Say hello to Abydosaurus mcintoshi

February 26, 2010

The hot news on the block right now is the description of the new sauropod Abydosaurus mcintoshi, which, amazingly, is known from four more or less complete skulls (Chure et al. 2010).  This is unheard of — absolutely unprecedented.  There are few enough sauropods for which a skull is known at all; but four of them, all in decent nick, is breathtaking.

And here is one of them, the holotype:

Abydosaurus mcintoshi holotype skull DINO 16488, from Chure et al. (2010:fig. 3)

It’s a real shame that, presumably due to space limitations, this is the only one of the skulls that’s figured in the paper; but the good news is that some of the referred material is illustrated in the supplementary information, which — like the paper itself — is freely available, thanks to the wonder of open-access publishing.

According to the phylogenetic analysis in the paper, Abydosaurus is a brachiosaurid — it is recovered in all MPTs as the sister taxon to Chure et al.’s “Brachiosaurus” OTU (on which, see below).  Since it’s from the mid Cretaceous (Cenomanian-Albian, from the Mussentuchit Member of the Cedar Mountain Formation), it’s likely about 105 million years old, which means it lived the best part of 50 million years after the better known brachiosaurs Brachiosaurus and Giraffatitan.  It was evidently attracted by the Giraffatitan component of the compound OTU, since the skull and neck are effectively unknown in Brachiosaurus (see Taylor 2009 for a review of the holotype and referred material).  Because it lived in pretty much the same time and place as Sauroposeidon, there is the tantalising possibility that it is actually the skull of that animal; on the other hand, the four recovered skulls are all too small to fit the Sauroposeidon holotype, so unless they were all subadult, that appealing idea is probably wrong.

Unlike Giraffatitan — the only other brachiosaur with decent cranial material, so far as I recall — Abydosaurus has narrow teeth , superficially similar to those of diplodocids and titanosaurs.  Chure et al. show that this seems to be part of a general trend of sauropods evolving progressively narrower tooth crowns through time, perhaps because narrow teeth can be replaced more quickly and turnover rate is more important than robustness.

Abydosaurus mcintoshi, reconstruction of skull and anterior neck based on holotype and referred specimens (from Chure et al. 2010:fig. 4). Note your weekly helping of sauropod-vertebra goodness in the upper-right corner, in the form of a transverse slice though cervical 3 just behind the diapophyses.

One aspect of this paper particularly pleases me, and that is that the new species is named after John McIntosh.  For anyone out there who doesn’t know who McIntosh is, he’s been working on sauropods since forever: he’s produced a stream of important papers on the skulls of diplodocids, among many other things, and wrote the Sauropoda chapter in the original The Dinosauria (McIntosh 1990).  All of this in his spare time, mind you, because as his day-job he was a professor of theoretical physics at Yale and Princeton.  He’s probably seen more sauropod material than anyone else alive.  And on top of all that, he is one of the good guys.  I drew the long straw at the Austin SVP in 2007, and got to sit next to him at the informally convened sauropod-workers’ lunch, and it was a revelation to see his face light up as I tried to describe the weird morphology of the as-yet-unpublished vertebra that we now know as Xenoposeidon.  At an advanced age — I don’t know exactly how old he is, but you can get some idea from the fact that he flew over Hiroshima and Nagasaki less than a week after the bombs were dropped — his enthusiasm remains undimmed, and he is truly an inspiring example to every avocational palaeontologist.

So it’s sort of scandalous that it took so long before McIntosh got a sauropod of his own.  (Jensen did name Ultrasaurus after him, but as has been much discussed, that ended up synoymised with Supersaurus).  I know there’s at least one more new sauropod in the works that’s slated to be named after him, and I’m in favour.

Brooks Britt (a co-author on the paper) with one of the skulls of Abydosaurus. Stolen from Science Daily.

A note on brachiosaur taxonomy

I suppose I ought to mention this, only because if I don’t, everyone will just ask me about it.  Chure et al. (2010) refer to Giraffatitan by the old name “Brachiosaurusbrancai throughout, and explain why they do so on page 2:

Taylor (2009) recently suggested that the North American species Brachiosaurus altithorax is generically distinct from the African species Brachiosaurus brancai, which is known from abundant material including a complete skull and many craniodental elements. Based on numerous differences between overlapping parts of both holotypes, Taylor (2009) proposed that the African species should be known as Giraffatitan brancai. While we are open to this possibility, we do not believe that it is sufficiently justified at present because the identified differences have not been defended as separating genera, rather than species, populations, or individuals. The sister-taxon relationship between the two species recovered in the phylogenetic analysis performed by Taylor (2009) neither supports nor refutes their generic-level separation. At this point, we consider the decision to recognize the African species as a genus apart to be arbitrary. We choose to retain the original nomenclature in this contribution, distinguishing between the two species where appropriate.

I am sort of nonplused by this.  I’m certainly not saying that my 2009 paper is unassailable: as soon anyone comes along with evidence that Brachiosaurus and Giraffatitan should after all be considered congeneric, I’ll be first in line to hear them out.  But I do feel that now 26 osteological differences have been described between the species, the null hypothesis has shifted, and the burden of proof is now on those who wish to synonymise the genera.  “We choose to retain the original nomenclature” is not an argument, and doesn’t really advance understanding.  So I’m afraid I think this was a regrettable misstep.

Anyway — I don’t want to end on that note!  The big deal here is that we now have four fantastic new brachiosaur skulls, no doubt to be described in more detail hereafter, and John McIntosh has a beautiful sauropod named after him.  Happy days!

References

About these ads

40 Responses to “Say hello to Abydosaurus mcintoshi”

  1. Mike Keesey Says:

    “The sister-taxon relationship between the two species recovered in the phylogenetic analysis performed by Taylor (2009) neither supports nor refutes their generic-level separation.”

    How could it? The analysis is purely scientific and the separation is purely nomenclatural.

    “At this point, we consider the decision to recognize the African species as a genus apart to be arbitrary.”

    Of course it is! ALL DECISIONS ABOUT GENERIC-LEVEL SEPARATION ARE ARBITRARY.

    Anyway, cool stuff, and looking forward to subsequent analyses with more and finer OTUs.

  2. Nathan Myers Says:

    We may say that McIntosh earned the right to any number of regrettable missteps, but may not get to make them all. Chure et alia, on the other hand, will have plenty of time to regret their rash and inattentive choice. The sooner they are led to regret it, the better for all.

    This cranial abundance would seem to do in the crocosaur-at-the-waterhole theory of sauropod cranium scarcity, unless those crocosaurs that preyed on Abydosaurus carefully collected the crania after devouring the altogether more nutritious post-cranial remnants. (“Step 3: Profit!!!”) The cranial-detonation hypothesis fares a little better, indicating secondary loss of detonation ability in A.

    However, it may simply be that polycephalia in sauropods has been inadequately explored. With sauropods limited primarily by their ability to ingest enough greenery in limited daylight hours, surely the relatively minor investment of additional appropriately terminated cervical structures could multiply their intake rate.

    I wonder if the parallel evolution of dentition reflects changing flora.

  3. Jamie Stearns Says:

    I was actually at the site where Abydosaurus was found, by the way. It’s within sight of the road leading to the quarry building at Dinosaur National Monument, and along with the at-least-four Abydosaurus individuals was a partial Deinonychus skeleton.

  4. 220mya Says:

    Mike – Although you and I disagree about our philosophy on how to treat “genera”, I agree that their explanation for Brachiosaurus/Giraffatitan was a total cop-out, and not really a valid explanation. Mike’s comment says exactly what I thought when I read it. From my point of view, the best reason for justifying the use of Giraffatitan, is that if you put all three brachiosaurs in a phylogenetic analysis (even better – include Sauroposeidon too), your’re going to get a polytomy. With new brachiosaurs cropping up, there just is no longer much evidence that altithorax and brancai are sister taxa to the exclusion of other brachiosaurids.

  5. Paul Barrett Says:

    Four skulls for a sauropod taxon is excellent, but not unprecedented. Don’t forget poor old Camarasaurus, for which there are numerous complete skulls (and also a couple of complete disarticulated skulls). There are also at least three skulls apiece for Diplodocus, Apatosaurus (two undescribed) and Shunosaurus.

    Jack was also honored by the various sauropod volumes in 2005, one of which contains an interview with him.

  6. Mike Taylor Says:

    Paul, you’re right about Camarasaurus, of course: I suppose it feels like Cam has all these skulls because there is just so much material of it that there are going to be skulls in among them by the law of averages, whereas the startling thing about Abydosaurus is that it is, so far, known pretty much only from the skulls.

    The McIntosh interview is one of my favourite parts of the Curry Rogers/Wilson volume.

  7. Jamie Stearns Says:

    Don’t worry about Abydosaurus being poorly-known; plenty of the skeleton has been found and is probably being described as we speak. It’s just a matter of when it gets published.

  8. Aaron Natera Says:

    Could someone please post a link to the “freely available” paper?

  9. Mike Keesey Says:

    I think the concept of distance metrics is probably the most useful one for taxa like genera and species. (What kind of distance? Whatever suits your purposes.) So if species X is closer in your metric to brancai than altithorax is, and you do not consider species X to be part of Genus Brachiosaurus, then you certainly shouldn’t be including brancai in that genus.

    Mike Taylor provided his own metric, considering 26 osteological differences to be more than enough to warrant separation. This is arbitrary, but potentially consistent. Chure & al. don’t seem to offer a different metric, so it’s not much of an argument.

    Insisting that genera be monophyletic is untenable (since that leaves ancestral species with no genus), so I don’t think a polytomy should automatically rule it out. (It should, however, cast doubt.)

  10. Mike Taylor Says:

    Aaron, the paper is linked from the References section in the main post. Don’t tell me you skipped the references? :-)

  11. Aaron Natera Says:

    It’s strange, but my eyes glaze over and I lose track of time and space when I see the word “references.”


  12. [...] coverage of this new taxon over at SV-POW!, and more sauropods to come soon courtesy of Phil Mannion, stay [...]

  13. Matt Says:

    The discovery of Abydosaurus does indeed raises some interesting questions about sauropod phylogeny. It makes me wonder about Titanosauriformes and Titanosaur relationships in general.
    According to Wilson Titanosaurs are most related to Brachiosaurids. However I can not help but wonder if Titanosaurid relationship are more diverse than what we think. Were Titanosaurs really Brachiosaurids that developed Diplodocid like morphology? Or just the opposite were cretaceous “diplodocids” acquiring Brachiosaurid like features? Abydosaurus is a significant discovery that may help shed more light on Brachiosaurid/Titanosaurid relationships. I’m sure this new sauropod will add to the debate.

  14. qilong Says:

    Mike Keesey wrote:

    Mike Taylor provided his own metric, considering 26 osteological differences to be more than enough to warrant separation. This is arbitrary, but potentially consistent. Chure & al. don’t seem to offer a different metric, so it’s not much of an argument.

    This is a great idea, if it’s consistent. When Taylor (2009) was published, I suggested just this method on the DML and (if I recall) on this blog. The consistency issue however becomes complex beyond all reason when you start splitting “characters” or “traits” or when you define the aspect of a functional complex that reduces or inflate the number or value of the “characters” included. I also criticized this concept on the DML. There is no nonarbitrary method yet developed that can determine a “character” per given definition of one in the same way across all specimens in a species, given ontogeny, individual variation (different character, different expression, less value!), functional development, etc. The argument of how to determine a species by number of features becomes the subject of a mad scientist making each explicit measurement variation a “character.”

    There is NO best solution when every method involved in each perspective is frought with arbitrariness. If then the Chure et al. method is correct in arguing that the Taylor method is arbitrary, and one agrees with them, then why support a split off without any other difference? Note that taylor further supports this method by trying to argue about other “brachiosaurids,” but there are horrifically few papers sampling the many basal titanosauriform taxa yet to the degree in which splitting so many “valuable” characters allows us to test Taylor’s argument, or Mike Keesey’s above.

  15. Mike Taylor Says:

    “Arbitrary” does not imply “unhelpful”. We use arbitrary distinctions all the time: they help us to break the world up into think-about-able chunks.

    No-one — NO-ONE — disputes that genus boundaries are arbitrary. Many people would accept that the same is true of species boundaries. The purpose of a classification is be USEFUL, and that property depends mostly on consistency. The most we can hope for in terms of allocating sauropod species into genera is that we have something approaching a consistent degree of granularity — which is why I made the point in the paper that Brachiosaurus and Giraffatitan are more different from each other than Diplodocus and Barosaurus, which no-one has seriously suggested synonymising. (Actually, here in this less formal arena, I’ll go further and suggest that B. and G. differ more than Diplodocus and Apatosaurus.)

    And the real solution, of course, is to abolish one or other the notions of genus and species altogether, so that we don’t even have to have this discussion. Uninomials are the way to go — our current system conflates classification, nomenclature and phylogeny into one big classifinomenclogeny … so it’s no wonder we sometimes end up in a bit of a mess.

  16. 220mya Says:

    Mike said: “Uninomials are the way to go”

    Hey – a taxonomic point we agree 100% on! It would make this whole discussion unnecessary. But then what would you do with higher groupings? Oh, that’s right, place them in clades. Thus why the simple answer is to treat genera as clades ;)

    To the other Mike (Keesey): monophyletic genera are tenable – your objection (this approach can’t deal with ancestral taxa) is theoretical only, because phylogenetic analyses cannot recover direct ancestors. So a direct ancestor will always be recovered as a separate branch. If you use mononomial alpha-taxonomic units, then your “genus” becomes the least inclusive named clade, whether it be Titanosauriformes or Brachiosaurus.

  17. Mike Keesey Says:

    This assumes that all phylogenetic analyses are cladistic. Genetic analyses actually *can* show one species to be ancestral to another. For example, Ursus arctos is ancestral to Ursus maritimus. Of course, in these (fairly recent) examples, the species are congeneric anyway, but it’s still a potential problem somewhere along the line. Just because the road’s smooth now doesn’t mean it’s never going to lead over a cliff.

  18. Mike Taylor Says:

    Randy, your just-treat-genera-as-clades is not tenable, because whether or not we recover ancestor-descendant relationships by numerical analysis, we know that each taxon did have ancestors, and it’s theoretically grotesque to say that they didn’t belong to any genus (which is the only option if we insist on all genera being monophyletic).

    Also: it’s not good for nomenclatural stability to depend on phylogenetic stability. If you strictly insist on monophyletic genera, then you have to redo the alpha taxonomy — i.e. actually change the names of species — every time you do an phylogenetic analysis that recovers a different topology of, for example, the various candidate diplodocid species.

    Example: Hartman et al.’s SVP poster on the WDC [I always want to write WMD] Supersaurus included a phylogenetic analysis that recovered it close to Apatosaurus ajax than the other apatosaur species were. Surely no-one believes that that poster’s communicative value would have been enhanced if it referred to the new specimen as Apatosaurus vivianae, or if it put A excelsus back into Brontosaurus and made up a new interim genus name for the species formally known as Apatosaurus louisae? We should be free to play around with phylogenies without having to mess with nomenclature.

    [In the actual paper on that specimen, Lovelace et al. (2008) did not include the phylogenetic analysis, probably to avoid precisely this problem.]

    In short, Randy, I think you consistently conflate phylogeny with nomenclature.

    Of course the pragmatic solution to this — at least when dealing with clades so enigmatically represented, and therefore phylogenetically unstable, as sauropods, is just to give each new species its own genus, effectively treating the genus+species combination as a uninomial that happens to have a space in the middle. Of all the times in sauropod history when someone has referred a second or subsequent species to an existing genus, it would be interesting to see what proportion of those referrals are upheld today. My guess is that it’s way less than 50%, and quite possibly less than half that. Why even go there? Keep the suckers separate and then swirl your phylogeny around with impunity. Uninomials by stealth!

  19. 220mya Says:

    To the Mikes:

    1. I am not conflating nomenclature and phylogeny (I’m well aware of the differences), I simply want my nomenclature to reflect phylogeny (hence the phrase “phylogenetic nomenclature”).

    2. Just because someone produces a “cladogram du jour” does not require people to revise the nomenclature. As with any system, revising nomenclature and taxonomy require judicious and well-reasoned decisions. There is no taxonomic philosophy out there that is immune from poor taxonomic decisions.

    3. I reject the sentiment that applying phologenetic principals to alpha taxonomy is inherently less stable than any other system. In some ways the idea of what a “genus” might be is more stable, because its an explicit definition – sure the content might change, but content has been changing continually since the days of Linnaeus.

    4. No one said you only had to use node-based definitions for genera. Branch-based definitions can include the common ancestor.

    5. I don’t understand why it is “theoretically grotesque” to not have a genus associated with a species. That’s your subjective opinion, and such a solution only seems unreasonable because it is so different from the binomial system we have today.

    6. Mike T.: you still haven’t addressed how placing a uninomial of genus+species in a taxonomy of higher clades is any different from placing a uninomial of just a species (or whatever level you want) in a taxonomy of higher level clades, one of which happens to be a genus.

    7. The reason why a uninomial of genus+species is not preferable is that it is redundant. There are a couple of fine proposals for converting species names for unique names. Genera have always been considered more inclusive than species, so I would prefer not to make them part of a uninomial alpha taxonomy.

    What should be made clear here is that my “cladistic genus” is only a band-aid if we’re forced to retain the Linnean binomen. My preference is to have a uninomial alpha taxonomy, with a rankless phylogenetic taxonomy covering all higher-level taxa.

  20. Mike Keesey Says:

    “What should be made clear here is that my “cladistic genus” is only a band-aid if we’re forced to retain the Linnean binomen. My preference is to have a uninomial alpha taxonomy, with a rankless phylogenetic taxonomy covering all higher-level taxa.”

    Hear, hear!
    I think we all agree on that point.

  21. Mike Taylor Says:

    “What should be made clear here is that my “cladistic genus” is only a band-aid if we’re forced to retain the Linnean binomen. My preference is to have a uninomial alpha taxonomy, with a rankless phylogenetic taxonomy covering all higher-level taxa.”

    Hear, hear!
    I think we all agree on that point.

    If by “we all”, you mean you, me and Randy, then yes. If, however, want to extend the definition of “we all” to include the trifling matter of the other 6.8 billion people out there, I fear we may not find unanimity so easy to attain :-)

    No one said you only had to use node-based definitions for genera. Branch-based definitions can include the common ancestor.

    Work it through: use whatever combination of branch-based and node-based clades you want, you can’t partition up the tree in such a way that every individual belongs to exactly one of those clades; so those clades can’t function as genera as genera are understood by, well, everyone but you. Bottom line is that we have three structural desiderata for genera: that each individual belongs to a genus; that no individual belongs to more than one genus; and that each genus is a clade. It’s not possible to fulfil all three of those at once, so any system has to pick (at most) two; and since several hundred years of prior art mandate the first two, the ariviste new gotta-be-a-clade desideratum is the loser.

    I don’t understand why it is “theoretically grotesque” to not have a genus associated with a species. That’s your subjective opinion, and such a solution only seems unreasonable because it is so different from the binomial system we have today.

    You have to have some respect for prior art. All the rank-based codes require that every species belongs to one and only one genus. If you violate that, then what you’re doing won’t be recognised as alpha taxonomy. I realise that you don’t have any interest in perpetuating rank-based alpha taxonomy, but if you do something as fundamentally different from it as having indivduals and species that do not belong to any genus, then don’t be surprised if other taxonomists look at you in a funny way (and refuse to publish your papers).

    you still haven’t addressed how placing a uninomial of genus+species in a taxonomy of higher clades is any different from placing a uninomial of just a species (or whatever level you want) in a taxonomy of higher level clades, one of which happens to be a genus.

    There’s no problem with placing species or genera (whether uni- or binomial) in clades — none at all. The difficulty arises when the groups that you’re placing them into are genera, and that is because of the requirement that the set of genera partition the tree.

  22. Mike Keesey Says:

    Everything that Mike Taylor just said.

    (If only English had different forms of “we”, as some languages do….)

  23. Nathan Myers Says:

    Another alternative is to give up pretending that genera have the first thing to do with clades. Genera weren’t confused with clades in Linnaeus’s time. It was only when they began to be conflated with actual descent that the confusion really set in. The fundamental reason for conflict (beyond healthy scientific disagreement) is that accurate mapping between actual descent, cladistic approximation, and nomenclature will never be stable, but instability is inherently incompatible with the human mnemonic apparatus. Give people stable names to remember, and cladistic assignments and re-assignments harm nobody, so can waffle about entirely on their merits.

    I’m disappointed to find no one participating in the sauropod polycephaly revolution. We should have a four-necked Lernaean reconstruction of A. mcintoshi to admire.

  24. qilong Says:

    Mike Taylor wrote:

    “Actually, here in this less formal arena, I’ll go further and suggest that B. and G. differ more than Diplodocus and Apatosaurus.”

    Without invoking the argument of “number of differences” or numerical distance values between two “species” (which have nothing to do with the metric that differentiate either explicitly from other taxa as autapomorphies are not phylogenetically informative), can you tell me why the metric used to argue one taxon is NOT the same as another taxon can be used to explicitly distinguish another pair altogether?

    In addition to this, how far down can you split a complex or suite of features (something I brought up in my reply) and be reasonable about it? If I lump one suite too much while splitting another, am I being biased?

  25. Mike Taylor Says:

    Jaime, I don’t understand your question.

  26. qilong Says:

    This encompasses only the statement about differentiating “diplodocid” versus “brachiosaurid” “genera.”

    Specifically:

    1. Do you think using autapomoorphies (your reason for differentiating brancai from altithorax at the “genus” level) is adequate reason for defining taxa from one another?

    2. If so, do you use a baseline metric that allows you to differentiate a point or a grade where one cannot call one “species” as so different as to be a “genus”?

    3. And finally (adding on to the above) what prevents you from using this to affirm Linnaeist ranks? You argued that brancai and altithorax should or could clade differently, but this is untested and is also true of every other species complex (including all species of Apatosaurus).

    In this, you argue that you can find more differences in “Brachiosaurus” than in “Apatosaurus,” but this is a relativistic comment that is not, to my knowledge, verified by actual data.

  27. Mike Taylor Says:

    1. I used morphological differences between Brachiosaurus and Giraffatitan to separate them, not autapomorphies. Until we have a phylogenetic analysis that includes more brachiosaurs, we won’t know which of those characters are autapomorphic, or synapomorphic, for which clades. But since genera are a different kind of thing from clades, that’s not a problem. Separation is by morphology, not phylogeny.

    2. Of course there is no objective metric that can be used — everyone knows and understands that. The best we can hope for is consistency within a broader group. Everyone would love to have numeric methods that we can put observations into, turn a handle, and get an objective pronoucement on congenericity; but we have to accept that it is not going to happen: alpha taxonomy always has required, and always will, experience of the group in question, taste and judgement.

    3. Once more, I DO NOT CARE whether Giraffatitan and Brachiosaurus clade together or not (especially since whether they do is almost entirely a question of which other brachiosaur OTUs are included in the analysis at hand). That doesn’t mean that I “affirm Linnaeist ranks”; it means that then I am working within the framework of rank-based nomenclature, which is the only way to do alpha taxonomy, I use appropriate criteria.

    You argue that you can find more differences in “Brachiosaurus” than in “Apatosaurus,” but this is a relativistic comment that is not, to my knowledge, verified by actual data.

    Ah, the argument from ignorance. Never very convincing.

  28. qilong Says:

    Mike Taylor writes:

    Ah, the argument from ignorance. Never very convincing.

    I tend to be careful on how I phrase things. Occassionally I jumble it, but the cautionary language is there and prevalent. This is included in the quoted statement. It has NOT been published, as the theory you ascribe to for brancai/altithorax has not been applied to virtually any dinosaur grouping since Paul (1988) applied this in his article on sauropods and his book, and later by others following Paul’s typological argument that mere differences qualify to taxonomuic value for species — or valuable as your paper argues as rank-able variation. Your use of what might be called sheer-numbers-of-differences is itself a “species concept” that you did not apply, or show how it was applied by others, save the above cited works, and every time I’ve pointed this out, you seem to give some snarky comment like above rather than respond to it. Telling or no?

  29. qilong Says:

    1. I used morphological differences between Brachiosaurus and Giraffatitan to separate them, not autapomorphies. Until we have a phylogenetic analysis that includes more brachiosaurs, we won’t know which of those characters are autapomorphic, or synapomorphic, for which clades.

    The point I was going to lead in here was different, but this is a decent start. You had the opportunity to include OTHER brachiosaurs in the analyses, in a high profile paper — why didn’t you, and are you working on a more extensive analysis to test your hypothesis?

    However, I was attempting to drive at the point that simply rating proportionate variation as distinguishing taxa i9s generally an unresolved issue as there are ecological or potential sexually dismorphic reasons why some proportions/features exist, so where you have robusticity or proportiosn differ, you inflate the features differing between two small sets of specimens. If the features that differ are fundamental, relating to expression of morphological details such as extra lobes on bones or such in odd places, or a peculiar arrangement of fossa, or an inversion of a typical proportional cline, one could argue the features are fundamentally different enough, but critical evaluation of the features seems to be absent. You just counted them up, like Paul did in 1988 in PDW and his sauropod paper, and this is essentially typological in nature (it only seeks to find “similar” and “different” as qualifying taxonomy, like how man cannot be an ape because we’re so different morphologically).

    But since genera are a different kind of thing from clades, that’s not a problem. Separation is by morphology, not phylogeny.

    This was another point,. You endorsed Linnaean taxonomy in the paper by framing a genus, and a species. If they are really that different, why can’t we have different Families? The problems with ranks are rampant.

    2. Of course there is no objective metric that can be used — everyone knows and understands that. The best we can hope for is consistency within a broader group.

    I argued earlier that you didn’t affirm your own method in your own paper, just used it. You didn’t compare the method by calculating the differency-index or whatever calibrated by the individual-variation-index (or whatever) to assure yourself that simply counting up differences was even a useful tool in distinguishing species into monotypic genera.

    Everyone would love to have numeric methods that we can put observations into, turn a handle, and get an objective pronoucement on congenericity; but we have to accept that it is not going to happen: alpha taxonomy always has required, and always will, experience of the group in question, taste and judgement.

    And now who’s arguing from ignorance? It’s the ultimate shrug: There’s numbers involved, I don’t know how it works, but hey, it’s something.

    3. Once more, I DO NOT CARE whether Giraffatitan and Brachiosaurus clade together or not (especially since whether they do is almost entirely a question of which other brachiosaur OTUs are included in the analysis at hand). That doesn’t mean that I “affirm Linnaeist ranks”; it means that then I am working within the framework of rank-based nomenclature, which is the only way to do alpha taxonomy, I use appropriate criteria.

    I’m sorry, but NOTHING in “alpha-taxonomy” requires you to use Linnaean ranks, and it does not require you to treat a value of varying features as relevant to a “rank.” This is another topic where number of features = rank up is not supported in the literature by, say, testing, as it is an abstract concept. Moreover, the value of a rank depends on the taxa within it, which do NOT depend on the rank, and yet they are treated as theough they do by the ICZN. But even the ICZN doesn’t say how you determine what a “genus” or “species” is. Now, your argument as noted above seems to imply you ARE required to use ranks. I refer you to any of the works of Flynn et al. in Nature or JVP wherein Madagascan fossils are described without application of ranks, and so far, the taxonomies remain unchallenged.

  30. Mike Taylor Says:

    The problems with ranks are rampant.

    I am running out of different ways to say “I, too, am not fond of ranks, but use them because that is what convention demands.”

    The only Flynn et al. paper I could find on Madagascar was the 1999 report “A Triassic Fauna from Madagascar, Including Early Dinosaurs”, which does not name any new taxa.

  31. Jay Says:

    On a slightly different matter, but since it was mentioned, can i draw your attention to the 2004 Upchurch et al monograph on Apatosaurus ajax, which did infact attempt to test the species-level relationships of species & specimens of Apatosaurus.

  32. Mike Taylor Says:

    Upchurch et al. (2005) [not 2004, though it does sometimes get cited that way] is an invaluable paper — I refer to it all the time. But because its specimen-level phylogenetic analysis didn’t include Supersaurus or Suuwassea, it doesn’t really tell us anything about the monophyly of Apatosaurus as traditionally conceived.

    [Of course at the time that work was done, Suuwassea hadn't been described and no-one had any clue that Supersaurus might be apatosaurine -- everyone assumed it was closely related to, or even synonymous with, Barosaurus (Curtice 2003), so that is not in any way a criticism of Upchurch et al.]

    Right now, I think that Apatosaurus monophyly is one of the more interesting outstanding issues. But please note, all you genera-must-be-monophyletic advocates out there, that even if it turns out that (say) Supersaurus clades with Apatosaurus ajax, then the element of judgement that we’d all like to eliminate will still be necessary: someone would have to decide whether to sink Supersaurus as a species of Apatosaurus, or whether to blow Apatosaurus apart, reverting A. excelsus to Brontosaurus and raising a new genus for A. louisae. In short — to repeat myself, I know — you CANNOT escape from the need for subjective but informed taste when doing genus- and species-level taxonomy.

    REFERENCE

    Curtice, Brian D. 2003. Two genera down, one to go? The potential synonomy[sic] of Supersaurus with Barosaurus. Southwest Paleontological Symposium 2003, Guide to Presentations, Mesa Southwest Museum, January 25 2003, unpaginated.

  33. Jay Says:

    ‘Upchurch et al. (2005)’ – In the FM, one page before the contents, “Published on 28 December 2004″

  34. Mike Taylor Says:

    I know it says it was published in 2004, but it wasn’t. I have this from Paul Upchurch, who ought to know.

  35. Jay Says:

    Fair enough. I thought you might say so, but only after I posted that damn comment.
    And fair call about apatosaurine phylogeny, Upchurch et al certainly did make some excellent inroads into figuring out species interrelationships of Apatosaurus, but as you allude to, this area of research is ripe for alotta further work.


  36. Nathan wrote:

    I’m disappointed to find no one participating in the sauropod polycephaly revolution. We should have a four-necked Lernaean reconstruction of A. mcintoshi to admire.

    Mr Myers I think your hypothesis worthy of note (and I had thought ‘Hydra’ before I saw ‘Lernean’ – oddly, I had thought of The Labours of Hercules earlier today, specifically, the Cleansing of Augean Stables – but I digress); however I wonder if you may have it barse-ackwards, since I fail to see how your hypothesis of polycephaly would explain rather than exacerbate the encephalisation deficit, whereas MY new theory which I hear unveil – this is it – which it is – is that sauropods budded off their young, sprouting tail-first a-dangling from the adult’s head, and dropping when ripe, thus explaining the elongated (stretched) neck as a developmental not genetic novelty. The head of the young did not form immediately (nor always) but they filled an ecological niche as blind, snuffling suck-uppers of tasty primeval soup: algal suspensions in shallow lakes, etc. Only later in life (and if the soup ran out) did they mature to full headhood: metamorphosing neocranially to a toothed, eyed form and thus adopting a new habit of browsing on the foliage which they only then began to perceive growing so lushly overhead.


  37. ..which I here unveil.. (I meant to write).

    Um, we could call my theory “Facultative Neo-Cephalisation”,

    or – for short – New Faculty Head.

  38. qilong Says:

    Mike, it is in fact

    Flynn, J. J., Parrish, J. M., Rakotosamimanana, B., Simpson, W. F. & Wyss, A. R. 1999. A Middle Jurassic mammal from Madagascar. Nature 401: 57-60.

    http://www.nature.com/nature/journal/v401/n6748/full/401057a0.html

    Of course, it is a note, as well. Is this what you’re thinking of?

    I should not need to remind anyone why “convention” is a fools’ paradise when it comes to following the pack and not following what is or is not the most scientific way. Ranks are meaningless without taxa, and are only defined by the positions for the taxa that they are assigned to. This isn’t the only paper on the subject of naming taxa (and practicing the method) without assigning ranks. No one has yet to legitimately challenge Ambondro mahabo‘s exisitence.

    Flynn has also published in Novitates, PNAQS, and JVP and avoid3ed the aweful burden of ranking his taxa. It does not mean that other people cannot rank them for him, in their own schema, but it does not mean that that is his intended preference. mThere are other works that ascribe the method of phylogenentic taxonomy without ranks, and it is not hard to conceptualize rankless taxonomy (to me, at least). It makes issues like Brachiosaurus vs Giraffatitan a lot easier than to create things like monotypic classes or orders for birds or squamates or insects or whatever.

  39. Mike Taylor Says:

    I don’t have access to that Flynn et al. article. Can you shove a PDF my way?

    “I should not need to remind anyone why “convention” is a fools’ paradise”

    On the contrary: convention is what enables us to communicate.


  40. [...] to decide over the next few years.  I was disappointed to see that Chure et al. (2010), for poorly explained reasons, rejected this name; and I was pleased to see that Sander et al. (2010) accepted it.  But only [...]


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 401 other followers

%d bloggers like this: