Anchiornis: a new basal avialian from China

Woo-hoo I’m on a new dinosaur paper. I have got plenty of papers done (I think) but I have yet to be on one naming a new dinosaur so I’m really quite pleased about this, even if I didn’t actually do too much on the paper. This one is coming out in the Chinese Science Bulletin but the paper is already up online so I’m happy to shout about it now before the thing is formally published in the January issue.


As you can see from the picture there is a fair bit of this guy (arms, legs and a partial vertebral column) and with a few feather impressions too, oh and it’s a juvenile, or at least not quite an adult, too. (a) is the main plate, and b) the ‘negative’ counterplate.

Onto the beastie itself and what it means for dinosaur evolution. The full name is Anchiornis huxleyi which roughly means ‘near bird’ or ‘close to birds’ of Huxley (by the way it’s pronounced ‘an-key-ornis’, not ‘anch-ee-ornis’). The fossil is another from the Liaoning beds and thus is the latest in a very long line of important Chinese fossils that relate to the origin of birds (though we don’t know exactly how old this one is). In this case analysis shows that Anchiornis is a basal avialian, that is to say it is on the lineage that leads up to birds, but not actually a bird itself. Fortunately there is a really good reference point for this: Anchiornis comes out in the phylogenetic tree immediately basal to Archaeopteryx, so you can see why it is potentially very interesting and important despite its relative incompleteness as it further narrows the gap between birds and dinosaurs (as if that was not narrow enough already).

The phylogenetic analysis that produced this result is itself quite a big step from previous ones that dealt with this part of the dinosaurian tree. (Unfortunately I have yet to get to my posts on cladistics, so my apologies if I lose a few people at this point). We have included a bunch of new characters and especially those with very fine distinctions in the character states (i.e. small differences applied to continuous characters). This of course risks the problem of character atomisation (over-splitting characters until they are no longer independent and thus cause problems with pseudoreplication of data in the analysis) but the results of the analysis suggest that this has not happened here (at least not yet) since the results are fairly well resolved all things considered. However, it is a reasonable approach with such a group where the number of taxa is building up rapidly with only fine distinctions available between them, and more reversions and convergences appearing.

The presence of Anchiornis (and other new taxa added to the analysis and the new characters) somewhat inevitably narrow the gap between birds and other theropods quite literally by having a number of characters in intermediate states between the two. Importantly it has a highly derived wrist that allows great flexion (or more accurately abduction) allowing the hand to be folded against the lower arm, which is thought to be an important precursor to flight. This is, of course, to be expected from something which is clearly closely related to Archaeopteryx, but the wrist is such an important feature in bird evolution this is an important discovery.

Although this specimen might appear to be a little disappointing, I can tell you that already new material has been recovered that is far better preserved and far more complete and there are several of them. Work is actually at a pretty advanced stage on this so expect a lot more on Anchiornis in 2009 or 2010. I really should add that I did very little on this paper, having been brought onto the project quite late in its development, but I am delighted to have been involved though I have just noticed that my middle initials are missing from the reference even though I did point this out twice. Oh well. Feel free to fill them back in if you wish.

Xu, X., Zhao, Q., Norell., M., Sullivan, C., Hone, D., Erickson, G., Wang, X., Han, F., & Guo, Y. A new feathered maniraptoran dinosaur fossil that fills a morphological gap in avian origin. Chinese Science Bulletin (in press).

37 Responses to “Anchiornis: a new basal avialian from China”

  1. 1 Mike Taylor 23/12/2008 at 11:21 pm

    Well, congratulations! Nice Christmas present.

    So what WAS your contribution?

  2. 2 Nick Gardner 24/12/2008 at 12:36 am

    Why was the phylogenetic analysis data not included in a supplement online?

  3. 3 Zach Miller 24/12/2008 at 2:26 am

    Congrats, sir! Bill Parker sent me the paper a few days ago. I’m confused by one point, though. In the abstract, as well as the rest of the paper, the authors constantly repeat that Anchiornis is a non-avian theropod dinosaur, but then they place it in Avialae, which has traditionally been thought of as BIRDS.

    So are the authors must be redefining Avialae, but they never state that they are.

  4. 4 David Hone 24/12/2008 at 9:10 am

    Mike: I did some stuff on the lenght and mass estiamtes and helped fix some of the characters in the analysis, plus hunted down a few odd references. As you know this takes a lot of time, but it was hardly a major contribution.

    Nick: Sorry, no idea. The paper was destined for Nature till a relativly last minute change where we lost out, I don’t know why the MS went through either without the analyiss included in the appendix or online. It is freely available from Xu if you want it though.

    Zach: that all depedns on how you define ‘birds’ in the first place. If you look at the phylogeny in the paper, Aves is marked at the node that gives us Archaeopterux, thus anyhting below that is *not* a bird (including Anchiornis) and everything above is. Thus Anchiornis is an avilain (it’s on the stem lineage to birds that is separate from the other maniraptorans, but not a bird itself). Plenty of people would call any avilaina a bird, and plenty would call only things more derived than other stem avialians (including Archaopteryx) birds. It is entirely dependendt on your defintiion and soemthignt hat is by no means fixed. We make it clear what we call Aves (which let’s face it is pretty much synonymous with birds, though it depends *where* you put Aves) and Anchiornis is not in that group.

  5. 5 Brad McFeeters 26/12/2008 at 2:57 am

    Two big questions that I don’t see addressed in the paper:
    1) Is the age of the Anchiornis holotype merely not settled yet, like many other feathered dinosaurs when they were first described, or was that information lost forever due to the collecting history of the specimen?
    2) Why no mention of Epidexipteryx, which was also just described as the most basal avialan by some of the same people? What happens when Epidexipteryx is entered into the Anchiornis paper’s matrix, and vice versa?

  6. 6 David Hone 26/12/2008 at 9:48 am

    1. I think we just don’t know. A great many of these specimens come to us through local farmers / collectors and although we always ask, they often do not know where they got it from. The specimens often change hands several times as well before they reach us, so information is unreliable. As I said there *are* other specimens and my understanding is these were collected by a proper field team and so we should be able to work out the age of these ones at least.

    2. It is in the context of the analysis (i.e. since Epi. is not in the Anchiornis paper, it can’t really come up as more basal or more derived than anything). If you know how papers work (especially with Nature) the reason is basically that the Anchiornis paper was completed long before Epi. was. We couldn’t include that taxon until the preparation / research was completed, but then they couldn’t include the Anchiornis stuff as it was unpublished. Just bad timing in terms of how the two papers were prepared and published. In either case, Epi. technically can’t be the most basal avialian since it’s not on the stam linegae, it can be the most basal scansoriopteridge and this clade can be the most basal avialian clade, but Epi. itself cannot.

  7. 7 David Hone 26/12/2008 at 9:49 am

    Oh, and as for what happens with the analyses, I simply don’t know. That takes a long time to sort out, and while I am interested, I have neither the time nor motivation to do it myself. It will, I am sure, be handled by one or the other group shortly. My guess is that Anchiornis will appear above the scansori-s given that basically it looks far more like Archaeopteryx that they do (though statements like that can be dnagerous assumptions, I think it’s a pretty safe bet).

  8. 8 Alexander Vargas 06/01/2009 at 9:11 am

    Hi David
    Can you please give me these measuremets for Anchiornis?
    Femur length
    Humerus length

  9. 9 David Hone 06/01/2009 at 2:24 pm

    43.2 and 41,5 mm respectively.

  10. 10 David Hone 06/01/2009 at 3:07 pm

    I just want to get in here with an extra quick note. It seems that the media is starting to pick up on the Anchiornis story and thus it might well make a few newspapers / websites over the next few days. I would therfore like to predict well in advance that each of the following will happen in at least one article and probbaly all three in at least one:

    1. Anchiornis will be called a “missing link”.

    2. Anchiornis will be called a “bird ancestor”.

    3. Anchiornis will be spelled wrong, the species name will be capilalised and / or the name will not be italicised. (And yes I know I haven’t done that becuase I can’t do the tags in the comments).

    *None* of these things should happen, probably all of them will.

  11. 12 Manu 08/01/2009 at 7:28 am

    4. ‘Anchiornis’ will be claimed as a blow to evolutionary theory. Someone will manage to do it…

  12. 13 Nathan Myers 14/01/2009 at 5:15 pm

    Even I’m impressed.

    What exactly is wrong with calling this a bird ancestor? Has it lost some feature that is present in modern birds, indicating it’s not actually ancestral, but an offshoot? Or just that nobody can really say whether it’s actually ancestral or not, and so shouldn’t? Given the limits of newspaper vocabulary, what should they say instead?

  13. 14 David Hone 15/01/2009 at 6:01 pm

    Ancestors cannot pretty much by definition be identified so we don’t recongise them as such. They share the most recent common ancestor, but one is not ancestral to the other, to do so it would have to be on the stem on a lineage, whereas of course it does not. An acceptable phrase would be something like ‘shared a very close relationship to’ or ‘shared an ancestor with birds’.

  14. 15 Nathan Myers 18/01/2009 at 7:01 am

    That’s pretty weak. They may just as well say it “might have been an ancestor to all birds”.

  15. 16 David Hone 18/01/2009 at 1:19 pm

    It’s not weak at all, it’s correct.

  16. 17 Nathan Myers 18/01/2009 at 5:19 pm

    And my alternative above is, somehow, less correct?

    You can feed journalists awkward constructions that they (or their editors) will feel obliged to alter and get wrong, and then complain about them; or you can express results so that it’s less work for them to get it factually right than to mess it up. There’s no reason why what they print has to match what you would put in an abstract. It suffices if it’s informative and catchy.

  17. 18 David Hone 18/01/2009 at 6:41 pm

    But not if it’s worng. It IS NOT and ancestor. SO to even say it might have been is wrong. Is that really so hard to understand. I can feed them awkward constructions if it is factually correct – I notice it never bothers good writers like Carl Zimmer. There are plenty of things that can be said that are accurate and helpful “it gives us an idea of what a bird ancestor probably looked like”, “it’s the closest thing to an ancestor of birds”, “it was an excpetionally close relative of the first bird”, “it was immediately related to the evolutionary line that led to birds”. All of these get the point of bird origins across, but none are complex, or hard to understand or call it something it is not.
    Once more in these endless debates I feel it is the job of the journalist for getting the report right, you can’t criticse us if they chose to change our words and phrases from right ones to wrong ones.

  18. 19 Nathan Myers 19/01/2009 at 2:26 pm

    It all depends on whether you want what’s published to be right, or you just want be sure all the mistakes published are somebody else’s. Me, I’d rather they get it right, and I’d rather do what it takes to help them get it right. Complaining afterward doesn’t do anybody any good, not even you.

    Not to belabor the matter, but I haven’t seen any hint of evidence that it isn’t ancestral to modern birds — and I did ask. Sure, it’s not especially likely, but it’s about equally as likely to be an ancestor as any other specimen. If you can identify anything about “might be an ancestor” that is actually incorrect, I’m all ears.

  19. 20 Zach Miller 19/01/2009 at 2:52 pm

    Well, we weren’t there to see the transition take place. Even today, there are very few instances where we can point to one creature and say “A evolved from B,” or “X is the ancestor of Y.” Ancestors in the fossil record are necessarily hypothetical.

    I think most journalists do a horrifying job of getting their facts straight in news articles. I’ve seen way too many stories calling pterosaurs “flying dinosaurs” or dromaeosaurs “missing links.” Explain to me how Microraptor is a missing link between T.rex and a swan. Please, tell me how that works. In all cases, a simple Google search or Wikipedia check would tell the journalist all he/she needs to know to get the story right. Pterosaurs are, in fact, NOT dinosaurs. Neither are saber-toothed cats or plesiosaurs. You don’t see journalists calling marchairodont cats “dinosaurs.” I expect the same low level of journalistic excellence regarding non-crocodilian archosaurs.

    And it’s not like you have to a trained scientist to understand the difference. As you know, Nathan, I do this on my spare time, yet I figured out that very basic fact more or less on my own a fifteen years ago. And that was BEFORE Google or Wikipedia. With virtually any piece of information a mere click away, I don’t think there is any excuse to get very basic facts about dinosaurs (and other prehistoric critters) right.

    Back to the original question–Anchiornis cannot be directly ancestral to Archaeopteryx because it has its own suite of unique characters, among them a pitted coracoid and metacarpals II and III, which appear to be fused at their proximal ends. These features are not present in Archaeopteryx, so Anchiornis is not a suitable ancestor.

  20. 21 David Hone 19/01/2009 at 3:50 pm

    Thanks far that Zach but what Nathan fails to notice despite my explicit staement about it is that we fundamentally DO NOT identify ancestors as part of phylogenetics. It is something that is not done, full stop, by default. Thus it is irrelevant if there are or are not characters, combinations of characters or anyhting else going on that would possibly identify Anchiornis, or anyhting else for that matter, as an ancestor, no researcher would even call it one anyway. It is therefore, scientifically not an ancestor, so we are not calling it one.

  21. 22 Nathan Myers 20/01/2009 at 2:35 pm

    Thanks, Zach. Your last paragraph is an actual answer to my original question. Even your typical harried news editor would have no difficulty following it.

    It’s completely obvious that nobody can demonstrate that some random fossil *is* an ancestor of some modern creature. Repeating the fact at length enlightens nobody, but insults some.

  22. 23 David Hone 21/01/2009 at 10:37 am

    I’ll assume tht is directed at me. I was not jsut trying to blindly rpeat a fact but make a perfectly valid point which it that it is irrelevant if Anchiornis has distinct charaters or not, even if it had only features that *could* mean it was a bird ancestor, no professional would call it one since we cannot and do not identify ancestors or talk about them in this manner. As a result, we would not want anyone else to call it one (i.e. the newsies) because this would be a odds with how we do our science. To do otherwise would be incorrect and inappopriate, hence my statements to this effect.

  23. 24 Nathan Myers 21/01/2009 at 12:11 pm

    You’ve explained all of this at length, several times. I understand it perfectly. However, it’s beside the point.

    Phylogeneticists are carefully conditioned not to be interested in direct ancestry. Normal people — readers of newspapers — are keenly interested. You consider the details that make it not-quite-ancestral to be trivial distractions. Newspaper readers want to know those details. They get frustrated at being teased with facts that means nothing to them, while the (to them) meaningful details are nowhere to be found.

    Yes, superb science writers can get it right without losing their audience. What of the rest? You can excoriate them for trying as well as they can to construct something that will inspire their readers, or you can give them what they need so that they can inspire without getting it wrong. Details that affect your career belong in your own paper. Details about the actual animal belong in the newspaper. They are not the same.

    The odds are overwhelming that no specimen will be on the direct line of ancestry to modern creatures; if one was, we wouldn’t be able to determine it. Nevertheless, even normal people know there is such a line, and they are interested in where a new species lies in relation to that line. That you don’t think they should be interested in such trivialities is your problem. What matters is that they are interested at all.

    Biology has a unique place in Science. It is able to inspire, to engage the enthusiasm of people from all walks of life. That power can lead them to be willing to pay for all manner of scientific investigation. Squander it, and they mutter, instead, about dusty, arrogant eggheads. Which seems better to you?

  24. 25 David Hone 21/01/2009 at 3:41 pm

    But I never said any of that. All I said was that I expected at least one paper to incorrectly report that it was an ancestor. I expected this becuase long experience tells me that they will say that even if you explicitly say otherwise. I think your opinions of the avergae readership and indeed the average journalist are far too high, not that they are stupid, but specifically that they do not know about ancestors and ‘lines’ and where things fall on those lines (or otherwise) and they are simply not that interested. You are and you project this onto others, but even my family with a great iunterest in biology and my career research do not care if it was an ancestor or not.
    The journailists tend to keep it very simple indeed as a result (hence the over simplification that leads to things being called ancestors). Long experience with journalists (both my own private work with the BBC as well as those of many colleagues) tells me that no matter how careful one is, and how much you explain in how uch detial, or how simple you keep it, mistakes of this kind are all but inevitable. The journalists (in my experience) will oversimplify from ‘close to the ancestor’ to ‘the ancestor’ or even fail to spot the distinction, even if you stress it.
    We care because it is important to us and we want it passed on, but we cannot control it. My endless lack of patience comes from their inability to take what we say and maintin the interity of it. Their fundamental job is to take our words and make them interesting and exciting and understandable for the public. But, this must be done without *changing* the meaning of those words, and in this the majority fail. This is not our fault, it is teir responsibility. We know abiout it and take steps to avoid it, which is why it is so frustrating when it happens so often and so predictably.
    The journalist is there (in the case of science reporting) to bridge the gap and when done ably (as it can be) this does indeed inspire and inform, when doen badly it does not. You seem to insist that this is the faulyt of the scientist, but their job (primarily) is to do research, the journalists are there to pass it on. If this is done badly, I fail tos ee why the scientist must ake the blame. They can be clearer sure, but surely the journalist is there with the experience and skills to extract the information properly. This is what they are there for! What frustrated me as I have often said is where simple mistakes are made through the journalist imposing his icnrrect will on simply science pterosaurs becoem dinosaurs, relatives become ancestors, and so on. We do not make those mistakes, we do not even often leave windows for tose mistakes (I now always explicitly write that pterosaurs are not dinosaurs for these kinds of exchanges) and yet they are still made. Don’t blame us, we are not doing it. These are not dry arrogant eggheads these are reporters making mistakes that if they had more training (i.e. science background) they would not make and I would not get mad and the public would be better served.

  25. 26 Nathan Myers 03/02/2009 at 3:04 pm

    I’m not interesting in blaming anybody. I’m interested in having an informed, engaged, and inspired public. I’d rather there be nothing to blame anybody for. Complaining about, or to, reporters doesn’t achieve that.

    When reporters change statements, they’re changing them to be more interesting to (who they imagine to be) their readership. If they change “relative” to “ancestor”, it’s because they think that will engage their readers. Likewise if they change “pterosaur” to “dinosaur”. If the stories were engaging enough “as is”, they wouldn’t be motivated to change them. They typically have six or eight other stories they’re supposed to be finishing at the same time, so if anything can save them the trouble of rewriting, they will take it eagerly.

    I think we both know that you are not personally equipped to present material in such a way that a reporter won’t be tempted to violate it. That just means the work needs to be filtered through somebody else who has demonstrated such a skill. That person is almost certainly not a reporter or a publicist.

    One of the key differences between scientific presentation and public presentation is that the reported result typically has, in science, some esoteric, intolerably dull, and extremely temporary consequence, such as firmly negating an academic rival’s obviously insane proposition (e.g., azhdarchid skim feeding).
    What engages the public, instead, are the actual animals, how they fit into the natural world, how they lived, what they looked like, and whether they could swallow oneself or one’s offspring whole.

    What makes pterosaurs interestingly different from dinosaurs, for example, is not that they are neither saurischia nor ornithischia, but that their lineages split off, and evolved independently, for eons before any of the dinosaurs that readers ever heard of existed, and then were driven to extinction once dinosaurs finally got around to developing flight themselves using the markedly superior “feather” technology, leaving only the enormous azhdarchids and tapejarids to be devastated by that damned meteorite.

    Nobody can teach you to present in such a way that reporters won’t violate the work, but it’s certainly possible that a corps of presenters could be rallied to reform results into a form that is stable against reportorial violation. That would be much better than hurling abstracts over the transom and complaining about the result.

  26. 27 Nathan Myers 03/02/2009 at 3:29 pm

    Oops, I guess not even tapejarids.

  27. 28 David Hone 03/02/2009 at 5:37 pm

    ‘If they change “relative” to “ancestor”, it’s because they think that will engage their readers. Likewise if they change “pterosaur” to “dinosaur”.’

    But it will be wrong. Full stop. End of argument. As I have said many times before, this serves no-one and is a frustrating and simple error, not made by *good* journlaists with *training*, and is made by poor journalists who lack the knowledge and experience to do it properly. You would not send a football expert to cover a political convention so why send a non-scientist to cover dinosaurs? I can repeat this till I am blue in the face and perhaps we have have to agree to disagree, but if they change wat the words mean, they are getting things wrong. Relatives are NOT ancestors and pterosaurs are NOT dinosaurs, just as Obama is NOT a republican and water is a molecule and NOT an atom. For some reason some of these errors are considered acceptable and some are not. Journalists should increase understanding and pass on knowledge, if they are not doing that, I fail to see what they are there for? Things can be made understandable without being changed.

    ‘I think we both know that you are not personally equipped to present material in such a way that a reporter won’t be tempted to violate it. That just means the work needs to be filtered through somebody else who has demonstrated such a skill. That person is almost certainly not a reporter or a publicist.’
    Thanks for the studiedinsult, obviously writing papers, running blogs and science education website, teaching students, writing for popular magazines, and oh, yes, writing a dinosaur book and working as a consultant for the BBC clearly marks me out as inept. As I havewearily pointed out repeatedly to you, THIS IS NOT JUST ME. I can name dozens of scienists who will report exactly the smae failings of a multitude of journalists that I have laid out here, not least including BBC TV and Radio presenters who I happed to know who feel undermined by their own researchers (who I won’t embarass by naming here) not to mention entire blogs that are dedicated to this phenomenon alone (most notably the incredibly popular Bad Science). To suggest otherwise marks you out, not me. I have had a great deal of contact with journlaists of all kinds and all medias and am well awar eof the variety of styles of reporting and gaining of information.

    I am not sure what you mean by a ‘publicist’, but yes, my work often goes through university PR departments, as does that of others. Again, it is irrelevant what one says when (as happend to Witton and Naish) pterosaurs are described as bird ancestors. How exactly does one proof against that? Must we point out each and every taxon that pterosaurs are not related to in each press releare, interview or blog post so it is not misreported?

    ‘What engages the public, instead, are the actual animals, how they fit into the natural world, how they lived, what they looked like, and whether they could swallow oneself or one’s offspring whole.’
    What, like the work I and my colleagues do you mean? You really should read the press releases that go out, if you thing we jsut send abstracts to journalists and newspapers you are sorely misinformed. Half the time they call us having read the abstract and want to know more. I have to say you strike me as speaking from a position of ignornace here, I do not know about you it is true, but I have done a massive amount of media work and know many others who do, as well as people in the media itself, all this is based on my personal experience and that gained from close friends and colelagues. You criticse my way of writing to the media (apparently I do not have the ability) but I can’t remmeber the last time I sent you a press release. One more, and I hope for the last time, the good people get it right, if they can the rest can, they do not, how is this the error of the scientist. (And again a contradiction, you do not want to blame people, but apparently I am not doing it right – that looks like attributing blame form where I stand).

    I suggest we close this argument – easy to say when it’s my blog, but then that’s largely the point. This is my blog and my forum for my ideas, endless repetition of arguments is pointless. I am interested in questions and debates, but you want the scientists to pander to the jounalists and I want them to pander to us. If you truly want an “I’m interested in having an informed, engaged, and inspired public.” then the must be given accurate information, if you not you might as well tell them anything. And if you think i am not helping, then I would point to the stuff I dod do outside of the media – it may reach far fewer people, but over 50 000 people have used Ask A Biologist (not to mention my other projects), so I think I am getting some members of the public interested, informed and inspired.

  28. 29 Nathan Myers 03/02/2009 at 6:34 pm

    If you do the same thing again, you will get similar results. That’s the fundamental principle of science. If you’re not happy with similar results, logic suggests considering something different.

    That’s my last word.

  29. 30 David Hone 03/02/2009 at 7:09 pm

    That may be true, but once again, not all results are the same.

  30. 31 Nathan Myers 05/02/2009 at 11:38 am

    I post only to note that towering figure of science journalism, Ed Yong of “Not Exactly Rocket Science” fame, described a basal cetacean unabashedly as “ancestral” in

    I have seen, elsewhere, claims that mammal specialists tend to be particularly sloppy in this regard, so I doubt that Ed has personally interpolated the description.

  1. 1 Birds of a (leg) feather « Dave Hone’s Archosaur Musings Trackback on 15/07/2009 at 11:20 am
  2. 2 Anchiornis – again « Dave Hone’s Archosaur Musings Trackback on 25/09/2009 at 11:03 pm
  3. 3 Anchiornis and the temporal paradox « Dave Hone’s Archosaur Musings Trackback on 01/10/2009 at 3:15 pm
  4. 4 What colour was Anchiornis? « Dave Hone’s Archosaur Musings Trackback on 26/04/2010 at 7:51 am
  5. 5 50 Best Blogs for Paleontology Students | Bachelor's Degree Trackback on 12/09/2010 at 2:06 am
  6. 6 Začíná darwinovský rok 2009 | DinosaurusBlog Trackback on 31/10/2012 at 5:10 pm
Comments are currently closed.

@Dave_Hone on Twitter


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

Join 579 other subscribers

%d bloggers like this: