Changes through growth in sauropods and ornithopods

March 29, 2012

By one of those happy coincidences that you sometimes get, today saw the publication of not one but two dinosaur ontogeny papers: this morning I was sent a copy of Woodruff and Fowler (2012) on ontogenetic changes in the bifid spines of diplodocoids, and tonight I was alerted to Werning (2012) on Tenontosaurus growth trajectories based on osteohistology.

It’s interesting to compare them.  The obvious conclusion is that, while sauropod vertebrae are intrinsically better than ornithopod long-bones, the latter make much, much better subjects for ontogeny studies.

Sauropod Vertebrae

Woodruff and Fowler (2012:fig. 3). Ontogenetic development of diplodocid anterior cervical vertebrae documenting bifurcation from absent, through incipient, to fully developed. (A) MOR 790 7-30-96-132, Diplodocus sp. (B) MOR 790 8-10-96-204, Diplodocus sp. (C) MOR 592, Diplodocidae sp. (D) ANS 21122, Suuwassea, (image provided by J. Harris). (E) CM 555, Apatosaurus excelsus, (image provided by M. Wedel), (F) CM 84 (from Hatcher, 1901). Scale bar = 5 cm.

The problem that Woodruff and Fowler have is that they’re working from a small selection of mostly isolated elements, nearly all of them damaged by breakage and/or crushing, and uncontrolled for serial position beyond some basic binning.  As a result, the taxonomic identifications are really rather arbitrary and unsupported — as they say (p. 2), “for the purposes of this study, original taxonomic designations were not reexamined” — and indeed re-examination would not necessarily help much.

As a result, they are left with rather a circular argument, as follows:

  • Small specimen X is Suuwassea.
  • Small specimen Y is assigned to Apatosaurus, because someone once said so.
  • But Y has a less bifid neural spine than adult Apatosaurus.
  • So spine bifurcation increases through ontogeny in Apatosaurus.
  • So small specimen X belongs to Apatosaurus, too.
  • => Suuwassea is Apatosaurus [Note: I stupidly wrote “Diplodocus” here in the first posted version.  Now corrected to Apatosaurus.]

I don’t find this at all convincing.  (Neither does Matt: we discussed this briefly today, and at more length at the Bonn workshop where Cary presented this work.)  Leaving aside the observation that the conclusion fits in neatly with the Horner Lab’s ongoing everything-is-just-a-Triceratops-growth-stage project, there isn’t really much that Cary could have done differently here: the necessary specimens (i.e. multiple near-complete associated individuals of unambiguous taxonomic identity) just don’t exist.

(Mind you, figure 7A is about the least convincing evidence of bifurcation I’ve ever seen.)

Ornithopod long bones

Werning (2012:fig. 2). A. Cross-section of OMNH 10144. This bone was invaded by bacteria before fossilization and thus much of the primary tissue is obscured. It is presented here in cross-section to illustrate vascular density and arrangement. B. Detail of A, showing general vascular patterning. The cortex is dominated by longitudinal canals arranged circumferentially. C. Cross-section of OMNH 8137. D. Detail of C, showing primary cortical tissues. The bone is woven, and most canals are longitudinal primary osteons (some anastomose circumferentially). Two LAGs (arrows) are shown. E. Cross-section of FMNH PR2261. F. Detail of E, showing mostly primary tissues of the midcortex at a transition to slower growth. Deeper in the cortex (upper left), bone is woven and osteocytes are dense and disorganized. Some secondary osteons are visible, but they do not overlap or obscure all of the primary tissues. Past the LAG (arrow), canals remain dense but decrease in diameter, bone tissue is weakly woven, and osteocytes decrease in number and become more organized. Scale bars: A = 2 mm; B, F = 1 mm; C = 4 mm; D = 0.5 mm; E = 10 mm.

By contrast, Sarah Werning has a much better story to tell, largely just because she’s working with much better material.  Tenontosaurus is known from many complete and near-complete individuals, there is no real uncertainty about the material all belonging to the same taxon, and the elements (long bones) are much less subject to crushing and breakage than vertebrae.

Of course it also helps that Sarah has done astonishingly careful, detailed work.  I don’t think I’m giving much away when I say that this paper has been many years in the making — as the Acknowledgements say “This work was completed in partial fulfillment of the requirements for the Master of Science degree, Department of Zoology, University of Oklahoma.”  And it’s been a looong time since Sarah was at Oklahoma.  But I get the sense that this reflects taking the time and putting in the effort to Get It Right, rather than standard-issue procrastination.

Perhaps the most impressive part of this paper, though, is methodological:

High-resolution histological images of the cross-sections are digitally reposited online for scholarly use at MorphoBank (http://MorphoBank.org), project p494; see Table 2 for a list of slides and accession numbers. Digital images larger than 25,000 pixels in either dimension were digitally scaled (reduced to 17,000–20,000 pixels in the larger dimension) to allow processing on MorphoBank and because most image editing software does not support editing of gigapixel jpg files. These edits were made after scale bars had been added. Images in full resolution can be obtained from the author.

As you can see, these are ridiculously high-resolution images.  By making them freely available, Sarah is doing everything she can to enable others to replicate or correct her work — or to use the data for other purposes that she’s not yet thought of.  This is a big win for the progress of science, and I want to publicly congratulate Sarah for doing it.  I would love to see this become standard behaviour.

(It’s a bit strange that the paper links onto the the MorphoBank home page rather than directly to the Tenontosaurus project.  Here you go.)

PLoS ONE vs. Journal of Morphology

A final thought, before I finally go to bed.  When I saw Cary’s paper in the Journal of Morphology I thought to myself, “good for him, he’s got that into a really good journal” — which is true.  But then when Sarah’s came in, I found myself comparing PLoS ONE.  Here’s what I came up with:

Access.  A no-brainer: PLoS ONE wins hands down, being open-access while J. Morph. is paywalled.

Charges.  The flip side of the first category: J. Morph. wins because (as far as I can tell from the Author Guidelines) there is no publication charge.  PLoS ONE charges $1350.

Length.  Another PLoS ONE win, because it imposes no limits whatsoever on length, figure count, etc.  Cary’s paper is no lightweight at 11 pages, but PLoS’s liberality with page-space means that Sarah’s is well over twice as long at 25 pages.  Now of course not all papers need to be long; but for those that do, cutting to a journal’s length limit is a painful and stupid process.

Image Colour.  Again, PLoS ONE wins — it’s very rare these days to see a greyscale specimen image in a PLoS journal.  But to be fair to J. Morph., the author guidelines do say: “All color figures will be reproduced in full color in the online edition of the journal at no cost to authors. For the printed version free color figures are at the editors discretion.”  So it looks like Cary and Denver dropped the ball on this.

Image Resolution.  No question.  PLoS ONE wins by a mile.  Click through the two representative images above: the PLoS one has 3.5 times as many pixels, and that’s after I added a 10% wide margin around the J. Morph. image.  And if you want to use these illustrations as basis for new images, remember PLoS also lets you download the original full-resolution submitted image in a lossless format.

DOI Resolution.  When I added the references below, I noticed that the DOI for the J. Morph. article doesn’t resolve yet, which is careless.  PLoS ONE wins in this category, resolving just fine.

Impact Factor.  Yes, it’s a stupid number and we all hate it.  But some people still seem to take it seriously, so we may as well look.  And PLoS ONE wins with 4.411 to J. Morph.’s 2.087 — more than double.

Putting it all together, based on the seven categories that I evaluated here (and no doubt I missed some), it looks like the only reason to go with J. Morph. ahead of PLoS ONE is to avoid the publication fee.  By this point, PLoS ONE has made itself an absolutely mainstream journal for palaeo, and the obvious first choice for most projects.

References

  • Werning, Sarah.  2012.  The ontogenetic osteohistology of Tenontosaurus tilletti. PLoS ONE 7(3):e33539. doi:10.1371/journal.pone.0033539
  • Woodruff, D. Cary, and Denver W. Fowler.  2012.  Ontogenetic influence on neural spine bifurcation in Diplodocoidea (Dinosauria: Sauropoda): a critical phylogenetic character.  Journal of Morphology, online ahead of print.  doi:10.1002/jmor.20021 [Direct Link, since the DOI doesn’t seem to work.]
About these ads

82 Responses to “Changes through growth in sauropods and ornithopods”


  1. Mike, nice summation, but:

    there isn’t really much the Cary could have done differently here: the necessary specimens (i.e. multiple near-complete associated individuals of unambiguous taxonomic identity).

    You’ve left us hanging here with the clarification of your colon!

    Mind you, figure 7A is about the least convincing evidence of bifurcation I’ve ever seen.)

    Sadly, I lack the reference to evaluate this statement: perhaps you could include the figure or discuss the reasoning? (Not that i doubt you.)

  2. Mark Robinson Says:

    Minor quibble, in the last para under References I believe it should be “Cary” without an ‘e’. Nice to occasionally see you write an article about something as silly as bones that aren’t sauropod vertebrae.

  3. Mike Taylor Says:

    Thanks, Jaime and Mark, for spotting those mistakes — now both fixed. As you can guess, I wrote this in a hurry while desperately wanting to get to bed at 2am :-)

    In the references, I was obviously thinking about the Joni Mitchell song.

    I’ll add Fig. 7A later.


  4. Your opinion does not have to agree with the data that we present, but if so, I invite you to falsify our hypothesis. Also, please read the paper thoroughly and be careful not to misrepresent the material. It is untrue to suggest that the material is “nearly all … damaged by breakage and/or crushing, and uncontrolled for serial position beyond some basic binning”. MOR 592 is a partly articulated (and near complete) dorsal to cervical series (plus partial skull). The Mother’s Day site is full of abundant quality material (multiple individuals; so yes, a bit jumbled). It would be nice to have a histological test of the hypothesis (ad this is underway, where possible).


  5. I’m also curious about this:

    “As a result, they are left with rather a circular argument, as follows:

    Small specimen X is Suuwassea.
    Small specimen Y is assigned to Apatosaurus, because someone once said so.
    But Y has a less bifid neural spine than adult Apatosaurus.
    So spine bifurcation increases through ontogeny in Apatosaurus.
    So small specimen X belongs to Apatosaurus, too.
    => Suuwassea is Diplodocus.

    I don’t find this at all convincing.”

    Where do we say that “small specimen x is Suuwassea”?
    Where do we assign the Suuwassea holotype to either (specifically) Apatosaurus or Diplodocus.

    Or maybe you didn’t read the paper properly. As far as I can see, we just say that Suuwassea has features simialr to all other immature diplodocids, and that we can’t tell which taxon it belongs to because we don’t have enough intermediates linking medium-sized to adult Diplodocus/Apatosaurus morphologies (whatever “adult” means in this context).

  6. Mike Taylor Says:

    Here’s that figure 7 that I mentioned in a drive-by:

    Woodruff and Fowler (2012:fig. 7)

    The caption is:

    Fig. 7. Ontogenetic development of diplodocid caudal vertebrae documenting incipient to fully developed bifurcation. (A) MOR 592, Diplodocidae sp. (B) CM 84, D. carnegii (from Hatcher, 1901). (Inset) apex of MOR 592 neural spine in lateral view. Scale bar = 5 cm.

    And the relevant portion of the text (from pages 4-5) is:

    Caudal vertebrae. Some anteriormost caudal vertebrae also develop bifurcation (Fig. 7). In MOR 592 (vertebral arch 34.5 cm high), the anterior caudal vertebra exhibits a tapering neural spine with a widened, nearly ovoid-shaped apex and shallow transverse trough. In the adult form (CM 84, vertebral arch, 104.9 cm high; Hatcher, 1901), the neural spine is a tall “paddle”-shape, with a shallow “V”-shaped bifurcation at the apex.

    Note that the leftmost part of fig. 7A is in lateral view.

  7. Mike Taylor Says:

    Hi, Denver, thanks for joining us.

    Where do we say that “small specimen x is Suuwassea”

    That part of the argument is not in doubt — the relevant elements are part of the holotype, aren’t they?

    Where do we assign the Suuwassea holotype to either (specifically) Apatosaurus or Diplodocus.

    From the abstract: “the small diplodocid Suuwassea is more parsimoniously interpreted as an immature specimen of an already recognized diplodocid taxon.” And from page 8, “a more parsimonious explanation is that the Suuwassea holotype specimen ANS 21122 represents an immature ontogenetic morph of an already recognized diplodocid taxon”. So you don’t explicitly synonymise it with another specific taxon, but you make it pretty clear where you’re headed.

    More importantly:

    Or maybe you didn’t read the paper properly.

    Correct (so far) — I’ve only skimmed. This post is based more on Cary’s presentation of the study in Bonn. That’s why I didn’t attempt a point-by-point response, just put a “we don’t find this convincing” stick in the ground.

  8. Cary Woodruff Says:

    I’ve dreamed of the day my work would make the front page of SVPOW! Seriously Mike, thanks for considering our research important enough to warrant further sauropod discussion. Two things I’d like to say first: 1) I will fully admit that I’ve not had a lot of publication related experience, so I was not really aware of the whole open access issue (but thanks to your series of in depth posts I am now!), 2) this is still an early online version and I am in discussion with J. Morph. to correct some errors that were still not fixed after the proofs stage (don’t read the paper and assume I’m completely illiterate!). Now on to science! Frankly I’m glad to hear doubters. Ontogeny in all of its forms can a difficult concept to wrap your brain around. Most folks are most familiar with the ongoing Triceratops vs. Torosaurus debate, and I’ll be the first to admit our work is nowhere near as sexy. But I will firmly stand by our most basic result, that bifurcation does develop ontogenetically. Listen up doubters, here is the foolproof way to prove us wrong: find a small diplodocid will fully bifurcated neural spines. As yet in every small diplodocid or Camarasaurus our work has held true. Now I’m in the boat that believes that Toni (SMA 0009) is a diplodocid, but I know opposing viewpoints do exist. Regardless of Toni’s phylogenetic position, our work still holds true. So is it more logical to say that since we know dinosaurs underwent radical ontogeny, that finding ontogenetic change in sauropods is not really that surprising, or is more logical to say that all of these specimens should in fact represent numerous Morrision dwarf taxa? I seem to recall that Matt was the one to propose that pneumaticity is ontogenetic as well. So some ontogeny is ok, but not others? Personally I’m going with ontogeny. And if someone thinks that small sauropods (ok bigger than Toni) are rare, then they certainly haven’t spent much time in museum collections. So to be fair a lot of our dataset at that point came from the Mother’s Day Quarry. This is a jumble of many sauropods, and almost none of the MOR’s material had any articulation (same for the Kenton material). So you’re completely right that for the MDQ material it is out of place and identifications were rudimentary. However nearly every other specimen is known from a decent vertebral series. But while in Switzerland I spent a week with Kirby Siber’s diplodocids, and they are immature animals that are similar sized and articulated. So now I can say what position an isolated vert is. And figure 7A. I’m glad Mike that you said, “…about the least convincing evidence of bifurcation I’ve ever seen.” And I do mean this with the utmost respect to sauropod workers, but I’d say at least 90% of folks have no idea what bifurcation really is (hint hint for a upcoming paper). So where does that leave us? The future! So all of this work was basically my undergrad thesis and as Denver likes to say we wanted to shake the bottle. For my Master’s I’m diving in head strong. I’ve already built the dataset up loads, I’m re-evaluating the taxonomy (which to be fair thus far has been nearly spot on), and I’m doing the histology for every specimen. Histology is the final answer. That’s what we say for Suuwassea. And a sneak peak, but thus far the histology is also backing us up. But we really cannot talk about the future because it isn’t here now. Honestly I thought that more folks would be commenting about Suuwassea or what we said about phylogeny. Maybe that’s to come on the DML. But thanks for the opportunity to discuss our work, and hopefully continue to do so!


  9. Ok I didn’t follow the x-y thing properly. We merely point out that all small to medium body size (Morrison) diplodocids (also, tentatively Camarasaurids) show incipient/less developed neural spine birfucration than the larger forms (where body size is assumed to be at least roughly correlative with ontogeny; histological corroboration is underway). We suggest Suuwassea belongs to a known diplodicid (perhpas Diplodocus or Apatosaurus) but we cannot tell the difference because we don;t know the difference between an immature diplodcus and immature Apatosaurus,

    The hypothesis is that either (and we consider the merits of both):

    1. Every different bodysize that has a different morphology is a different species
    2. We have one or more growth series

    We suspect that H2 is more parsimonious as it makes less species and we don’t get left with the curious situation where each species is only known from the same growth stage. What are the chances that al 1/3 grown diplodocids that we find are one species, all 2/3 grown are another? Isn;t it more likely that they are growth stages of the same species.

    Do you think that growth is isometric: that a juvenile diplodocus should look identical to the adult form (just smaller), or would it change through growth? What would you predict a juvie Diplodocus would look like, and how would it be different from Apatosaurus. Better, how would you test these ideas? What would you look for in the fossil record? What metadata (ontogeny and stratigraphy) would you use to support this view?

    We make a stab at this in the paper, which isn’t supposed to be a final answer. Rather, we take one character and show that it varies consistently (and without exception) with body size, which looks like a strong case for ontogeny (if you disagree, falsify the hypothesis: it is a falsifiable hypothesis, unlike “diversity first”). It is likely that many other characters (if not the majority of important characters) will have ontogenetic variation. Some will be obvious, others less so, and there is a stratigraphic component too (7 million years in the Morrison; plenty of time for a bit of change).

  10. Mike Taylor Says:

    Thanks, Cary and Denver, for your thoughtful comments. They deserve a proper reply, but since I don’t have time to give you one now, I’m leaving this placeholder here. I’ll return when I can. It’s a shame, in a way, that your and Sarah’s papers came out on the same day, because they were each worthy of discussion in isolation.

    One thing: to mitigate the low resolution and grey-scaling of the figures in the published PDF, have you considered posting a set of full-sized submission-scale images (in colour where you have it) like the ones we did for our neck-posture paper? These days it would probably make more sense to put them on FigShare or similar rather than on a bespoke site as I did back then.

  11. Andy Farke Says:

    With all due respect to Cary and Denver, it is not a simple either/or case. Due to the vagaries of the fossil record, sometimes we’re going to have only a juvenile specimen of a rare taxon. I would expand the hypothesis list to:

    1. Every different body size that has a different morphology is a different species. (a ton of species)
    2. Some different body sizes with different morphologies are different species, but not all. (a moderate number of species)
    3. We have one or more growth series (one or two species)

    #1 is a straw man hypothesis – with few exceptions, nobody has seriously followed this method of estimating dinosaur diversity for decades. Sure, we ended up with things like Morosaurus and Procheneosaurus, but all of those were mopped up some time ago. Even Gilmore recognized juvenile sauropods, way back in the 1920s. In actuality, the situation today is choosing between #2 and #3.

    The hypothesis (apologize that I’m straw manning here a little) that most things are juveniles of one or two other species also makes some ecological and taphonomic assumptions that aren’t warranted. For instance, due to ecological partitioning it’s entirely possible that we might only find juvenile specimens for some taxa in a given facies or formation. Some great examples are the giant hadrosaur bonebeds up on the North Slope – almost exclusively (if not exclusively) juveniles. Fruitadens is also known only from juvenile material. This doesn’t mean that it’s the young form of Dryosaurus.

    Hypothesis #3 also makes some unwarranted evolutionary assumptions. In many cases, the juvenile features of one taxon are the adult features of another (no surprise to anyone reading this). As an independent example, the supraoccipital participates in the boundaries of the foramen magnum in some “primitive” ceratopsians. Juveniles of “derived” ceratopsians like Triceratops have that same morphology, whereas adult Triceratops exclude the supraoccipital from the foramen magnum. Would you then suggest that Protoceratops has to be a juvenile ceratopsid? I don’t think anyone would – similarly, if the neural spines bifurcate during the ontogeny of some sauropods, this doesn’t automatically mean that it happened in all sauropods, or that a sauropod with an unbifurcated neural spine is automatically a juvenile of another taxon.

    And here is the part where we’re probably all in agreement – independent measures of age, like histology – are needed to resolve the issue. But. . .for the aforementioned reasons (ecological partitioning, poor sampling, whatever), some taxa will only ever have juvenile specimens. For this reason, histology (even in concert with stratigraphy and other data) is very important, but it is *not* a magic bullet.


  12. Falsify the hypothesis. Talk of bias in the fossil record does not provide testability: this is “diversity first”. We find large-bodied sauropods in the same Morrison beds as small-bodied. Does this mean that they are the same species? No, but it at least removes one of the potential taphonomic biases (body size).

    Show me a small-bodied diplodocid or camarasaur that has bifurcation that is equally developed as in adults. That will falsify the hypothesis. One specimen, out of the hundreds of small-sized vertebrae available (that Cary has looked at), from 7 million years of the Morrison.

    And anyway, we are merely putting the idea out there that bifurcation might be ontogenetic. A currently supported and unfalsified hypothesis that is testable. I’m happy to have it sit alongside the diversity first hypothesis. Test them.

    “In many cases, the juvenile features of one taxon are the adult features of another”

    Right. It’s possible that Suuwassea is some weird heterochronic
    species. It’s also possible that its a juvenile of a recognized taxon. What would a Suuwassea-sized juvie of Diplodocus or Apatosaurus look like?

  13. Andy Farke Says:

    1) Recognition of bias in the fossil record *is* important and *is* testable. Perhaps the most relevant one is the recent paper in PLoS ONE by Horner et al. on the predominance of juvenile material in the Hell Creek Formation. You can have examples of large and small animals in the same beds, and still have bias (collection or pre-collection taphonomic in nature). As you know from your work in the Hell Creek, it is only through extensive collection of various facies that this bias can be identified.

    2) I do not disagree (at least based on the evidence I’ve seen so far – Mike & Matt may know something I don’t) that bifurcation developed ontogenetically in some species. This doesn’t mean it developed ontogenetically in *all* species. That is all I’m saying. We have lots of examples today where you get taxa with “derived” and “basal” characteristics side-by-side. To pick an example from my experience, frontal sinuses develop ontogenetically in some bovids. Yet if you go to Africa, you can find adult individuals of taxa with big sinuses living next to adult individuals of taxa with small sinuses. For instance, impala (decent-sized sinuses) and waterbuck (tiny sinuses). Yes, sinuses develop ontogenetically, but finding a skull with small sinuses is not enough to say that it came from a juvenile impala. It could just as easily be from an adult waterbuck! The same certainly would have happened if you had multiple sauropods from disparate clades living together.

    3) As for what a juvenile of Diplodocus or Apatosaurus looked like. . .I can’t speak for Apatosaurus off the top of my head, but according to Whitlock and colleagues (paywalled – free PDF here), the skulls of juvenile Diplodocus are easily distinguished from Suuwassea.

    An alternative here is that Suuwassea *is* a juvenile of a valid taxon distinct from those previously known.

  14. Mike Taylor Says:

    Andy has said much of what I would have said — nice work! :-)

    So to clarify, I certainly don’t disagree with Cary and Denver’s most fundamental finding that bifurcation is generally more complete in adults than in juveniles, or with their first inference that this is because bifurcation develops ontogenetically in at least some taxa.

    The problems come further downstream, when specimen growth-stage, serial position and taxonomic identity (through association) are unknown. Histological work to determine growth stage will help with two of these issues; the other two will need more and better specimens.

  15. Andrea Cau Says:

    In absence of histological works, could we use other indications of maturity? Wedel (2003, 2009) showed that tail pneumatisation occurred “late” during sauropod ontogeny: is it correlated in some way with caudal neural spine bifurcation? Which started first? Are there studies describing/comparing the degrees of both vertebral pneumatisation and neural spine bifurcation in diplodocid tails?

  16. Cary Woodruff Says:

    Andrea,
    We are in the process of CT scanning cervical and dorsal vertebrae to truly document development of the pneumatic structures through a growth series. I think Matt was dead on with his theory of ontogeny linked pneumatization (at least externally it is dead obvious!). So who knows, will one vertebral feature proceed another, or vice versa, or what if they occur simultaneously? Since pneumatization in diplodocids is rather dull (compared to a cervical), and coupled with the rather late development of bifurcation, I would theorize that in regards to biomechanics and ontogeny that the neck and torso would have loads more information. And to address your first question, histology must be done! Something may look immature or mature, but how do you know? Suuwassea bears numerous features indicative of an immature animal; so is it the case of an animal with juvenile-like characters, or is it legitimately immature? Histology is the only foolproof way to answer that. Coupled with the numerous juvenile characters, we state that pending a histological analysis that Suuwassea may be a juvenile diplodocid. We are working on the histology now, but we would like to get to the point that someday based on a number of characters we could be more definitive in asserting relative age.

  17. Andrea Cau Says:

    Thank you very much, Cary: I’m currently studying a specimen, but actually cannot do histological analysis (and probably I’ll never be allowed to do it), so I have to infer at least a rough ontogenetic stage based on external features: ossification degree and pneumatic invasion would be the best available.
    I provisionally consider the combination of a complete sinsacral fusion (all sacral centra, arches and spines are fused) and presence of caudal pleurocoels as evidence that the specimen was not juvenile. This is, for me, a conservative but not too much unrealistic, interpretation. Any suggestion would help much.


  18. I agree that it’s possible that you have a species that might only show incipient bifurcation as a final morphology (although I would expect this also to develop ontogenetically), but since Suuwassea shows other features indicative of immaturity, it’s special pleading that this taxon is in some way different to all other species that exhibit bifurcation. On it’s taxonomy, we don’t provide a specific conclusion, we just say that we need to be careful about taxonomic and phylogenetic conclusions drawn on immature specimens (the main point of the paper).

    On the preservation bias issue: I have presented at SVP on bias in the Wealden, and there are places where it can occur, but the Morrison is replete with small and large bodied specimens from the same horizons. You must keep bias in mind in all studies, but it is hardly parsimonious (especially in the case of the Morrison) to argue that the juvenile dinosaurs we find are likely to belong to a species for which we have not yet found the adult (given the abundance of adult, or at least diagnosable large-bodied, material).

    We’re trying to open up Morrison sauropod taxonomy to the sort of stratigraphic and ontogenetic scrutiny that is currently taking place for dinosaurs from the Campanian-Maastrichtian Western Interior. I fully expect there to be ontogenetic and stratigraphic components to morphological variation, which may lead to some taxa being sunk (ontogomorphs), but this same approach leads to new taxa being erected (stratomorphs/chronospecies). I get pretty tired of the snarky comments that we’re all “hornerites”: a league of rabid indiscriminate sinkers with a secret hit list. Far from it.

  19. Matt Wedel Says:

    I agree that it’s possible that you have a species that might only show incipient bifurcation as a final morphology

    Uh, yeah, it’s certainly possible, as evidenced by Apatosaurus excelsus (Gilmore 1936) and Diplodocus carnegie (Hatcher 1901), both of which show incipient or incomplete bifurcation in the anterior cervicals. So in those ostensibly adult animals one finds unsplit spines, incipiently split spines, and fully split spines, and all three ontogenetic stages are coexisting serially in one individual. Same goes for CM 555, which is about half the size of CM 563 CM 3018. You guys figured the one neural spine of CM 555 that is incipiently split, but not the completely unsplit spines anterior to it or the fully split spines posterior to it. So, once again, one animal is showing the full range of conditions within just a few serial positions. This is the point that Mike made in the post: we agree that spine splitting changes through ontogeny, but it also changes serially, and you can’t tell which you’re looking at unless you have a reasonably complete presacral series. In CM 555, the most anterior vertebra that is more than incipiently split is C7, same as in CM 563, so from those two specimens one would infer that neural spine bifurcation did not increase in at least the latter half of ontogeny in Apatosaurus.


  20. And we state in our discussion that bifurcation does not proceed at the same rate serially:

    “Bifurcation first occurs in the posterior cervicals and anterior dorsals, although confirmation of this trend requires
    collection and analysis of more complete specimens”

    And we even divide the cervicals into anterior and posterior sections to accommodate the serial variation. Interestingly, this “works” phylogenetically, since the non-neosauropod Mamenchisaurus only shows bifurcation (and very shallow bifurcation at that) in the posterior cervicals and anterior dorsals.

    Thus, yes, you get differently expressed depth/width of bifurcation depending on serial position. We see this in MOR592 which is 4 dorsals and 11 cervicals (pretty decent completeness).

    I don’t see how this falsifies ontogenetic development in bifurcation. What’s your argument? Are you saying that all the incipient bifurcation that we describe are anterior cervicals of adult that are misidentified? the (near) completeness of MOR592 falsifies this argument. The cervicals that are most bifurcated in adults are also the most bifurcated in MOR592, just at a lesser state of development; as is the whole point of our paper.

    Anteriormost cervicals (as evidenced by necks of adult diplodocids) are less bifurcated than posterior cervicals, but we say this in the paper.

  21. Matt Wedel Says:

    Yes, and you also say that Haplocanthosaurus and Barosaurus might be juveniles of Apatosaurus and Diplodocus. Shall I falsify the hypothesis that Haplocanthosaurus is a juvenile Diplodocus? Go read Hatcher (1901, 1903). Better still, go to the Carnegie and see the material for yourself, and compare it to the juvenile Diplodocus material in the same collection. A list of characters as long as my arm separates Haplocanthosaurus from Diplodocus, and you’re proposing synonymy on the basis of one character–bifid spines–that is absent in Haplocanthosaurus, when it is already present and extensive in similar-sized juveniles of Apatosaurus and Diplodocus. There, hypothesis falsified.

    As for Barosaurus, there are juveniles (at the AMNH and BYU) and ‘normal-sized’ adults (at the AMNH, Carnegie) and some monsters with cervicals a meter long (at BYU). The morphology is consistent among all of them. They aren’t Apatosaurus, they aren’t Diplodocus.

    You asked in a previous comment:

    What would a Suuwassea-sized juvie of Diplodocus or Apatosaurus look like?

    We don’t have to speculate, we can just go look, because those specimens exist. Primarily at the Carnegie and BYU. They don’t look like Suuwassea, they don’t look like Haplocanthosaurus, and they sure as heck don’t look like Barosaurus.

    You propose synonymies based on a single character while ignoring numerous contradictory characters, ignore known juvenile specimens of Apatosaurus and Diplodocus (and Barosaurus), and then complain about being tarred as a Hornerite. Well, do your damn homework. The age of single-character phylogenetics ended some time ago.

  22. Mike Taylor Says:

    (Side-note: I just noticed that the Journal of Morphology DOI now works.)

  23. Cary Woodruff Says:

    Matt,
    1. I could agree that Haplocanthosaurus is not a Diplodocus. But you’ll have to agree with me that the material is quite small. Now size should not be immediately indicative of maturity, but when I examined H. priscus and H. utterbacki at the Carnegie in December, those specimens scream ontogeny. In the case of those particular specimens I saw nothing at the time that could not falsify that H. utterbacki was a juvenile H. priscus. Also there is no neuro-central fusion in H. utterbacki and only incipient bifurcation, contray to H. priscus. Haplocanthosaurus is lower Morrison so potentially it could be distinct from the upper diplodocids.

    2. As far as Barosaurus and Diplodocus I’m a bit more hesitant to say they are not the same animal. Again I’ve looked at both of the beautiful baby Barosaurus at the AMNH, and they look identical to known juvenile Diplodocus. Now is the similarity between Barosaurus and Diplodocus because the split between the two happened sometime during the lower Morrison? The Morrison stratigraphy is screwed up, but do you know if all “Barosaurus” are restricted to the lower or upper Morrison, or continuous throughout? Once the stratigraphic resolution is cleared up we can decipher this. Maybe we are dealing with anagenesis among many Morrison taxa. And why is Barosaurus so rare? Apart from some cervical lengths even the AMNH specimen really isn’t that big (smaller than CM 84). And what makes the two so different from one another? Is this like Triceratops v. Torosaurus? That is that “Torosaurus” specimens are really rare while Triceratops are ubber common. Matt I can certainly agree that the Morrison taxa should not be synonymized. God knows we all despised it when Henry tried to do it with “A. brontodiplodocus”. This particular passage in the paper was meant to be more thought provoking (and I can tell it worked!). If Seismosaurus hallorum can now be recognized as just a large Diplodocus, why can this not go the other way for smaller taxa? Yes I don’t think that all encompassing synonymization should be applied, but to not question the possibility of ontogeny is ignorant. The plain fact is that sauropod ontogeny is hard as hell. Ask Martin Sander. Due to fast growth rates LAGs are nearly always obliterated, and they lack horns, crests, or frills which are relatively easy features to recognize and track through ontogeny. I’m not saying the recognition of amount of bifurcation is perfect, but it is a start.

    3. Not to be argumentative but when we asked the question of “What would a Suuwassea-sized juvie of Diplodocus or Apatosaurus look like?”, we already knew this answer. And you are spot on Matt with “We don’t have to speculate, we can just go look, because those specimens exist.”. Right. And that’s exactly what I did. And it is of my utmost professional opinion that at that particular size range Suuwassea and MOR 592 look identical to other known juvenile Diplodocus and Apatosaurus. Heck in 1999 Jeff Wilson said that MOR 592 was a long lost Amphicoelias! Now to be fair upon closer inspection there are differences between juvenile Diplodocus and Apatosaurus, but those same differences are expressed likewise between Suuwassea and MOR 592. One is an Apatosaurus and the other a Diplodocus. But as we said in the paper we are going to wait until a formal histological analysis until a proper sinking. And I know John Whitlock thinks Suuwassea is some kind of lost dicraeosaurid based on the dentary, but go out and look at the dentary of Suuwassea and the juvenile Diplodocus John described. Both look rather dicraeosaurid. And why not! Juveniles express characters more common in the basal form than their own derived form. And John beautifully shows how the dentary changes shaped through ontogeny. So not to be argumentative, but the dentary and pneumatization can instantaneously be recognized as ontogenetic, but vertebral change is harder to swallow?

    4. Now we’re talking. Many phylogenetic characters are onotgentic characters. Fact. Just go and ask Thomas Carr. We even state in the paper that synonymization based on one character is sketchy, but bifurcation is a freaking huge character in Diplodocoidea phylogeny. What separates Rebbachisauridae from the other two? Some cranial features and lack of bifurcation. Now known rebbachisaurids are being found with definitive bifurcation that even Ray Charles could see. Even in these specimens bifurcation appears to be ontogenetic. So how does Rebbachisauridae stand? My bet is that soon other workers will sink Rebbachisauridae. And originally for the paper we did have a very small, very basic Carr inspired “ontogram” which placed our smallest form, then the middle, and finally CM 84, just as we expected. Now we are going back and doing a proper kick ass one. But as Thomas has shown, at relatively small growth stages, even a very small number of characters may pull an individual away. But through ontogeny the presence, absence, or even further character combinations change. And we did our damn homework Matt. We recognized the importance and potential significance of a character, not write it off or ignore it as everyone else did before.

  24. Matt Wedel Says:

    Cary, thanks for the substantive reply.

    I agree that the Carnegie Haplocanthosaurus material is small and probably subadult, and that the two species there are probably ontogenetic morphs. There’s nothing new there, people have known that for a century. As for the possibility that H. is a juvenile of Apatosaurus or Diplodocus, I’ll just point out that both of the latter taxa have 10 dorsals and H. has 12 or 13. And its dorsals have upwardly directed transverse processes and short, wide neural spines, in marked contrast to the horizontal transverse processes and tall, narrow neural spines of the diplodocids. So unless the Ontogeny Fairy came along and stole two dorsals and worked some serious magic on the rest, there is no chance that H. is a juvenile of one of the known diplodocids. I won’t even get into the differences in the caudals or the appendicular skeletons.

    Now is probably a good time to point out that Barosaurus has 9 dorsals, not 10 like Diplodocus. Did the Ontogeny Fairy strike again?

    Again I’ve looked at both of the beautiful baby Barosaurus at the AMNH, and they look identical to known juvenile Diplodocus.

    Did you look at the mount, or the material back in collections? Because if it looks identical to juvenile Diplodocus, you missed a few features. The juvenile Barosaurus cervicals are proportionally longer with lower neural spines, more anteriorly overhanging prezygs, and larger ventrolateral flanges on the cervical centra–all features consistent with adult Barosaurus.

    And it is of my utmost professional opinion that at that particular size range Suuwassea and MOR 592 look identical to other known juvenile Diplodocus and Apatosaurus.

    Regarding Suuwassea, are you speaking here just of the presacral vertebrae, or including the skull and appendicular material as well? Because your take is strongly at odds with the extensive, careful work of Jerry Harris and John Whitlock on this point.

    I haven’t seen MOR 592 so I don’t have a dog in that fight. I have seen juvenile Apatosaurus, Diplodocus, and Barosaurus, most notably at the AMNH, Carnegie, and BYU, and all three are distinct from pretty early on in ontogeny.

    What separates Rebbachisauridae from the other two? Some cranial features and lack of bifurcation.

    I’m sorry, that’s just not true. Rebbachisaurids have a very distinctive scap, distinctive dorsal neural spines, distinctive caudals, and tend to have postcranial pneumatic features that are more invasive than in diplodocids whereas the pneumaticity in dicraeosaurids is less invasive. Go look at the Nigersaurus papers and Phil Mannion’s recent work. Sweeping statements like that, which ignore mountains of published evidence, do nothing to help your case.

    Look, I don’t want to be mean. It’s fantastic that you guys have advanced our knowledge of neural spine bifurcation. If you had stopped there, no one would be arguing with you. But it’s one character. I strongly encourage you to have some humility about all the other characters that contradict your proposed synonymies. When you say stuff like Haplocanthosaurus might be a juvenile Diplodocus, you’re just encouraging anyone who has read Hatcher (1903)–or who can count dorsals–to dismiss your work entirely. And that’s too bad, because some of your points are valid.

  25. Cary Woodruff Says:

    Matt,
    All are very good points. Just FYI, I was aware of the dorsal count of Diplodocus and Barosaurus (but remember vertebral counts are known to change), and the Haplo characters, but for now let us agree to disagree. I respect Jerry and John’s work, but again let’s agree to disagree. When we finish the Suuwassea work we can discuss it in more detail again. And you are completely right about the rebbachisurid characters. But bifurcation is a very important one; that character is what Wilson used to originally presume Rebbacisauridae as the most basal of the three subclades. Now I can fully agree that any future reader could examine some of our claims and quickly dismiss our work. But I consider myself a “glass half full” kind of guy. That’s the point of science, not to agree with everyone, but to challenge it. You’ve already done a laudable job at distinguishing Haplocanthosaurus. Well done. What we’re saying is for more people to go out and challenge us further! Sure not every sauropod will be an ontogenetic stage of another, but surely some are, and let us go forth and examine this in greater detail.


  26. So to summarize the paper

    1. All indication is that bifurcation develops ontogenetically (I still can’t tell if Mike agrees with this or not).

    2. We don’t synonymize anything: the invoked cause of this kerfuffle was overinterpreting three words in the conclusion. We just say that caution is advised when assessing the taxonomy/phylogeny of specimens that have less-than-fully -developed bifurcation, and flag up a few taxa where this seems to be relevant. It could be that other features that made these specimens appear distinct may also be ontogenetic/stratigraphic.

    Beware of heterochonry (ontogeny plus stratigraphy). It can make juveniles of one taxon look like adults of a later or earlier taxon (depending on paedomorphosis/peramorphosis; we see this in Triceratops, and it has been mentioned as far back as 1965 for other taxa), so you have to be cautious relying on reading morphology literally (ie. without strat/ontogenetic context). A way we can add resolution it is by getting high-res strat and ontogeny data, and identifying just how each character changes ontogenetically and stratigraphically. In this paper we look at one character ontogenetically: neural spine bifurcation; now we need to look at it stratigraphically.

    This was an undergrad project intended to flag up an issue that had previously gone unpublished.

  27. Mickey Mortimer Says:

    This has been an incredibly interesting thread to read. From my perspective, Mike, Andy and especially Matt won this hands down. While sauropods aren’t really my thing, I fully agree that “Sweeping statements like [“What separates Rebbachisauridae from the other two? Some cranial features and lack of bifurcation”], which ignore mountains of published evidence, do nothing to help your case.” Even the first paper to properly place rebbachisaurids (Calvo and Salgado, 1995) only had bifurcation as 2 out of 10 flagellicaudatan characters. Yet in theropods, we have the parallel of Juravenator being recently hypothesized as a juvenile megalosauroid, which is at least as drastic as Haplocanthosaurus being a juvenile diplodocid. So while I’d like to be able to entertain the latter possibility, things like the rebbachisaurid comment, “let’s agree to disagree” and then backing off and saying you just want people to question things and that Matt’s done a good job in a blog comment countering a point in your published paper… these all just lead me to not take Cary’s results seriously.

    As a final question, Cary said “remember vertebral counts are known to change.” I’d like a citation for dorsal counts changing ontogenetically in dinosaurs without it being due to the incorporation of sacral vertebrae. Haplocanthosaurus, Barosaurus and Diplodocus all have five sacrals, so that can’t be what happened there.

  28. Cary Woodruff Says:

    Mickey,
    I am not saying that scientifically it is a rival, but there was a heaping load of information that the Earth was flat and that all planetary bodies circled the Earth. Just because there is a lot of literature on a subject doesn’t automatically make it right. To be honest I’m beginning to think Denver and I must be speaking a different language. Haplocanthosaurus is not a Diplodocus. El Haplo no Diplo. That one sentence is meant to state that ontogeny through the Morrison sauropods should be explored. Good Lord all the reviewers didn’t throw such a fuss over that sentence! And to be honest I don’t know what more can be said. I don’t care where you stand, bifurcation is an important character in all sauropods with that feature. The particular significance, well take it or leave it. And the “lets agree to disagree” bit well Mike, Matt, Andy, Denver, and I can all agree that blogs are useful sources to bring forth and discuss new ideas, but do you want to spend all your time wasting away on a blog, or actually publishing scientific contributions in the form of papers. All of us do the second one. I bet you spend a lot of time on the DML don’t you? And while I am legitimately grateful to Mike, Matt, and Darren for creating SV-POW, I find it extremely sad that your conclusions about a scientific paper are based solely on the outcomes of a blog and not from reading the work yourself. Frankly I’m glad “scientists” such as yourself don’t believe me. I’ll sleep better tonight knowing my “crap” will never grace the lips of higher dialogue from such. And ok make a reference. Well Mama doesn’t wipe your ass any longer, and neither will I when you have easy access to Google Scholar. But I’ll throw you a bone: Miragaia. Mike Matt, & Andy, did you knows we were in competition, cause I sure didn’t. Point for you guys.

  29. Mike Taylor Says:

    So sorry that I am too busy to get into this properly. But I thought I ought just to address this point that Denver addressed specifically to me: “All indication is that bifurcation develops ontogenetically (I still can’t tell if Mike agrees with this or not)”. I certainly agree that there may well be an ontogenetic component to spine bifurcation. At this stage, I’m not sure how much more can be said but remember that’s still based only on Cary’s Bonn talk and a skim of the paper — I’ve yet to read it in detail. It’s interesting to think about the developmental implications of this: IIRC the left and right halves of the spine arise from separate condensation: it’s strange to think of them uniting in very early ontogeny, then re-dividing as the animal gets older. At any rate, I have no quibble at all with Cary’s meta-point that we should all go out and look in more detail at more sauropods.

    It’s true that this paper does not synonymise anything. But it’s also pretty clearly a staging post towards doing so, and I fear that because it will create a taxonomic mess that we’ll then have to navigate around for a few years until someone nails it down. (That won’t be me — I have too much other stuff to work on.) I don’t want to be negative, but I do have some sympathy with Matt’s “do your homework” admonition, because it’s not really clear from the paper that you’ve looked at the relevant specimens or even necessarily the monographs. To go back to Haplocanthosaurus, there is simply not the ghost of a chance that it’s a juvenile Diplodocus or Apatosaurus. None. So what I am saying is: before publishing a synonymy, please be very careful to cover all the relevant ground.

    I think you’re right that much of our dissatisfaction with the paper comes from its over-reaching for exciting conclusions rather than the core point “look out for ontogenetic changes in bifurcation”. And, yes, it’s good to prod people into awareness of this stuff.

  30. Mike Taylor Says:

    I should also just chip in to say that the idea of rebbachisaurids being dicraeosaurids is just wildly contradicted by the evidence, and unfortunately that gets bodged in the very first sentence of the abstract: “Within Diplodocoidea (Dinosauria: Sauropoda), phylogenetic position of the three subclades Rebbachisauridae, Dicraeosauridae, and Diplodocidae is strongly influenced by a relatively small number of characters.”

    Go and look at the fine appendices to Wilson 2002 (p. 269ff). You will see that he lists eight synapomorphies for Rebbachisauridae, none of them to do with spine bifurcation; and eighteen(!) synapomorphies for the (Diplodocidae + Dicraeosauridae) clade that we now know as Flagellicaudata; and within that clade, a further eight synapomorphies for Dicraeosauridae. I’ve not checked them for reversals, so the total number of characters in that analysis separating Rebbachisauridae from Dicraeosauridae may be somewhat less than 8+18+8 = 34 characters, but it’s not going to be much less. And, really, to try to counteract that mass of character evidence with “but some rebbachisaur neural spines are a little bit bifid” is way beyond mere special pleading. It’s extra-special, super-special.

    So the idea that “Rebbachisauridae may instead represent a clade of late surviving derived dicraeosaurids” honestly hurts the credibility of the paper. Which is a shame, because its core point is reasonable.

  31. Mike Taylor Says:

    Cary asks: “Mike, Matt, Andy, Denver, and I can all agree that blogs are useful sources to bring forth and discuss new ideas, but do you want to spend all your time wasting away on a blog, or actually publishing scientific contributions in the form of papers.”

    Ohohoh, now we get to a real issue!

    What counts?

    Ten years ago, blogs didn’t. In ten years I am pretty sure they will. Now? Hard to say. I converted on SV-POW! post into a Nature Precedings article at http://precedings.nature.com/documents/6878/version/1 — Precedings often appear on CVs whereas blog articles don’t, but really: what’s the difference? In this case, none at all — the Preceding and the blog post that it’s based on are word-for-word identical. I find myself less and less inclined to buy into the idea that the significance of a scientific idea depends on the medium in which it’s promulgated.

    In short, I’m not sure I recognise the distinction between blogs and papers.

    And I am afraid I didn’t understand the second half of that comment at all.


  32. I appreciate the reasonable response Mike. Usually I am the fiery one; here Cary (also Matt earlier) used some choice language reflecting their passion for the research, but which is perhaps best avoided (and I am guilty in the past, just not here).

    Anyway…

    Bifurcation is only 1 character among many, I agree, and we’re not claiming bifurcation trumps all. The language used in the paper might not have been precise enough to convey this, but its hardly a reason to dismiss the entire work.

    The methodology we use at the MOR is new (there was that nature-news article on it, but that is all). We break down morphology into its components: stratigraphy, ontogeny, and phylogeny are the most important ones here.

    Bifurcation:
    Stratigraphic: unknown, in that we haven’t studied whether the pattern of birfurcation (including developmental timing) is different for lower, middle, or upper Morrison taxa. It probably will be.

    Ontogenetic: bifurcated in oldest animals, unbifurcated in smallest, the ontogenetic “stage” at which bifurcation begins is undetermined, the rate at which it proceeds is undetermined (but probably determinable, even with only the current data).

    Phylogenetic: very basal sauropods with bifurcation (e.g. mamenchisaurus show shallow bifurcation in posterior cervicals & anterior dorsals. This pattern is followed in more derived sauropods in that this is where bifurcation both begins (ontogenetically) and is most strongly developed, but derived adult diplodocids have more extensive bifurcation, and it extends further along the neck and towards the sacrum. I suspect that bifurcation happens very late in ontogeny in mamenchisaurs (based on limited juvie material, but we do have some for mamenchisaurus). It probably occurs earlier and earlier through dipodocids as we go from basal to derived (ie. heterochrony). This remains to be demonstrated, but its a good hypothesis based on what we typically see in dinosaur ontogeny.

    So we’ve done this for one of the characters. What about the rest? I don’t know, but you need to understand that just about all characters that are useful vary ontogenetically (this is probably how they evolve: heterochrony). Andy was spot on when he said (I think) that the autapomophy of one taxon becomes ontogenetically changable in other (more derived especially) taxa. that’s heterchrony for you, and it overrides my past criticisms that ontogentically variable characters are not useful (point for andy). You can use them but you need to understand how they are changing (point for us), and you can’t go comparing a fully grown adult with a tiny juvenile, or even a half grown with a juvenile, probably. And if you don;t know where your specimen fits stratigraphically… then you’re basically just using a literal interpretation of morphology. We do this: our conclusion for bifurcation is literal, and without the strat signal it will remain more rough than we’d like, but it is easily falsifiable: find a juvie diplodocid with full bifurcation. refinement will come with further study and better strat data.

    Juvies of one species might look like adults of another:

    We’ve taken 16 species of Triceratops (ignoring Torosaurus for the time being) and found that all that morphological variation can be understood as a combination of ontogenetic and stratigraphic change over about 1.8 million years. Some characters that seemed to unite 2 species of Triceratops turn out to be heterchronic artifacts: juvenile T. prorsus (Upper Hell Creek species) retain many features of T. horridus (Lower species) until they are about half-ish grown, such that if you only had one or two specimens, you might misassign them, but we have lots (just not as many as I would like).

    Might we see this in sauropods? That’s all we’re really saying here. I agree completely that seeing every character as in a constant flux (stratigraphic, ontogenetic) causes a breakdown in the current taxonomy, and I think that this is the issue that most people have. in order to rebuild the phylogeny, you need to consider the ontogenetic variability of each character. We don’t have that for every character as yet. We just published a paper about one character. take all the characters mentioned in this disucssion and try and look at them in the way outlined above: stratigraphic, ontogenetic, phylogenetic.


  33. *I meant to say you can’t put a tiny juvenile into a cladistic analysis with a full grown adult

  34. Mickey Mortimer Says:

    Cary,
    Let’s see here. You use the Galileo Defense, malign me for not publishing and imply I’m not a scientist (ad hominem much?), falsely accuse me of saying you think Haplocanthosaurus could be a Diplodocus when I actually said ‘diplodocid’, put the word crap in quotes which makes it look like you’re quoting me when in actuality I never described your work as crap… The entire point of my comment was that the nature of your comments here don’t lend credibility to your conclusions, and the above just adds to that.

    It’s as if I published a paper on ontogenetic orbit shape change in theropods, pointing out correctly that young theropods have rounder orbits. This is fine, just as I have no issue with your hypothesis bifurcation increases in sauropod ontogeny. But then if I said “What separates Coelurosauria from the megalosauroids and carnosaurs? Some pelvic features and round orbits”, that Coelurosauria would probably be sunk soon, and that Ornitholestes may be a juvenile allosaurid since it wasn’t adult and has round orbits, you’d be right to respond as Matt did and tell me to do my homework. I’d have taken a valid point and buried it in a mass of unsupported extensions. Then I claim geocentrism was also well evidenced at one point and bring your mother wiping your ass into the conversation and yeah…at this point you’d probably be better of just letting my own words make your point for you.

    But back to science. You claim Miragaia as an example of dorsal counts changing ontogenetically in dinosaurs without it being due to the incorporation of sacral vertebrae, but the holotype preserves 15 cervicals and two dorsals, while the juvenile referred specimen preserves two dorsal centra and three dorsal arches. So unless there are specimens I’m unaware of, this is not a valid example.

  35. Tomasz Skawiński Says:

    On a side-note, referring to Dr. Taylor’s comment: today I got an email from NPG:
    “As of April 3rd 2012, we will cease to accept submissions to Nature Precedings. Submitted documents will be processed as usual and hosted provided they are uploaded by midnight on April 3rd. Nature Precedings will then be archived, and the archive will be maintained by NPG, while all hosted content will remain freely accessible to all.” (probably many people already know that but maybe not all.)

  36. Nick Gardner Says:

    All back and forth sniping aside, the above comment regarding NPG closing Nature Precedings leads me to ask, does anyone have an opinion on the new journal being put out by the Faculty of 1000 group, F1000 Research (http://f1000research.com/)?

    Would be interested in comments by all.

  37. David Marjanović Says:

    Note that the leftmost part of fig. 7A is in lateral view.

    Is that actually true, or is that a lapse that the reviewers and editors overlooked?

    What separates Rebbachisauridae from the other two? Some cranial features and lack of bifurcation.

    Let me just say how refreshing such a statement is when I’ve been working on a craniocentric matrix all the time. 195 skull/lower jaw/tooth characters out of a total of 284 characters… It’s about tetrapod-sensu-lato phylogeny and the origin of Lissamphibia.

    (but remember vertebral counts are known to change)

    …Are you trying to say that dorsals change into cervicals in sauropod ontogeny? I could just barely imagine that, though things like the deltoideus muscle would make it… difficult.

    IIRC the left and right halves of the spine arise from separate condensation: it’s strange to think of them uniting in very early ontogeny, then re-dividing as the animal gets older.

    The left and the right neural arch indeed arise as separate entities. In “larval” temnospondyls they take a long time to fuse, and in the vaguely paedomorphic lysorophian “lepospondyls” they suture but never fuse. And yes, the contact surface extends all the way up through the neural spine.

    It’s definitely unique, but I can imagine the left and the right neural spines continuing to grow in their middles and not at their contact surface anymore.

    (Pleurocentra and intercentra are paired structures, too.)

    just about all characters that are useful vary ontogenetically (this is probably how they evolve: heterochrony)

    Seconded.

    *I meant to say you can’t put a tiny juvenile into a cladistic analysis with a full grown adult

    You can, but you have to score all of its ontogeny-dependent characters as unknown*, and that means you have to identify them first, which is probably your point.

    * Or rather as “the observed state or ontogenetically later” – partial uncertainty, of which complete uncertainty is the extreme case.

  38. David Marjanović Says:

    I forgot to mention that there are taxa with individual variation in the count of presacral vertebrae even when there’s just one sacral vertebra. The stereospondylomorph temnospondyl Sclerocephalus haeuseri, which is known from shitloads of articulated skeletons, is a case in point.

    Rainer R. Schoch & Florian Witzmann (2009): Osteology and relationships of the temnospondyl genus Sclerocephalus. Zool. J. Linn. Soc. 157: 135 – 168

    Also, the left and right neural arches of the atlas commonly don’t even meet in temnospondyls, not even when they fuse to the intercentrum. Each has its own neural spine, and a process of the fused neural spines of the axis reaches between them!

    ===============================

    So, now I’ll take a bath in the flamewar.

    but for now let us agree to disagree. I respect Jerry and John’s work, but again let’s agree to disagree.

    WTF? No. This is science you’re talking about.

    We can agree to wait for the paper, fine. But when there are n hypotheses/speculations/ideas, at least n – 1 of them are wrong.

    Science is the search for all falsehood, so that only reality may be left standing.

    And you are completely right about the rebbachisurid characters. But bifurcation is a very important one;

    The thing about unweighted parsimony is that there is no such thing as an important character. No, not even when the authors of a phylogenetic analysis claim there is* – they may care, but PAUP* and TNT simply don’t.

    * No idea if Wilson or anyone did.

    You’ve already done a laudable job at distinguishing Haplocanthosaurus. Well done. What we’re saying is for more people to go out and challenge us further!

    …Shouldn’t you have challenged yourself as far as possible before submitting your manuscript?

    And I am afraid I didn’t understand the second half of that comment at all.

    That’s because it’s an incoherent, brainstorming, wall-of-text rant that has nothing to do with the paper.

    Cary, when you’ll learn to write in paragraphs, maybe you’ll learn to think in paragraphs, too…


  39. To second David, but for different reasons, you can put a juvenile specimen of any sort into a cladistic analysis with adult specimens, so long as you are aware of what you’re doing, and what you’re intended to accomplish with the act. Cladisitc analysis is a form of cluster analysis, which also includes eigenshape analysis. Setting shapes of bones and skulls and vertebrae as eigenvalues and computing relative similarity can track resemblances across a broad range of shapes, which can then be transformed into eigenvalues and those can be analyzed using heuristic search algorithms and employing parsimony. This is higher end stuff, and few people (Whitlock excluded, of course) have attempted to push their shape arguments and ontogenetic change variation to this level in dinosaurian morphomerics. It really, really should be done to vertebrae, scapulae — well, just about anything!


  40. You shouldn’t code a juvenile as an adult because it hasn’t got its full suite of adult characters yet (if any). You can make new characters for the species:

    e.g. for half grown specimen:

    “level of bifurcation (some sort of ratio) at 1/2 grown”

    and then code

    “level of bifurcation at full grown” as unknown (unless you know both 1/2 grown and adult states.

    But if you code the juvie the same as an adult, it is going to misrepresent the taxon, and -typically- get placed much more basally than would the adult.

    The problem is that if most useful characters change throughout ontogeny, you end up only being able to properly compare like for like growth stages. Fine if you have all adults, not as useful if you only have one species at half grown, and another at full grown. If you’re lucky, there might be some adult characters present in the half grown, but how would you know unless you know the ontogenetic trajectory of each character?


  41. I would like to point out that what I was talking about is not a phylogenetic analysis, merely a cladistic analysis. The two are different in that the former is a subset of the latter, just as eigenshape analysis is a form of clustering analysis, as is cladistic analysis (and therefore phylogenetic analysis using cladistic algorithms).

    Thus, when one is placing specimens of variable ontogeny within a single matrix, the process by which the matrix is produced and the purpose of the analysis must be separated from phylogenetics.

    One can, for example, assess a range of shape variables through eigenspace and express it as a tree of relatedness, as Dodson did with Protoceratops andrewsi skulls, followed by Chapman, although the processes by which those authors achieved their results was anticipatory of current geometric morphometrics for which eigenshape analysis is currently employed.

    Applying this form of analysis to very complex structures, such as vertebrae or skull shape, can and will elucidate both a plot-based “clustering” and a tree-based “cladistic” output, but neither are intended to infer actual relationship, merely similarity. And yes, Carr shows that applying this thinking to tyrannosaurs will result in juvenile animals clustering at the bases of phylogenetic dispersal, due to a lack of distinguishing “adult” features. But it cannot tell you that the animal is, in fact, a juvenile of a particular species, meaning that the types of Suuwassea emilieae, Haplocanthosaurus priscus and delfsi, juveniles that they may be, may be juveniles of taxa that are distinct from other taxa, regardless of apparent juvenility.


  42. and that is a decent explanation of “diversity first”: unfalsifiable belief that somehow these specimens are immature forms of unknown adults rather than juveniles of known adults.

    Which is where stratigraphic parsimony comes in.

    And it doesn’t really matter if you are only interested is which taxon is more closely related to other taxa, but when people strat talking about diversity and radiations, you need the number of species to be as correct as it can be.

    In many ways, this is why I think in phenes instead of species. Every specimen’s morphology is a single phenotype expressed as a combination of its position on the axes of ontogeny and stratigraphy (plus individual & sexual variation, which are what you are left with when excluding the former two). What happens when stratigraphy stays the same but position on the ontogeny axis changes (ie. looking at different growth stages at a single stratigraphic horizon, e.g. a bonebed; making some assumptions about the bonebed formation), and vice versa (same growth stage through geologic time).

  43. Mike Taylor Says:

    Denver, you keep writing about “diversity first” as though it’s some dumb trap that everyone except your lab has fallen into. But what we’re in fact talking about is “diversity when supported by numerous apomophies and a reasonable judgement of ontogenetic and stratigraphic factors”.

    In Dirk Gently’s Holistic Detective Agency, Inspector Gilks tells Dirk that he’s fallen into a trap that clever people often fall into — assuming everyone else is stupid. Whether or not it’s your (and Cary’s and Horner’s) intention, the impression you give is that you think you’re the first people ever to have thought of considering ontogeny. The reality of course is very different — it’s well over a hundred years ago that Riggs synonymised poor old Brontosaurus with Apatosaurus on ontogenetic grounds. Y’all give the impression that you’ve come up with a new golden key (that woefully underspecified Nature article didn’t help) as though there hasn’t already been a century of careful, detailed work by intelligent, thoughtful palaeontologists. No-one’s saying their work is sacrosanct and can’t be provewn wrong; but it’s going to take a lot more to overturn all that work than “aha, but neural spines split during ontogeny, ha, you never thought of that!”.


  44. You’re not doing ontogeny and stratigraphy. You’re just doing morphology.

    Don’t keep trying to put words in my mouth. I ask, how many of these characters have been assessed ontogenetically and stratigraphically, and you say that I am saying everyone else is dumb.


  45. So, I’m not commenting again. I don’t see what other points I can make other than refuting things that you report here that we have said, but we didn’t. We made a few errors in choice of words used (I’ll admit that, although our intent is not what was conveyed here), but you propose that it invalidates the entire work.

    And stop sending out emails to my co-author suggesting his only route back to credibility (in your terms) is to throw himself publicly at the feet of your blog. I find that approach to be little short of bullying a junior researcher. I am guilty of being aggressive before (these days I am making a conscious effort to be civil in correspondence), but you are guilty of bullying here, you’ve bullied other researchers before, and its about time you stopped.

    We published a paper suggesting that bifurcation is ontogenetic. No-one had published this before. Our reviewers, other sauropod workers at meetings, and many people at the recent sauropod biology workshop (I wasn’t there; 2nd hand info) thought it was a decent hypothesis.

    I think that bifurcation ontogeny gives an indication that many other apomorphies may turn out to be characters that vary ontogenetically and this may affect phylogenetic positioning / validity in a range of taxa, especially those based on immature specimens. I does not matter how many apomorphies you think is enough to justify a taxon: pretty much ALL “apomorphies” in Triceratops (and tyrannosaurs) change through ontogeny, and I expect the same to be the case in sauropods.

    As a consequence, I think that big question marks hover over Haplocanthosaurus and Rebbachisauridae, and the way we might resolve those issues is tracing characters ontogenetically and stratigraphically. Maybe we overstate our position in the discussion, but I don’t think it changes the implications.

    Falsify our hypothesis.

  46. Matt Wedel Says:

    And stop sending out emails to my co-author suggesting his only route back to credibility (in your terms) is to throw himself publicly at the feet of your blog. I find that approach to be little short of bullying a junior researcher.

    Wow, and you complain about having your words misrepresented! If you have seen the email in question, you know that this is a gross misrepresentation of it. If you haven’t seen the email in question, kindly shut up.

    For those of you who are unfortunately hearing about something that Mike was trying to handle quietly and tactfully, I’ll paraphrase the content of the message that Mike sent to Cary.

    First, he thanked him for engaging here in comments section of the blog, because it makes this a much more interesting place to visit.

    Second, Mike said that he felt bad about contradicting Cary on some points, because the paper was something he was happy to see.

    Third, he suggested that Cary was not doing himself any favors by refusing to back down on certain speculative points, namely the potential for Barosaurus and Haplocanthosaurus to be junior synonyms of Apatosaurus and Diplodocus. Those synonymies are just not possible, given that we have a reasonably good ontogenetic series of Barosaurus that is distinct from both Apatosaurus and Diplodocus all the way along (I have photos that show this and I’ll be happy to post them when I get time), and that Haplocanthosaurus is so wildly different in just about every way. Mike merely pointed out that vociferously defending such unlikely hypotheses might be unwise.

    And…that’s it. Mike never asked Cary to throw himself at anyone’s feet, here or anywhere else. He saw a junior researcher doing something he felt was unwise and attempted to give him a little advice, privately.

    If you’re keeping score, Mike’s email was roughly one part thanks, one part apology, and one part advice. No incrimination, no bullying, all out of the public eye.

    So, thanks for blowing that wide open by lying about what was said.

    As for the paper, you keep endlessly going on about ‘falsify the hypothesis’. As we’ve made clear at just about every possible point, we don’t have a big problem with the idea that bifurcation proceeds ontogenetically. We have a medium-sized problem in that there is very little control for serial variation in your sample. And we have a huge problem with the synonymies proposed in the paper, especially Barosaurus and Haplocanthosaurus, and with the general method of ignoring shedloads of characters in favor of one (e.g., rebbachisaurids are separated from dicraeosaurids by “some cranial features and lack of bifurcation” vs. the 30+ character differences in Wilson 2002).

    Now, as you’ve repeatedly pointed out, the proposed synonymies are not the main thrust of the paper. We get that. But they’re still in the paper, and your names are on the paper. So man up and take some responsibility when we criticize those hypotheses. If you really think that Barosaurus and Haplocanthosaurus are juveniles of Apatosaurus and Diplodocus, you’re going to have to get used to taking some heat. And if you no longer think that, just say so. We know something about that–we’ve had to admit that we were wrong a fair amount (Exhibit A, Exhibit B; there are more). So far it has been Mike and I that have been trying to keep the focus on the science, bringing up contradictory evidence and citing it, and you and your coauthor who have been whining about bullying and telling people to wipe their own asses. That grandstanding isn’t lending any credibility to your arguments.

  47. Mike Taylor Says:

    Well, this is rather unpleasant. I find I’m accused of bullying. My message to Cary was sent privately in the hope of helping him out, but since Denver has raised the subject and lied about the contents, I am going to have to publish it now. Thanks to Matt for paraphrasing, but since Denver’s made this about what I said, anyone wanting to judge the validity of his accusation needs to see the actual text.

    So here it is.

    Date: 1 March 2012 22:54
    From: Mike Taylor mike@indexdata.com via gmail.com
    To: sauropod4@gmail.com
    Cc: Mathew Wedel
    Subject: SV-POW! discussion of your paper

    Hey, Cary,

    Two things. First, thanks for engaging — it makes for a much more interesting blog.

    Second, I feel a bit bad about having to contradict you so much, when this is a study that I wanted to see, and encouraged you with. I hope it’s now clear that the issues are with the speculative conclusions more than than body.

    But a word to the wise: you’re not doing yourself any favours by not backing down over some of these. Contra the protestations in your last comment, the paper does say “Taxa defined on small specimens (such as Suuwassea, but also potentially Barosaurus, Haplocanthosaurus, and ‘‘Brontodiplodocus’’), might represent immature forms of Diplodocus or Apatosaurus”. And I am afraid that is Just Plain Wrong. It’s not open to interpretation and it’s not something where you can validly “agree to disagree”. (To be clear: I have no opinion on Suuwassea.
    I’m talking here about Barasaurus and ESPECIALLY Haplo.) Your
    credibility will be better served by just admitting that.

    Hope you read this in the spirit I write it. Truly, I’m on your side.

    Feel free to forward this to Denver if you wish, but it’s basically you that I’m talking to.

    — Mike.

    That’s it. You be the judge, dear reader, of whether this is “little short of bullying a junior researcher”.

    I’d be interested to know whether that’s Cary’s interpretation, too. I can tell you that it certainly isn’t what he’s conveyed to us off-list, but I won’t quote his response without permission.

  48. Cary Woodruff Says:

    Well this certainly isn’t how one wishes to start his day. I will first and foremost say what I have said from day one, I thank Mike and Matt for finding this topic interesting enough to warrant further in depth discussion. And second I appreciate ALL of the advice everyone has given me. While this discussion has had ups, and very low downs, it’s has certainly taught me a lot. Most folks follow and agree with bifurcation happening ontogenetically. Fine. Some concern is raised over the serial position. Fine. Some question the phylogenetic implications. Fine. I don’t know of a single paper I view as “perfect”. In any work there are bits you like, bits you don’t, and bits you think could use some tweaking. I’m certainly not done with the story of bifurcation and all of these are great points that I can modify, reword, or certainly clarify in greater detail in future works. I certainly do agree with Denver that stratigraphy and ontogeny need to go hand in hand (I’m doing the strat now in case anyone was wondering), and this combo and way of thinking is not terribly widespread in our field and needs to be used much more often. And I fully agree with Matt, if we are going to be saying some controversial things, then we need to have a thick skin and get used to the heat. Now I’m sure everyone at some point or another in this discussion has said something they might have addressed differently. Lord knows I’m guilty of that! Mickey I’m sorry I went off on you. It wasn’t professional or personal (just sort of the wrong place at the wrong time). This whole thing is a snowball effect. Now I will fully admit that I wish everyone thought this was the greatest paper in the world and that parades would be thrown in my honor. As of this post I have not received any parade instructions. So I wish some comments were more uplifting, again as Matt said, big deal. If you can’t stand the heat get out of the kitchen. And as the hypocrite that I am, I hate this snarky, personal aspect. I don’t want to burn professional bridges and make folks think I cannot carry on an intelligent conversation. Selfish case in point #1: I plan on working on sauropod pneumaticity coming up. Now Matt is far more professional than this, but I don’t want him to hate me or be repulsed by conversing with me. Selfish case #2: Maybe someday I want to work on the ontogeny of the Tendaguru sauropods, and for the same reasons I don’t want Mike to blacklist me. At the end of the day it comes down to chocolate and vanilla. I like chocolate, you may like vanilla. We may disagree at times over which flavor is the best; but I’m not going to hate a “vanilla lover” (and I would hope the same applies to me!). We can debate for hours the merits of each case, but at the end of the day it all comes down to our own feelings about what the other has presented. I have no animosities or feelings of ill will, and I do value all of the useful comments and advice; I may not agree with particular aspects, but they are still valuable none the less. Mike and Matt, I’m sorry this discussion has gotten out of hand. You guys are running the show, but I would suggest that for now we call this particular post done. I’d love for us to pick this up at a later time and look forward to valuable conversations. If anyone has something to ask or discuss you are more than welcome to write me at sauropod4@gmail.com.

  49. Mickey Mortimer Says:

    Thank you. Apology accepted.


  50. Sigh. iIsaid I wasn’t going to comment again, but I guess I have to straighten out a few things, at least from my perspective.

    1. I am unconvinced by the evidence that Suuwassea is a dicraeosaurid. See our paper for discussion for more than just bifurcation.

    2. You’re right that Haplo is not Diplodocus or Apato, and we misstated MY intent in the paper (which makes ME wrong). this said, I think that there is much more to the Haplo & Barosaurus phylogeny story than is currently published, which leads us to the next point.

    3. It doesn’t matter how many characters separate one taxon from another if the context of those characters is not known. The superficial implication from our paper is that bifurcation is ontogenetic, so a single character once thought to have some significance has a more complex background that needs to be understood if it is to be used again. The deeper significance of this is that a specimen that exhibits incompletely formed bifurcation may have other aspects of its anatomy (one of those 10 or 22 other apomorphies) that are similarly affected, unknowingly, by ontogeny (or in the case of different stratigraphy: heterochrony). Each character needs to be investigated for how it changes stratigraphically and ontogenetically. Cary’s working on this for a few things. When we have a good ontogenetic and stratigraphic map for apomorphies, then I suspect we will have a much better ideas to where Rebbachisaurids (etc) fit into the picture. I remain unconvinced by the current cladograms, have doubts about some characters, and we need to do more work to get a clearer picture. Right now, that’s all I have, or at least am willing to share.

    4. Bullying. Regardless of your intent, the tone of the original post belittles this paper, and the Triceratops work. You don’t have to agree with the findings or the discussion, but my first post asked you not to misrepresent the material we use (so i’m not the only one at fault here). Doing so unreasonably invites the reader to draw doubt on our findings, and our ability to conduct research. You state that we use “mostly isolated elements, nearly all of them damaged by breakage and/or crushing, and uncontrolled for serial position beyond some basic binning”: MOR592 is a decent neck (11 cervs), 4 dorsals, and a partial skull. hardly fragmentary, and it’s in decent condition. The Mother’s day site contains mixed remains of many juveniles. We don;t know which vert belongs to which individual, but there are many specimens (probably representing most serial positions; not all isolated specimens, quite the opposite) and they all show bifurcation that is less developed than in large diplodocoids. Obviously the degree of bifurcation varies by (inferred) serial position, but this doesn’t disagree with our main point.

    You don’t come back and say that your initial wording was a misrepresentation and you don’t change the text in the post (although you do correct spelling of Cary’s name). Later replies belittle my supervisors research approach and use charged language to do so. I think I have been nothing but civil in my own replies (for a change: as I say, I am making an effort).

    However nice you think your wording is, you suggest in your email that people will think our work has no credibility because of the mistakes we (or at least I) made in the discussion (fair enough), unless we admit on your blog that we were wrong (not so fair enough). Noone wants their peers to doubt their credibility, and maybe you’re right that we need to put a public note somewhere. I do this on my own website usually; I added a note to the Alamosaurus paper that the vert we identify as a caudal is actually (probably) a crushed dorsal (this affects the interpretation of size, although the cervical is still fine). By thinking we somehow need to put an apology of sorts on your blog, this feels like some sort of superiority thing, regardless o how friendly you think you are being. Cary’s thick-skinned, he’s not going to care too much, but I wouldn’t be the only person who is dissatisfied by the barrage approach taken on this blog.

    So anyway, I think I cleared up the Haplo thing, and clarified the other points. You can write whatever you like on your blog. I only ask that you don’t misrepresent our data, and I would rather you avoided the snarky tone of other bloggers with regards to the ontogeny work.

  51. Matt Wedel Says:

    Okay, just asking for a point of information, one researcher to another. Cary or Denver, do you know the serial positions of the MOR 592 vertebrae pictured in Figs. 3 and 4 in the paper, and if so would you be willing to share that information? Approximate positions are fine, I’m just trying to figure out which vertebrae I’m looking at. Thanks in advance for any help.

  52. Mickey Mortimer Says:

    Since this conversation is continuing (which is good), I’d like to say that while I don’t think Woodruff and Fowler’s published remarks accurately reflect the state of rebbachisaurid monophyly, this in no way means that they’re wrong about the family being unstable. It’s perfectly possible to have a clade be supported by numerous synapomorphies, yet also collapse easily due to conflicting data. Whitlock (2011) finds that while Limaysaurinae+Nigersaurinae is strongly supported, the incompleteness of the more basal Rebbachisaurus makes its inclusion less certain. And if it goes, its family goes with it. Similarly, my in prep. coelurosaur analysis finds some decidedly non-traditional groupings, largely due to the poor state of most published analyses. So I’d not be surprised if Sauropoda were similar, though on the surface Whitlock’s analysis seems excellent (but so did Smith et al.’s 2007 Cryolophosaurus one until I delved into it). Thus I’m not opposed to rebbachisaurid polyphyly, but would be more impressed by “when bifurcation characters are deleted from Whitlock’s analysis, rebbachisaurids are internested with flagellicaudatans in trees 3 steps longer than the most parsimonious one” than “bifurcation is ontogenetic and a dentary chin may be symplesiomorphic, thus Rebbachisauridae is questionable.”

  53. Mike Taylor Says:

    That’s a pretty classy apology for your totally unsubstantiated accusation of bullying, Denver, thanks for that.


  54. Hi Mickey,

    Fair point. Cary did actually run a phylogenetic analysis as part of the original paper, but this was removed at request of the reviewers. Bifurcation is only one character and it is unlikely to make a huge difference when you have 10 other characters, however, I am interested in seeing what the ontogenetic and stratigraphic histories are for these other characters. in some sense we are going down one hill in order to go back up another, and this does involve backward steps to begin with, but that’s the way it works.

    Since I’m here, I’ll take the opportunity to apologise (to Mickey) for being unreasonable/mean in past communications (elsewhere). Mickey, I’m sorry; I get a bit riled up about things and have let this get out of hand in the past. If you are interested in a research project (that I think would suit you) then get in touch.

    Mike: You misrepresnted our dataset in your original post (and have not acknowledged this). I have admitted our/my scientific faults where you rightly pointed them out, but you’re going to have to accept that regardless of your intent, I was offended by your comments (“everything is triceratops”; “tarred with the hornerite brush”; “ontogeny fairy”; etc which would never get in a published journal). Ganging up on a junior researcher and proposing that noone will take their research seriously unless they publicly apologise on your personal blog looks like bullying to me (Cary doesn’t care), no matter how nicely it is worded. If this was not your intent, then you shouldn’t precede your “nice” email with the belittling comments above, which I think are fully intended to cause offence. If public admission is your only issue, I will add a note to my personal website on the paper as I have done for previous papers (I also often add enlarged images as you also suggested).

    Matt: Cary can probably tell you which serial positions he thinks are missing / which are present. in the meantime, I have posted a picture of the mount on FlickR, for anyone that wants to see it. Apologies for the big Alamosaurus cervical model in the foreground: this is the only photo I had to hand that shows all the verts at once.

    Diplodocoidea indet

  55. Matt Wedel Says:

    Ganging up on a junior researcher and proposing that noone will take their research seriously unless they publicly apologise on your personal blog looks like bullying to me

    Denver, you have to stop saying this. Mike has already posted the full text of the email and as anyone who is literate can see, there is no demand for a public apology and certainly no suggestion that Cary has to throw himself on the mercy of our blog. Maintaining this fiction in the face of contrary evidence is just pathetic.

    It doesn’t matter how many characters separate one taxon from another if the context of those characters is not known. The superficial implication from our paper is that bifurcation is ontogenetic, so a single character once thought to have some significance has a more complex background that needs to be understood if it is to be used again. The deeper significance of this is that a specimen that exhibits incompletely formed bifurcation may have other aspects of its anatomy (one of those 10 or 22 other apomorphies) that are similarly affected, unknowingly, by ontogeny (or in the case of different stratigraphy: heterochrony). Each character needs to be investigated for how it changes stratigraphically and ontogenetically. Cary’s working on this for a few things. When we have a good ontogenetic and stratigraphic map for apomorphies, then I suspect we will have a much better ideas to where Rebbachisaurids (etc) fit into the picture. I remain unconvinced by the current cladograms, have doubts about some characters, and we need to do more work to get a clearer picture. Right now, that’s all I have, or at least am willing to share.

    Yeah, sorry, no. That’s no how science works. You don’t get to wave away dozens of characters that might be affected by ontogeny just because you’ve published a paper on one. That’s not falsifying anything. If you have evidence that all or even a majority of those other characters are invalidated by ontogeny, then show it and make your case. And if you don’t have that evidence, then drop this “we can’t know anything unless we know everything” line. We all understand that science is a process and that all hypotheses are subject to revision in the light of new data. But the last two decades of work on sauropod phylogeny are not automatically invalidated by the possibility of (as-yet-undocumented) new data.

    And as long as we’re having this discussion again:

    We’ve taken 16 species of Triceratops (ignoring Torosaurus for the time being) and found that all that morphological variation can be understood as a combination of ontogenetic and stratigraphic change over about 1.8 million years.

    Wow. Very modest of you to claim credit for this. It’s almost as if Ostrom and Wellnhofer hadn’t already synonymized all the other species of Triceratops into T. horridus back in the 90s, on the basis of intraspecific variation. But they did. So why do you get to claim credit for that?

    Anyway, the point is pretty well irrelevant for sauropods. We’re not talking about variation in just the shape of the horns and frills here. All of the synonymies proposed in the paper–Suuwassea, Haplocanthosaurus, Barosaurus, and the rebbachisaurids–involve ignoring a large set of character evidence in favor of a smaller one, and usually just one. That’s special pleading, and you shouldn’t be surprised that people are disagreeing with you on it.


  56. Matt: Please stop being deliberately antagonistic. I have asked your blog to not to misrepresent our dataset, which it currently does (in my opinion), and you haven’t acknowledged this. I was offended by the tone of your personal posts, and if this means that I misinterpreted the email then I’m sorry, but you get my back up by saying the things you did, and that’s why I am purposefully not using any charged language as it doesn’t help anyone.

    Ostrom & Wellnhofer proposed synonomy of all Triceratops species into one: T. horridus. They also suggested that maybe Torosaurus was a male Triceratops.

    Forster’s morphometric research showed that Triceratops morphologies fell out into 2 species: T.horridus & T.prorsus.

    These are both good studies, but they can’t both be right. We looked at the stratigraphic and ontogenetic distribution of morphologic variation of Triceratops (inclduing going out there and relocating old quarries, along with digging up new specimens) and found that the 16 original species can be explained as ontogenetic and stratigraphic morphs of more than two kinds (including intermediates). John is inclined to not name multiple Triceratops chronospecies, I think names are useful (for communication purposes) so long as people don’t think it represents diversity.

    Obviously I/we are not taking credit away from Ostrom, Wellnhofer, and Forster. The best way to think about this is that we started with a big pile of morphology (ie. ignoring all current taxonomy; but using the characters they described) and tried to make sense of it by analysing the morphology in stratigraphic and ontogenetic context. This involved a necessary backward step in the construction of the phylogeny, and in the recognition of the characters that had been previously studied. we found that some morphologic trends are stratigraphic, others are ontogenetic (usually a combination of both).

    We don’t propose synonomies in the Woodruff & Fowler paper, and I have added a note to my website to reflect what we do mean (I acknowledge the wording in the paper is sloppy and conveys an incorrect meaning; you were right). We suggest that currently unrecognized morphologic information may become apparent through studying each character within an ontogenetic and stratigraphic framework. We’re just attacking the problem from different angles than you and mike.

    I’m not getting angry about this. We’re just looking at aspects of each character, I don’t understand why this is a problem? we showed that ontogeny is an issue in bifurcation such that caution is advised when using degree of bifurcation as a character. Why can’t this be the case for other characters? We do need to demonstrate this, but for now it leaves a question mark hanging over characters that have not been assessed in this way. This might affect phylogeny and/or taxonomy. That’s all we mean.


  57. I wish I was physically present at all your homes to give all of you a slap on the back of the head.

    ;)


  58. I haven’t read anything here since my last post.

    I apologise for the email comment, if I got your intent wrong.

    However, I still do not think your characterisation of our work on your blog was fair.


  59. […] discussion over the new paper by Woodruff and Fowler (2012)–see this post and the unusually energetic comment thread that follows–made me want to go back to the […]

  60. Hikaru Amano Says:

    Good Day, everyone:

    This looks like a spin-off of what is happening with the TAST (Torosaurus as senescent Triceratops) hypothesis, only that this time it focuses on Morrison sauropods. It appears that both Dr. Taylor and Dr. Fowler have good points-nevertheless I would take the side of Dr. Farke here. Closely-related species often have very similar trajectories of ontogeny(but not in all cases-and in instances that they do, not to the same extent or same speed). Nowadays, it turns out a clash between lumpers and splitters are heating up to a whole new platform based on morphological, stratigraphic, biogeographic, and histological analyses. I know that I have said this many times before and some of you have already read it, but the most effective way to see which and/or when such kinds of analyses will give more accurate results is if scientists would be willing to break dinosaur bones to look for viable soft tissues to be used for biomolecular profiling (amino acid sequence analyses are fine, but nucleotide sequence examination would be far better-assuming scientists are lucky enough to find some of the latter in good quality and in sufficient amounts). Since many closely-related yet distinct taxa have more-or-less identical skeletal mophologies and differ only at their external mophologies(i.e. integument pattern and/or color distribution), behavioral patterns, and/or differences in their identical genes’ consensus DNA sequences, the earlier mentioned four methods may lead to SUPERFICIAL LUMPING if not used in combination with molecular examinations.

  61. Hikaru Amano Says:

    And oh, even though I’m primarily inclined towards biomolecular examinations, I strongly advocate using it together with morphological, biogeographical, stratigraphic, and histological analyses.

  62. Mike Taylor Says:

    Biomolecular analysis would be great if we could do it. But suffers it from the obvious problems that biomolecules are hard to find in fossils and difficult to avoid contaminating. Do you know of any cases where a biomolecular cladistic analysis has been used to elucidate the relationships of extinct animals? Ideally, Mesozoic animals, since more recent extinct animals have a better chance of preserving the relevant molecules.

  63. Hikaru Amano Says:

    The only instances so far are those 2007 and 2009 articles of Dr. Schweitzer’s team plus the 2009 article of Dr. Manning’s team(although those studies’ results are still very preliminary due to the problems you have fleshed out and the limited number of taxa included in the analyses, they do show that we could do it on Mesozoic taxa).

    References:

    Manning,P.L., Morris,P.M., McMahon,A., Jones,E., et al. (2009). Mineralized soft-tissue structure and chemistry in a mummified hadrosaur from the Hell Creek Formation, North Dakota (USA). Proc. R. Soc. B, 276(1672): 3429-3437.

    Schweitzer, M.H., Wittmeyer, J.L., and Horner, J.(2007). Soft tissue and cellular preservation in vertebrate skeletal elements from the Cretaceous to the present. Proc. R. Soc. B. 274, 183-197.

    Schweitzer MH, Suo Z, Avci R, Asara JM, Allen MA, et al. (2007) Analyses of soft tissue from Tyrannosaurus rex suggest the presence of protein. Science 316: 277–280.

    Schweitzer, M. H. et al. 2009. Biomolecular Characterization and protein sequences of the campanian hadrosaur B. canadensis. Science 324:626-631.

  64. Mike Taylor Says:

    Right — these are studies that assert the existence of biomolecules in Mesozoic fossils (though they are no uncontested). But it seems to me that it’s a long way to go from here to a point where we have the possibility to include molecular data in phylogenetic analyses.

  65. Tor Bertin Says:

    Mike,

    Höss et al. (1996) successfully reconstructed >1,100 base pairs of mitochondrial rDNA extracted from extinct ground sloths ~ 13,000 years in age, and were able to make some fairly interesting phylogenetic inferences, though most of their comparisons were made with extant taxa. I’m also aware of, though can’t find the reference for, some molecular work that’s been done with machairodontine felids, as well as cave bears and mammoths. Then there’s the work that’s been done with extinct humans…

    That said, I’m extremely skeptical of the feasibility of using molecular phylogenies for non-avian dinosaurs, given that one would not only need to successfully amplify DNA from what would almost inevitably be highly degraded material, but would need to amplify phylogenetically informative sequences. These problems are then further compounded if one would want to make comparisons across multiple extinct organisms.

  66. David Marjanović Says:

    Cary did actually run a phylogenetic analysis as part of the original paper, but this was removed at request of the reviewers.

    ~:-| May I ask why?

    amino acid sequence analyses are fine, but nucleotide sequence examination would be far better-assuming scientists are lucky enough to find some of the latter in good quality and in sufficient amounts

    DNA simply doesn’t survive for longer than 100,000 years.

    Unless it’s frozen – but the only place that may have stayed frozen since the Mesozoic are the Gamburtsev mountains in Antarctica, and they’re covered by about a mile of ice.

    Proteins probably can survive for much, much longer; but contamination is extremely difficult to rule out.

  67. Mike Taylor Says:

    Cary did actually run a phylogenetic analysis as part of the original paper, but this was removed at request of the reviewers.

    ~:-| May I ask why?

    I’ve had this request on occasion, too — always seems dumb to me.

    Still, even if you remove the actual analysis from the paper, there’s no reason you can’t still say thingsa like “when we re-ran the phylogenetic analysis of Taylor (2009) with Rebbachisauridae constained to be contained within Dicraeosauridae, 19 further steps were needed”.


  68. DNA simply doesn’t survive for longer than 100,000 years.

    Unless it’s frozen – but the only place that may have stayed frozen since the Mesozoic are the Gamburtsev mountains in Antarctica, and they’re covered by about a mile of ice.

    I’ve read about 1 million years [under ideal conditions] (Waggoner 2001, Willerslev and Cooper 2005) but it is still far from Mesozoic.

  69. LeeB Says:

    Quite a bit of Biomolecular cladisitc analysis has been done on Pleistocene horses.
    One of the results was the discovery that bone from south-west Siberia did not correspond with any known living or fossil horse whose DNA was known, further studies on the morphology of this horse led to the description of Equus ovodovi.

    Other results show that Hippidion from South America belongs within Equus as a distinct species and Equus(Amerhippus) species actually appear to belong in Equus caballus.
    And the “stilt legged horses” from North America are a distinct Equus species related to the South American species previously considered to be the distinct genus Hippidion.

    Some work has been done on Pleistocene Bison, Mammuthus and Panthera (Cave Lion and American Lion) as well.

    LeeB.

  70. Mike Taylor Says:

    I didn’t know about that — how cool!

    Do you have references to hand?

  71. LeeB Says:

    The original paper noting a new horse species using DNA was:
    Orlando et. al. 2009, PNAS 106(51) 21754-21759; Revising the recent evolutionary history of equids using ancient DNA.
    The susequent description of E. ovodovi is in:
    Eisenmann and Sergej, Geodiversitas 33(3) 519-530; Unexpected finding of a new Equus species (Mammalia, Perissodactyla) belonging to a supposedly extinct sungenus in Late Pleistocene deposits of Khakassia (S.W. Siberia).

    The paper discussing Hippidion and Equus(Amerhippus) is in:
    Orlando et. al. 2008, Journal of molecular evolution 66(5) 533-8; Ancient DNA clarifies the evolutionary history of american Late Pleistocene equids.

    The last paper has in the references an earlier paper where the DNA of a number of North American extinct equids was sequenced along with Hippidion from Southern South America and the outcome was that there were far fewer species in the Late Pleistocene of North America than had been thought based on morphological studies,

    LeeB.

  72. LeeB Says:

    The earlier paper is by Weinstock et. al 2005.: Plos Biology Aug. 3(8): e241; Evolution, systematics and phylogeography of Pleistocene horses in the new world: a molecular perspective.
    And some of these papers are open access.

    LeeB.


  73. Sometimes extinct taxa are also included in major analyses like Goloboff et al. (2009, http://doi.wiley.com/10.1111/j.1096-0031.2009.00255.x) analysis of eukaryotes, where Smilodon was coded for both molecular and morphological characters (both molecules-only and morphology+molecules analyses placed this taxon within Felidae).

  74. LeeB Says:

    And there is a good paper on lions here:
    Barnett et. al. 2009: Molecular Ecology 18(8) 1668-1677,
    Phylogeography of lions (Panthera leo ssp.) reveals three distinct taxa and a late Pleistocene reduction in genetic diversity.

    Similar diversity loss has been found in living wolves, bison and horses elsewhere; and puma apparently went extinct in North America at the end of the Pleistocene and then recolonised it from Northern South America.

    Molecular methods are proving really useful in studying late Pleistocene fauna from cold or dry deposits.

    LeeB.

  75. Hikaru Amano Says:

    I do agree with many people here that right now biomolecular examinations for ancient animals are still far from fully established. However, scientists still don’t know much about the preservation potentials of different biomolecules(the works of Dr. Schweitzer’s teams have overturned many of the long held assumptions of scientists based on experimental studies, so it would still be too early to say that there is no chance for Mesozoic DNA sequences to survive up to the present and someday be discovered by scientists.

    References:

    San Antonio JD, Schweitzer MH, Jensen ST, Kalluri R, Buckley M, et al. (2011) Dinosaur Peptides Suggest Mechanisms of Protein Survival. PLoS ONE 6(6): e20381. doi:10.1371/journal.pone.0020381

    Schweitzer MH, Wittmeyer JL, Horner JR. (2007a). Soft tissue and cellular preservation in vertebrate skeletal elements from the Cretaceous to the present.Proc R Soc Lond B 274: 183–197.

  76. LeeB Says:

    People have also done preliminary work on sequencing osteocalcin from bones, this may be preserved longer than DNA;
    I have also wondered if the related matrix GLA protein might be trapped in the dentine or enamel of sharks teeth, if so this might make it easier to work out evolution in shark lineages (osteocalcin (the bone GLA protein) apparently doesn’t occur in elasmobranchs).
    However the sequences in these proteins may evolve more slowly than DNA, possibly giving a resolution at above the species level.

    LeeB.

  77. Hikaru Amano Says:

    Resolving the taxonomy of closely-related fossil taxa is quite tricky when one only have the amino acid sequences of proteins. But even if the some of the DNA sequences of their identical genes are available, we may still have a hard time resolving their phylogenetic relationships. Due to the degeneracy property of the genetic code, genes with considerably different consensus DNA sequences may still produce proteins with identical consensus amino acid sequences (hence, they may still be considered as homologous). However, the advantage offered by the conservative nature of the genetic code is that the number of differences in the consensus amino acid sequences of their identical genes’ products will correspond to the minimum amount of differences in their equivalent genes’ consensus DNA sequences.

    Another problem to consider is if some of the taxa concerned have genes with rapidly evolving DNA sequences, increasing the chances that molecular homoplasies and/or convergences may and/or will inevitably occur due to the reversions and/or substitutions. That could and/or would change the amino acid sequences of those genes’ products. In turn, one may encounter the long-branch-attraction phenomenon (LBA; aka short-branch-attraction phenomenon). Hence, we need the biomolecular profiles of older taxa closely-related to the taxa being studied to establish the biomolecular evolutionary developmental trend/s of the clade concerned(again, using the greatest number of samples with different preservation conditions from as many taxa possible).

    One last thing here…I do wonder what the MOR team will say next if somebody performed biomolecular phylogenetic analyses of the taxa they are considering synonymous (but with signifcantly different morphologies) and found out that those animals have consensus biomolecular profiles that are distinct from one another(and in many different groups of animals, closely-related yet distinct species-and at times genera- are separated by as little as 1 % difference in their respective consensus biomolecular profiles[at times even less than that amount]). I don’t think ontogeny could be invoked in such a case since ontogeny only affects which genes get switched on/off at which point of maturity and how much of those genes’ products will be expressed at such phases, but the amino acid sequences of those genes’ products and the DNA sequences of the genes that produce them remain the same.

    References:

    Brinkmann,H., Giezen,M.V.D., Zhou,Y., De Raucort,G.P., and Phillippe,H. (2005). An empirical assessment of long-branch attraction artefacts in deep eukaryotic phylogenomics. Syst. Biol., 54(5), pp. 743–757.

    Brinkmann,H., and Phillippe,H. (2008). Animal phylogeny and large-scale sequencing: progress and pitfalls. Journal of Systematics and Evolution, 46(3), pp. 274-286. DOI: 10.3724/SP.J.1002.2008.08038.

    Chapmann,B.S., Tobin,A.J., and Hood,L.E. (1981). Complete amino acid sequence of the major embryonic β-like globin in chickens. Journal of Biological Chemistry, 256(11), pp. 5524-5531.

    Philippe,H., Zhou,Y., Brinkmann,H., Rodrigue,N., and Delsuc,F., (2005). Heterotachy and long-branch attraction in phylogenetics. BMC Evolutionary Biology, 5(50). DOI:10.1186/1471-2148-5-50.

    Wilson,S.J. (1999). A higher order parsimony method to reduce long-branch attraction. Mol. Biol. Evol., 16(5): pp. 964-705.

  78. Hikaru Amano Says:

    @ Dr. Mike Taylor:

    I think there are also figures of phylogenetic analyses performed by Dr. Schweitzer’s team in the journal articles (see afore-mentioned reference materials) regarding the amino acid sequences of the collagen samples extracted from the bones of Brachylophosaurus canadensis and T. rex, respectively. Also, other scientists have asserted that the preserved soft tissues inside some Mesozoic taxa’s bones are indeed original.

    Reference:

    Lindgren J, Uvdal P, Engdahl A, Lee AH, Alwmark C, et al. (2011) Microspectroscopic Evidence of Cretaceous Bone Proteins. PLoS ONE 6(4): e19445. doi:10.1371/journal.pone.0019445.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 422 other followers

%d bloggers like this: