Taxonomy Fail

Today’s breaking news in Ant Science is this:

Newly discovered pieces of amber have given scientists a peek into the Africa of 95 million years ago, when flowering plants blossomed across Earth and the animal world scrambled to adapt.

Suspended in the stream of time were ancestors of modern spiders, wasps and ferns, but the prize is a wingless ant that challenges current notions about the origins of that globe-spanning insect family…Inside the Ethiopian amber is an ant that looks nothing like ants found in Cretaceous amber from France and Burma.

Wow- that’s big news! I wonder what this amazing Ur-ant looks like? Fortunately, WIRED has a photo:

WIRED's caption- "Photos From Alexander Schmidt/PNAS: 1) Wingless ant"

Maybe I’m going out on a limb here, but I’ll venture that this ant looks nothing like the other ants because it is, in fact, a beetle. With clearly visible elytra, and everything.

And because the press coverage is coming out ahead of the release of the PNAS paper, we can’t check the study to see if this is WIRED’s error or if the researchers themselves actually mistook a beetle for an ant.

update: The PNAS paper (Schmidt et al., 2010, Cretaceous African life captured in amber, PNAS doi 10.1073/pnas.1000948107) is now out.  And yes, the mistake lies with the authors, as Fig. 3A shows the same beetle labeled as an ant.  They write:

The most outstanding discovery is a complete, well-preserved although enrolled, wingless female ant (Formicidae; Fig. 3A). Visible characters preclude affinities with the extinct Sphecomyrminae, which is the only subfamily recorded for contemporaneous and older ants in mid-Cretaceous Burmese and French amber (15, 16). Regardless of the subfamily, this discovery is significant because it is one of the oldest records of an ant and the earliest from Gondwana. It has been suggested that ants arose in Laurasia during the Early Cretaceous (16–18), but the present discovery challenges this hypothesis. Ants evolved concurrent with the rise of angiosperms but apparently remained scarce until radiating into the world’s most diverse and ecologically dominant eusocial organisms during the Paleogene (19). The discovery will aid in resolving the phylogeny and timescale of ant lineages.

Unless, of course, the ant is a beetle. Who the hell reviewed this paper?

update 2: on Roberto Keller’s visualization, I’m now viewing this thing as possibly not a beetle either. But still not an ant.

update 3: in the NYT, too? Ug.

31 thoughts on “Taxonomy Fail”

  1. Well DUH, ants are just highly derived social beetles! Ever wonder why so many beetles are myrmecophiles that closely resemble ants? Cause they’re actually primitive ants (er, I meant, the sister taxon of all other ants)!

    That’s bad enough to be an April Fool’s joke, but four days late. I hope this paper won’t actually say what I just did.

  2. I’m hoping they actually found what they say they found, and simply put the wrong picture along with the (correct) caption. If not, HUGE FAIL. Before I even read you post, I was thinking “cool, nice beetle find!”

  3. I have been staring at that picture for some minutes now (the pdf version), in all angles, and I think I see Our Lady of Guadalupe. No, but seriously, I can’t identify an ant worker there or any other insect for that matter.

    Giving them the benefit of doubt, the only possibility is that the large dark oval on the upper-right is the gaster in ventral view (since no tergo-sternal articulations are visible) and the acute tip almost at the edge of the amber piece is the apex of the hypopygium. To the left you will have the mesosoma and further left the head with a strange Daphnia pulex-like eyes.

    It just doesn’t work. You cannot identify any of the usual plates that will be present in those tagmata of a worker ant. None of the appendices on the bottom of the inclusion look like ant legs (and believe me, I have seen plenty of those).

    I really doubt that it is an ant. If it is, then the report fails in not giving enough information to support the authors assessment on their specimen.

  4. Bizarre.

    Looks like a cicada’s head glued on to beetles body to me (In case you haven’t noticed, my solution to any identification problem is usually to assume some form of skulduggery – It’s a conspiracy I tell you!).

  5. I certainly don’t know but I am loving the conversation. Just discovered your blog and think its great. Thank you and I am adding to our science news visualization map under blogs!

  6. Matthias Svojtka

    Dear Ladies and Gentlemen of this – seemingly – high-end-scientific blog!

    Neither is this amber fossil a beetle, nor has PNAS become “a complete joke” (thanks to Kai for this nice phrase). It is an ant, that curled up in itself during death. Because the inclusion is very hard to photograph while observed in LM, current research is done with X-ray-synchroton microtomy.

    I regret to say, but myrmecos, what kind of specialist are you that you can judge a “taxonomy fail” from one picture alone? You have never seen and investigated the original fossil. Do you think we are a complete bunch of idiots, including PNAS and the New York Times?

    The ant, out of my private collection, was studied for over one year now and further study will be published in a seperate paper. Maybe you could be so kind to wait a little before coming up with the next “intelligent” blog.


    Matthias Svojtka
    Department of Paleontology
    University of Vienna / Austria

    1. Dear Matthias,

      Given that both your paper and the main media story emerging from your study is how this fossil has implications for ant evolution, I would have thought you’d have provided support for your taxonomic interpretation of this fossil.

      Nowhere in the image, nor in the text, nor in the supporting information, is there anything that indicates that this fossil is an ant. There’s no documentation of a metapleural gland, petiolar constriction, no homology assessments of the sclerites.

      Perhaps that information will be forthcoming in your paper, and if so I do look forward to reading it. I will happily eat my words if I am mistaken. But from what has been shown so far there is insufficient evidence to support your conclusion. To this ant taxonomist, at least, that fossil does not look like an ant.

      That the story appears in PNAS and NYTimes does not by itself vouch for the quality of the research. It merely begs the question.

    2. Matthias,
      Why on earth was that photo used, then? Why not publish an image that actually looks like an ant? Does the X-ray synchroton produce images that look like ants? Most of the comments on here are about the image – which really looks nothing like an ant!

  7. Matthias Svojtka

    Dear Myrmecos,

    you may have noticed that our paper is an overview-paper, presenting the whole story about the new Ethiopian amber (amber chemistry, physical characteristics, geology, inclusions). Have you downloaded the SI-appendix? There is simply no room for details (morphology, phylogeny) on ANY taxon in PNAS, this will be done in forthcoming papers and is only hinted at in the present publication.

    What remains of this discussion is the fact – sorry – that you are arrogant enough to judge a complete error having seen ONE picture alone. The fossil and it’s preservation simply doesn’t allow a better display as far as conventional methods are concerned. Citing some entomological termini technici and listing features which you miss or not in our study, doesn’t make you a better scientist.

    To improve your knowledge on ants I will, of course, send you every forthcoming paper of interest to


    M. Svojtka

    1. Matthias,

      So we aren’t supposed to judge your conclusions based on the information you presented in the paper, but just by accepting your authority about details that are currently unavailable to the public? I don’t buy it.

      I’m happy to be wrong on this. There’s certainly nothing implausible about a 95 mya true ant in African amber. The discovery of this amber deposit is indeed big news- I’m not casting aspersion on the whole of your paper.

      But I do hope you understand that I can’t accept this fossil as an ant without further justification.

  8. Matthias Svojtka

    But you KNOW that “it is, in fact, a beetle”?

    I will stop our “scientific” communication at this point.

  9. Gordon Snelling

    Miss him everyday. He has debunked fossil ants before, would like to know which myrmecologist looked at this ant.

  10. Vincent Perrichot

    Well, sounds that the ant nature of our fossil is getting much controversy here! I understand that the photograph provided in our paper is not very clear, so I’d like to clarify things and try to convince everyone. First of all the photograph you are commenting on was published here and by Wired Science in reverse position. It should be viewed at 180º to give at least a better idea. I’ve sent a picture to Alex, which might help if he wants to post it on his blog. Now, the fossil is preserved enrolled and contrary to extant specimens, it is impossible to unroll, which makes its study much more difficult. Plus, anyone having tried to image some amber fossil one day knows that it is very challenging to obtain a good, well-focused, picture. So here I totally agree with one of my co-author Matthias Svojtka that it is very speculative to reject a determination based on a single photograph. Even with the specimen under your steromicroscope you will need time to figure out how this specimen is preserved and what it actually is. I’ve been examining and working on thousands of amber arthropods in the last 11 years and my first thought when I discovered this one was: “what the hell is this?”. It is only after a careful examination that I was able to see by transparency the antennae, a part of the mandible, and the petiole. So yes, it’s actually an ant!! No wing, no elytra… Sorry that the photo is not speaking better by itself. When it’s viewed on the correct position, the head is on the top with the right eye small, darkened. The vertically oval structure on the bottom right is the gaster with apex pointing upward, touching and overlying the mouthparts. The petiole appears cylindrical between the gaster and the legs that are mostly folded over the mesosoma and run on the left side of the pic. Our PNAS article is an overview paper of the amber deposit, thus it was not the purpose to describe the fossils in detail (will come elsewhere) and the space is too limited to justify our determination (otherwise I would have had to do so for each of the 30 arthropods). There are 12-segmented geniculate antennae with a very long scape, mandibles triangular to falcate, with a multi-toothed masticatory margin, a metapleural gland opening, a well-defined (although unusual) petiole, forelegs with a calcar spur… For now I am unable to tell what subfamily it belongs to (except it is not a sphecomyrmine) because too many characters are hidden by the legs and gaster, as preserved. This is why I am currently making a 3D reconstruction using high-resolution X-ray synchrotron imaging. The reconstruction of the specimen has been completed only very recently, which explains why such image was not provided in our paper but only the stereomicroscopic photo. The detailed study will be published elsewhere ASAP and you will understand that I cannot provide this reconstructed image here. Still, the diagnostic features are hidden on this view, so I am now working on a virtual dissection (e.g., reconstruction of the head alone, of the alitrunk alone. etc…) to access all characters. This is a long work on which I and two colleagues from the synchrotron are working for several months, so be patient and you’ll see a true ant! 🙂

  11. Pingback: The Amber Ant of Mysteries (Taxonomy Fail, updated) « Myrmecos Blog

  12. Pingback: Anatomy of an Error « Myrmecos Blog

  13. Pingback: The oldest known ant | Archetype

  14. Pingback: YourTechWorld » Saturday links | Not Exactly Rocket Science

Leave a Reply