; I don't see an azdarchid beak being very well at ripping off large chunks of meat. So it'd be an inconvenient and probably rare occurrence, I suppose.

      Delete
    10. Thanks for discussion, guys. Some interesting points here. Again, I don't want to tread on my own toes too much, but a few points:

      "I mean, what else is feeding on such big stuff,"

      Not all environments need macropredators capable of killing the biggest species around. Some environments cannot sustain such niches because other factors - say, generally limited resources - prohibit large populations of big prey animals. In the case of Hateg, Hatzegopteryx is certainly one of the largest predators there, but it's still only likely massing a quarter-tonne or thereabouts. That's still some way short of the masses seen in some of the larger Hateg dinosaurs.

      "even species not specialised for macropredatory habits, like gryphon vultures, are known to kill large animals on occasion."

      All we can do as functional morphologists is predict the general behaviour of these animals - what their skeletons are adapted to handle on a day to day basis. A qualitative assessment of Hatzegopteryx bones certainly indicates that they were more robust than typical azhdarchid material, and the ability to predate larger animals is possible one explanation for that. But we're also still dealing with fragments of this animal (the unpublished stuff is still just fragments), so we cannot make any predictions about its precise habits at this stage. Based on what we know of Hatz, the 'terrestrial stalker' model seems to fit, but I don't think we can say anything more specific than that.

      Delete
    11. Well, I can agree that the larger sauropods were probably off the menu at a certain size.

      Delete
    12. Carliro: I wasn't aware. Thanks for that. (I need to read moar.)

      Sean: "Someone must now draw that..."

      Pretty much why I barged into this comment thread!

      Delete
  5. But there is no claim that the fill-in forms a sort of natural endocast? If not, the posterior profile of the specimen is basically sculpted matrix. At the "waist", the specimen is severely fractured. This creates the possibility that the waist is an illusion, compacta fragments having been spread by the same force that broke them. There is also no unbroken line of bone fragments towards the postzygapophysal facet. I have to admit that, after magnifying the commendably high quality pictures, it certainly seems an authentic facet. That of course doesn't mean it's at its original position. The entire posterior end might have been caudocranially crushed.

    The again, if the shortness is authentic but the waist is an illusion, this would remove one of the objections against it being a third cervical. A few months ago, SVPW showed a picture of TMP 92.83.7, a presumed third cervical. The neural spine, hypapophysis and "pinch" differences with LPB (FGGUB) R.2395 seem very relative.

    Mark Konings

    ReplyDelete
    Replies
    1. I'm not seeing the issues you raise here. Having spent hours with this specific fossil (and drafting the description, and illustrating it) I can see no evidence to support your interpretation, and plenty of reasons to find it less parsimonious than our much simpler explanation (that being, it is what it looks like). I see little reason in arguing this point further.

      On the cervical ID: Definitive CIIIs of Quetzalcoatlus and Azhdarcho have relatively tall neural spines without a mid-portion break, as well as distinctive cotyle morphologies. The specimen shown at SVPOW! was associated with other azhdarchid material but no other cervicals - to my knowledge, no-one has presented a good case for it being a CIII (Godfrey and Currie 2005 certainly didn't, who described and illustrated it. They make no claim as to what vert it is). To me, it looks more like a CIV or V. Our paper does cover the distinctions between these verts - please see that and the citations we give for why we don't think our new specimen is a CIII.

      Delete
    2. Yes, but that is exactly the point. Precisely because we readers have not been able to examine the specimen personally, the value we'll ascribe to the article will depend on the degree to which you are able to bolster your interpretation. Now, if the material is near perfect and no controversial conclusions are drawn from it, simple practicality demands that we trust your primary observation. In this case however, a far-reaching claim is based based on the posterior end of a vertebra consisting at best of a partial and fragmentary single postzygapophysis. In such circumstances the "trust me, I'm a palaeontologist" stance does not suffice. You'll have to treat the claim as a real hypothesis, provide predictions and counterfactuals and test.

      Indeed, the vertebra at first blush seems short. But that could have been a preparation artefact. It would have been helpful if a picture had been provided of the condition of the specimen before it had been prepared. Was the matrix surrounding the postzygapophysis free of large bone fragments?

      It is certainly possible that TMP 92.83.7 is a fourth cervical. However, that would imply that there are three azhdarchid CIV morphologies: elongated with waist (standard); short with waist (LPB (FGGUB) R.2395); and short without waist (TMP 92.83.7). Assuming that the LPB (FGGUB) R.2395 waist is a artefact of taphonomy reduces this number with one and is, well, more parsimonious.

      Of course, all this is moot when the material you hinted at includes an articulated cervical series. Preferably two, one for each taxon :).

      Mark Konings

      Delete
    3. You're ignoring the posterior portion of the neural spine - how do you explain that? Is that a preperation artefact too, along with the postzyg and waist?

      Look, if we thought this was just the end of a vertebra, we would have said so. If we thought that was a compelling idea, we would have discussed it. But none of us have seen any indication that the specimen is half complete, and that includes the members of the research team which found it in the field. I'm not saying 'I know better than you': I'm saying I disagree with your ideas because there are features on the specimen indicating that it was very was near complete.

      And what is wrong with there being some variation in azharchid cervical morphology? I don't see how explaining the morphology of R2395 as serendipitous preservation is more parsimonious that simple variation between species.

      Delete
    4. I'd honestly find things more odd if a group lived throughout the entire Cretaceous, and was in this case influenced the weirdness of island evolution, and didn't have some odd variation.

      Delete
    5. I don't doubt that you didn't think it was just half a vertebra. But you should have entertained the thought. If the field team secured the specimen following a certain Gestalt, this will influence any subsequent preparation and prejudice analysis. Don't get prejudiced. If your main hypothesis is "this a short vertebra", the obvious alternative hypothesis to test against is "this is a long vertebra". From the first hypothesis certain predictions can be derived, such as the presence of a posterior neural spine peak. And yes, it's there. So that's a nice corroboration. The second hypothesis also generates predictions. One of them is that the vertebra should show the morphology of a CIV or CV. And darn, it does. Now I'm not saying "your pet hypothesis has been falsified and we all know since Popper that corroborations ain't worth a brass farthing" because that argument can easily be turned around. After all, the second hypothesis is corroborated and falsified too. The point is that to solve the contradiction we have to introduce ad hoc hypotheses. Can't be helped. One of these might be to assume that the peak is an artefact. The other might be that there existed a hitherto unsuspected clade of short-necked azhdarchids. Now, these ad hoc hypotheses are not equal in probability. It's much more likely that a single damaged and cracked specimen is deceptive than that an entirely new group of animals existed. Remember the story of Cope and the Elasmosaurus. Of course the conclusion should change dramatically if you have found half a dozen vertebrae with the same properties :).

      The same seems to apply to the identity of the vertebra. Is not some slight serendipity of a single specimen much more likely than the evolution of a specific extra neck morphology? Moreover, the serendipity isn't really there to begin with. It isn't as if you had predicted, on independent grounds, that such a vertebra existed. A specimen was damaged and cracked; the fortuitous result of this you made into a new hypothesis. Furthermore an alternative explanation is available in the well-known phenomenon that often interspecific variation exists in the first occurrence of traits along the vertebral series. In other words, the simplest explanation is that it is a third cervical with the illusion of a waist and an incipient "CIV morphology".

      Delete
    6. We did provide thoughts on the specimen being partial. You'll see that we provide an estimate of total length as well as measured.

      Beyond that last point, I don't see any point in discussing this further - I've said all I have to say on your ideas, all of which I consider less parsimonious than our interpretations. I think I've run my course in this discussion.

      Delete
  6. Mark K: what you are doing is called 'arguing for the sake of arguing'... You say that "predictions and counterfactuals" are required. You mean, like, additional specimens? In case it isn't obvious, they exist and are being worked on right now.

    ReplyDelete
  7. So, to recap... we have an island in which there are sauropods which are not the tallest animals around, a gigantic robust macropredatory terrestrial pterosaur, a mid sized terrestrial pterosaur, a small long necked gracile terrestrial pterosaur, a small short necked terrestrial pterosaur, a dwarf rhabdodont, a potential proportionally giant rhabdodont, a dwarf hadrosaur, a potential marine pterosaur (the pteranodont), and a potentially-herbivorous maybe-raptor (or bird). Awesome.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. Yep, except Thalassodromeus sebesensis was actually a turtle.

      Delete
    3. Turtles being effectively a boxed lunch, require opening by carriage to height and dropping, as eagles, and at least one Greek philosopher, are well aware; so who did eagles learn the trick from?

      Delete
  8. I wonder if there was some sort of niche partitioning between the azhdarchids of Hateg island. Large azhdarchids like Hatzegopteryx would take up the niche of top predator while medium sized azhdarchids like Eurazhdarcho would hunt medium-sized game. Finally, short necked forms like this new azhdarchid would take up the niche of small game hunters/ omnivores, using their short necks to pluck various fruits and seeds from fields. The neck of this azhdarchid might just be an adaptation to take advantage of the "newly-evolved" angiosperms which were thriving at the time.

    ReplyDelete
  9. The main argument of the paper, that a fragmentary isolated cervical represents a short-necked pterosaur, hinges on the correct identification of the cervical. One of the assumptions that seems to underlie the authors comparative arguments is that each cervical position is characterised by a particular morphology which, if I have understood the paper correctly, applies across azhdarchids rather than to particular taxa. There are two problems with this. First, it has yet to be demonstrated. Indeed, material I have studied (Quetzalcoatlus, Azhdarcho, Zhejiangopterus, Arambourgiania, Bakonydraco, the Solana azhdarchid) contradicts this, as upcoming papers will show. Secondly, and perhaps more importantly, the authors own conclusions, that their ‘fourth’ cervical is atypical, directly undermines the idea of morphological uniformity across taxa. Responding to a recent enquiry, I tried comparing R2395 with other azhdarchid cervicals and found that it compared best with posterior mid-series cervicals such as the seventh cervical of Azhdarcho (Averianov, 2010: fig 14) ZIN PH 138/44. The authors place some significance on the development of the neural spine which, they claim, distinguishes R2395 from seventh cervicals, but this can be highly variable within and between taxa as even the briefest of comparisons of published figures of Azhdarcho, Zhejiangopterus and Phosphatodraco will show. So, is R2395 a short-necked azhdarchid? Without a more or less complete neck in which individual cervicals can be identified with a reasonable degree of certainty, and some independent measure of size such as the humerus or dorsal-sacral series, it does not seem possible to answer this question.

    Dave Unwin

    ReplyDelete
    Replies
    1. Thanks for the comment, Dave - am pleased to have some feedback on this. Some responses to your points (over two comments - hit the character limit!):

      "material I have studied (Quetzalcoatlus, Azhdarcho, Zhejiangopterus, Arambourgiania, Bakonydraco, the Solana azhdarchid) contradicts this, as upcoming papers will show."

      You have us at a disadvantage in having had access to significant azhdarchid material now under embargo. We (again) sought access to Quetz for our recent studies on the Romanian pterosaurs, but again had no luck against the TMM lockdown. It's good to know there is new data coming on these other taxa (some of which our research team also has first hand experience with), but our conclusions are consistent with currently available data on these animals.

      "Secondly, and perhaps more importantly, the authors own conclusions, that their ‘fourth’ cervical is atypical, directly undermines the idea of morphological uniformity across taxa."

      The only manner R2395 is really atypical is the length/width ratio. Other than that, we consider it a pretty standard azhdarchid CV or CIV (see paper for discussion).

      "I tried comparing R2395 with other azhdarchid cervicals and found that it compared best with posterior mid-series cervicals such as the seventh cervical of Azhdarcho (Averianov, 2010: fig 14) ZIN PH 138/44."

      I have to disagree here. Our cervical has zygapophyseal articulatory surfaces which are much longer than wide - an opposite condition to those of Azhdarcho CVII, but consistent with a mid-series ID. The zypophyses of Azhdarchod CVII are also quite small compared to the long, relatively broad structures on R2395: these compare well with those of any mid-series azhdarchid vert. Details of the cotyle are also different, with a marked ventrolateral concavity (perhaps indicating a large hypapophysis) on Azhdarcho CVII. This is absent in R2395: the ventral margin of the cotyle is more consistent with the morphology of CIV and CVs across a number of taxa (Quetz, Azhdarcho, Phosphatodraco). It is difficult to compare neural spines of Azhdarcho CVII with R2395, but the former seems to have a wider neural spine base than R2395, which shows weakly developed, divided neural spines consistent with azhdarchid CIVs and CVs. Azhdarcho CVII also lacks the prominent, asymmetrical 'waist' common to many azhdarchid CIVs and Vs. Again, R2395 has such a waist. The only real similarity I see between Azhdarcho CVII and R2395 is the length/width ratio, but we agree with you - and note in our paper - that this can be variable between species and (probably) individuals of different ages. We therefore consider the specific anatomies of azhdarchid cervicals more useful guides to vert ID than basic proportions. With all that in mind, I don't think Azhdarcho CVII is a good match for our new vert, while the ID given in the paper remains more likely.

      Delete
    2. "The authors place some significance on the development of the neural spine which, they claim, distinguishes R2395 from seventh cervicals, but this can be highly variable within and between taxa as even the briefest of comparisons of published figures of Azhdarcho, Zhejiangopterus and Phosphatodraco will show."

      We argue contrarily here: there are indications of consistency in neural spine development in azhdarchids we are familiar with. For instance, the CIIIs of Zhejiangopterus, Quetzalcoatlus and Azhdarcho all show neural spines continuous along the neural arch, without the pronounced dip in the mid-region as characteristic of posterior cervicals. Cervicals IV, V and VI seem to consistently low spines with mid-region depressions (Arambourgiania, Quetz, Phosphatodraco, Zhejiangopterus, Azhdarcho). From CVI onwards, there are indications from at least Azhdarcho and Phosphatodraco that neural spine height and base width increases, with seemingly particularly taller spines in CVI (seen in Azhdarcho) and CVII (Phosphatodraco). More data from complete azhdarchid cervical series would be very useful here - no-one is pretending that this is a great dataset - but our interpretation of R2395 as CIV or CV is consistent with these preliminary observations.

      Also, from your comment I take it that there are better figures of Zhejiangopterus verts available in the literature. If so, please let us know where! The best photos I have are those in Chinese and Japanese volumes published in 2006 and 2007, respectively, and they still lack a lot of detail.

      "Without... some independent measure of size such as the humerus or dorsal-sacral series, it does not seem possible to answer this question."

      We acknowledge in the paper that it is difficult to get a size estimate from a single vert. However, R2395 is wider than any cervical from Zhejiangopterus, Eurazhdarcho, or the values provided for Azhdarcho neck verts in Averianov's monograph. At 46 mm wide, R2395 approaches the width of some Quetzalcoatlus CIIIs (the incomplete skeleton TMM 41544, which I recall having a 4-5 m wingspan, has a CIII 51 mm across). This suggests R2395 was not tiny, and likely occupies a size between Zhej, Eurazhdarcho etc. and Quetz (hence the 3-4 m wingspan suggested in our paper). Given that our CIII-CVII length estimates (computed from Zhejiangopterus, Phosphatodraco and Azhdarcho) are less than the measured neck lengths of Zhejiangopterus, the suggestion of a proportionally short, robust neck in this azhdarchid does not seem unreasonable.

      Delete
  10. Hi Mark,

    I must admit that it is very difficult to evaluate fine anatomical details on the basis of illustrations of incomplete specimens where even slight differences in orientation can significantly alter our perception of shapes and proportions, but I’m still not fully convinced that R2395 is a fourth cervical. For example, it looks as if the midline height of the cotyle of R2395 is much lower than the height of the neural arch, a relationship that, according to Averianov (2010, figure 8), is typical of late mid-series cervicals such as the seventh or eighth where the cotyle is a low, transversely elongate structure. If I have interpreted this aspect of R2395 correctly (and have not been misled by the orientation, shadowing, quality of the illustration) then the basic configuration in this vertebra is quite unlike the condition in early mid-series cervicals where the cotyle is relatively deep (nearly as tall as it is broad) and exceeds the height of the neural arch (op cit). In this respect R2395 doesn’t seem to look at all like a standard fourth or fifth cervical, whatever they might be. So, I’m not at all sure whether R2395 is a fourth, sixth or seventh cervical, mainly because of the lack of consistency in anatomical features between taxa which confounds our attempts to identify isolated vertebrae. That there is disagreement regarding the much more complete, articulated series of cervicals of Phosphatodraco, further emphasises this problem. So, I’ll stand by my point that we should be very cautious when dealing with isolated vertebrae, partly because we can already demonstrate that what we think are homologous cervicals can be quite dissimilar from one another, and partly because we still lack a well founded and widely agreed map of the azhdarchid neck with which we can make comparisons.

    ReplyDelete
    Replies
    1. "For example, it looks as if the midline height of the cotyle of R2395 is much lower than the height of the neural arch, a relationship that, according to Averianov (2010, figure 8), is typical of late mid-series cervicals such as the seventh or eighth... If I have interpreted this aspect of R2395 correctly (and have not been misled by the orientation, shadowing, quality of the illustration) then the basic configuration in this vertebra is quite unlike the condition in early mid-series cervicals where the cotyle is relatively deep"

      Except that, as we note in the description (and indicate in the illustrated line drawings), the cotyle is imperfectly preserved and been subject to crushing, whereas the neural arch appears to have escaped this. The bone surface of the cotyle is quite irregular and fractured, suggesting it might be quite significantly squashed along with the rest of the ventral surface of R2395. If I recall correctly, similar differentiated deformation has taken place on some Q. sp. cervicals - I suspect the complex structure and shape of the neural arch makes it more resilient to crushing than the cotyle. While I agree that a short cotyle would conflict with our identification, it seems unwise to base vertebral identification on this distorted feature when there are much clearer, better preserved, features at hand.

      "That there is disagreement regarding the much more complete, articulated series of cervicals of Phosphatodraco, further emphasises this problem"

      I'm not sure this disagreement really affects our interpretation that much: the Phosphatodraco holotype doesn't really provide than much information on complex vertebral anatomy, and the general trends it suggests (increasing shortness and complexity posteriorly; longer, simpler and more tubular morphologies in the mid-series) are consistent in either interpretation. As an aside, for this paper, we're following the advice of one of our team (GD) who examined Phosphatodraco first hand, agreeing with the interpretation given by Kellner that it represents CIII - CVIII. Some corroborating evidence for this comes from the vertebra length metrics, which seem to follow the 'standard' cervical length patterns.

      Delete