About the SA Blog Network

Tetrapod Zoology

Tetrapod Zoology

Amphibians, reptiles, birds and mammals - living and extinct
Tetrapod Zoology Home

Coming next: and the Dave Peters thing

Email   PrintPrint

Longisquama insignis, based on a skeletal reconstruction published by Dave Peters in 2004. I don't think that this insanely elaborate reconstruction is in the least bit accurate.

This is an annoying teaser for the very long article due to appear here soon: an attempt to counteract the massive web presence of the wholly misleading, mis-educational And, no, Longisquama did not really look like this… alas, Dave Peters tells us that it did [the drawing shown here - produced by me - is based on a skeletal reconstruction of Longisquama that Dave published in 2004]. More soon.

Darren Naish About the Author: Darren Naish is a science writer, technical editor and palaeozoologist (affiliated with the University of Southampton, UK). He mostly works on Cretaceous dinosaurs and pterosaurs but has an avid interest in all things tetrapod. His publications can be downloaded at He has been blogging at Tetrapod Zoology since 2006. Check out the Tet Zoo podcast at!

Nature Blog Network

Follow on Twitter @TetZoo.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 29 Comments

Add Comment
  1. 1. John Harshman 3:04 pm 06/30/2012

    Your link to (which I have never heard of) comes right back to your post. Was that the intention?

    Link to this
  2. 2. John Harshman 3:24 pm 06/30/2012

    OK, I went there, and now I’m back. That’s exceedingly bizarre. And, I note in passing, there’s not a single mention anywhere of any molecular data.

    Link to this
  3. 3. naishd 3:35 pm 06/30/2012

    Oops, sorry, the link was a mistake. Now corrected. Yes, I’ve been corresponding with Dave today and I put it to him that molecular data alone falsifies his phylogenetic hypotheses.


    Link to this
  4. 4. John Harshman 5:07 pm 06/30/2012

    I noticed also that there’s no indication of support on any of his trees. And with a characters:taxon ratio of around 1:1, I’d be very surprised if there were much support for most of it.

    Link to this
  5. 5. Therizinosaurus 5:07 pm 06/30/2012

    I look forward to this. I’ve been engaging Peters lately on his website and the results are not encouraging. Even ignoring his crazy imaginary morphologies and terribly formed characters, he declared Varanosaurus to not be an ophiacodont after only including 6 of 25 proposed synapomorphies, he thought Ophiacodon was closer to therapsids than Haptodus because he interpreted Romer and Price’s (1940) dotted lines as actual sutures, he declares Saurischia paraphyletic after including only 3 of 18 characters, and places many non-dinosaurs in Dinosauria after using only 1 of 12 dinosaurian characters, etc..

    Regarding molecules, I commented- “And yet both nuclear and mitochondrial analyses result in monophyletic snakes, as do morphological analyses like Gauthier et al. (2012) with over twice the number of characters as yours and over five times the number of squamates. So your claim is that without fossils, both the mitochondrial genome AND the nuclear genome happen to find the same, wrong phylogeny? A wrong phylogeny that is also found when the clade is analyzed in far more detail morphologically? That doesn’t seem suspicious to you?”

    Worst of all for a scientist is this statement- “As long as I can recover a single tree from all these reptiles, I’m not going to add another character. Sorry. I only take things this far. No further.”

    Link to this
  6. 6. John Harshman 5:08 pm 06/30/2012

    Actually, I’m surprised he gets a single tree at all. There should be millions of equally parsimonious trees from a data set like that. Assuming he did a real analysis at all.

    Link to this
  7. 7. naishd 6:52 pm 06/30/2012

    I’ve never understood Dave’s approach to compiling or analysing data. I’m pretty sure I remember him saying that he always expects a CI of 1, and if he doesn’t get this he goes back and adds to the analysis. Anyone else remember this?


    Link to this
  8. 8. naishd 10:37 pm 06/30/2012

    Incidentally, Dave has written a short article about this teaser on Pterosaur Heresies here. Therein, he promises to respond. I don’t know how this is going to play out. It may be an enormous waste of my time.


    Link to this
  9. 9. Therizinosaurus 3:58 am 07/1/2012

    John- Peters’ analysis really does result in 2 mpts. It’s probably due to his numerous correlated characters and imagined completeness of specimens.

    Darren- I don’t think Peters said that, since in a recent comment he correctly stated the CI of his reptile analysis is about 0.1 (actually 0.1040), though he didn’t know what that meant. You might be thinking of his numerous statements that he wants one MPT, then considers attaining that the goal. Which is stupid of course, as a partially resolved tree may be more accurate or rigorously tested than a fully resolved one.

    Link to this
  10. 10. Kenshin 7:49 am 07/1/2012

    Re: Peter’s response. I don’t understand why he considers your figure a “comic lampoon”. You treated his reconstruction pretty fairly without gross distortion

    Link to this
  11. 11. vdinets 8:42 am 07/1/2012

    Nice drawing. Makes it instantly obvious what the creature’s hunting technique was. I am surprised no one has realized it before. The animal would lure male butterflies by flashing dorsal quills of appropriate color, make a high jump and impale them on those terminal tail spines. Mystery finally solved.

    Link to this
  12. 12. John Harshman 9:21 am 07/1/2012

    This just isn’t credible. He now says his tree has over 500 taxa. That’s over 1000 nodes. Can we really believe that at least one of his 200 or so characters changes unambiguously on every one of those nodes? There just aren’t enough to go around. A CI of 0.1 tells us there are around 2000 state changes, total, which is technically enough, but they would have to be perfectly distributed to make it work. No real data set would behave that way.

    Link to this
  13. 13. Jerzy v. 3.0. 11:05 am 07/1/2012

    Just found Not, didn’t believe it. But loved it. I think in science it is important to keep a small percentage of fresh and alternative. Just a small percentage, to be sure. I guess of 100 crazy stuff on his website, 2 or 3 turns something right and new.

    No, didn’t believe tree-reaching Tanystropheus. Come on, the creature is just a few feet long. Maybe it was specialized for living in bonsai gardens.

    Link to this
  14. 14. naishd 3:27 pm 07/1/2012

    Kenshin (comment 10) says…

    “I don’t understand why he considers your figure a “comic lampoon”. You treated his reconstruction pretty fairly without gross distortion.”

    Thanks, I absolutely agree. If anyone has Dave’s 2004 article from Prehistoric Times titled ‘Pterosaurs from another angle’, look at the Longisquama illustration on the third page. My drawing above is a totally ‘accurate’ re-drawing of that GSP-style reconstruction, the only difference being that my drawing is meant to show the live animal in full colour. So, yes, I don’t get Dave’s thinking at all. His newer versions of Longisquama are even more elaborate.

    Jerzy – the largest forms of Tanystropheus are big: 5 m in total or more, and even the small ones are 2-3 m long in total. I don’t say this to support the interpretation, but you get the point.


    Link to this
  15. 15. John Harshman 8:43 pm 07/1/2012

    Still, even 5 meters is a tiny little tree.

    Link to this
  16. 16. Jerzy v. 3.0. 7:49 am 07/2/2012

    About 2/5 of the creature is tail…

    Link to this
  17. 17. THoltz 8:46 am 07/2/2012

    Well, I think the reconstruction of Longisquama got at least this right: the critter had a head, a torso, forelimbs, and (presumably) hindlimbs and a tail…

    Link to this
  18. 18. 7:49 pm 07/2/2012

    Dave Peters here. I’m looking forward to your critique, Darren. Let’er rip!

    With regard to the illustration, your readers should know that I invited you to copy and paste my latest version of Longisquama. Anything else done by any other hand, is not representative of my style, nor does a 2004 illustration represent the latest thinking about this difficult fossil. The “insanely elaborate” caption you placed below Longisquama is ironic in that the “insanely elaborate” plumes everyone has seen and written about. That’s not the issue! I simply added hind limbs, a pelvis, tail and some forelimb digits. Even if these elements were not present in the fossil, phylogenetic bracketing would have suggested long hind limbs and an attenuate tail as Longisquama nests between pterosaurs and Sharovipteryx.

    To those who think 228 characters are too few, I can only respond that so far that number has proven to be sufficient. I don’t need the presence of a carapace or plastron to nest the two basal turtles together, for instance. Remember this study has one purpose: to employ a very wide range of reptiles to see where they nest. That has never been done before. Currently there is no other study of comparable scope. I am as surprised by the novel nestings as the next guy, but all sister taxa resemble one another. If anyone can find a mis-nested taxon pair, please let me know. We’ll fix it.

    The study began over a hundred taxa ago. Subsequent additions have not changed the tree topology. The tree can also be pared down to 60 or so hand-picked taxa and the topology does not change.

    Support values? Some are weak (those associated with incomplete taxa) while others are strong. The very size of the matrix 300+ taxa and 228 characters is itself a strength. Errors, if present, can erode relationships when corrected. However, corrections can also cement relationships. So far, most errors have cemented relationships when corrected.

    And then, with regard to the molecules and their divergence from all or most morphological analyses, that’s a good question. Notably few DNA studies agree with one another. So there’s that to consider as well. The study is mostly about fossils and, unfortunately we will never have their DNA.

    The Gauthier et al. (2012) study is important, but did not include several fossil taxa that I found to be keys to resolving relationships. It’s notable that most previous studies, along with the 2012 Gauthier study, nest the highly derived burrowing snake, Leptotyphlops, at the base of the snake tree. Rather than being the most plesiomorphic snake, it is quite derived. The jaws move medially, for instance, not up and down like other tetrapods. So questions like this attend even major studies conducted by the major names in paleo. Other problems were discussed within the week at pterosaurheresies.

    Bottom line: whether by voodoo or prayer, the proof is in the pudding. All the sisters look like one another. One can trace the development of derived characters all the way back to basal tetrapods. And thus, many of the mysteries of paleontology have been solved. Pterosaurs were lizards. Mesosaurs lost their diapsid fenestra. Caseasaurs were closer to millerettids than synapsids. It’s all there.

    I’m hoping for some good scientific evidence to upset the reptileevolution tree and its topology. Let’s get this right together. All constructive criticisms are welcome.

    The term “unconvincing” is not convincing to anyone. I hope that term won’t pepper your criticism. Offer your alternative or present evidence to the contrary and you’ll win converts, as I have done. Or, throw out the “bad” evidence and see if the tree changes topology. Warts and all, the MacClade file is and has been available to one and all by request.

    Bring it on…

    Link to this
  19. 19. Geopelia 9:42 pm 07/2/2012

    Where do the Rhynchocephalians fit into the family tree?

    One still exists today, the Tuatara.

    Link to this
  20. 20. Therizinosaurus 9:45 pm 07/2/2012

    “To those who think 228 characters are too few, I can only respond that so far that number has proven to be sufficient.”

    It’s sufficient to find A phylogeny, but not sufficient to test the traditional phylogeny. If you don’t include most of the characters others used to make their trees, how can you say you’ve shown their trees to be wrong? You don’t realize that adding characters is as important as adding taxa.

    “If anyone can find a mis-nested taxon pair, please let me know. We’ll fix it.”

    This is disingenuous of you to keep repeating. I did exactly that on your blog for Scutellosaurus and Scelidosaurus, and even listed the characters missing from your matrix that would help group them together. Your response was “As long as I can recover a single tree from all these reptiles, I’m not going to add another character. Sorry. I only take things this far. No further. I’m not a dinosaur expert. I’m not a professional paleontologists. I’m just here to raise possibilities and provide evidence for those possibilities.” ( ) Thus you present the image of having a tree which none of your dissenters can justifiably criticize and claim to be open to engaging with such criticism, but then back off and say you’re not an expert and you’re not changing things once that criticism is actually provided.

    You say “bring it on”. I say I brought it and you ran.

    Link to this
  21. 21. Morgan Churchill 12:01 am 07/3/2012

    Producing a tree doesn’t mean that the tree is correct. With selective choice of characters, you can pretty much get any topology you wish. The only way to truly test the tree is, as Therizinosaurus said, to incorporate more characters to see what happens. How many characters support these controversial groupings, and how well supported are they?

    Link to this
  22. 22. naishd 4:06 am 07/3/2012

    I have to strongly agree with everything that Mickey says. Dave: you aren’t doing your phylogenetics right at all.

    And the comment about your Longisquama looking “insanely elaborate” comes from the stuff that only you claim to see, not the structures we all know are there.

    Anyway, I suggest we save our energy and time for the final article, I’m hoping to have it published later today. This is not an effort to ‘demolish’ Dave Peters; it’s an effort to counter the influence and mis-education of Dave’s hypotheses, reconstructions and trees (as seen at


    Link to this
  23. 23. John Harshman 9:46 am 07/3/2012

    Dave: OK, I’m a little confused. In some places it says your matrix has over 200 taxa, here you say over 300, and elsewhere you say over 500. While the first two are technically correct if the last is true, I’d like to know: how many taxa in your matrix? Regardless, I submit that 228 characters are way too few to give a fully resolved tree. Or at least it won’t happen without very good luck and careful planning. First, it requires massive homoplasy, since you have a thousand nodes to deal with and each character must therefore do multiple duty. And this creates the problem that all the homoplasy must happen exactly so, and balance exactly so, in order for all optimizations to be unambiguous. I would have difficulty producing such a matrix, even a purely hypothetical one.

    But now I’m curious. Please send me the matrix, as your web site offers. I want to see how it’s done. I’m jharshman (at symbol) pacbell (dot symbol) net.

    I also take issue with your facile rejection of molecular data, but we can get to that another time.

    Link to this
  24. 24. ohnosir 1:00 pm 07/3/2012

    As a longtime “lurker” of TetZoo and an art student with a particular interest in evolution and paleontology, I will be eagerly anticipating this article. I have recently been very interested in the synapsid evolutionary line, but with limited access to scientific journals (sadly, the online research database of art school is not nearly as generous as that of a public university), I have to wade through Google results in hopes of finding accurate and verifiable information.
    Reptile Evolution has been a source of major frustration for me ever since I’ve become interested in reptile-to-mammal evolution, as it takes up the vast majority of results that I am able to find on this topic, particularly if I am looking for some good skeletal restorations. Regardless of the validity of Mr. Peters’ proposed phylogenies, I am necessarily wary of any radical hypotheses out there that I can’t independently verify. Trying to sort out what is standard, accepted knowledge and what is controversial is extremely difficult when 99% of what I can find is from one source!

    That being said, I hope you will cite lots of publicly accessible sources for your evidence in this article, and I hope you will clear up some of the confusion over the synapsid lineages! Also, in the meantime, if anybody has any suggestions of some good online sources or books I can look up for more information on the subject, particularly ones with lots of nice, big, detailed fossil images and restorations, I could not be more grateful!

    PS – Does anybody know if there is a way to make your username link to your website rather than to your email? Or am I hoping for too much?

    - OhNoSir

    Link to this
  25. 25. ohnosir 1:01 pm 07/3/2012

    Hm, it appears the username doesn’t link to anything at all. How unfortunate.

    Link to this
  26. 26. naishd 1:16 pm 07/3/2012

    The article isn’t really about synapsids, sorry, so don’t expect much useful info on them. BUT I will be covering them (non-mammalian synapsids) at length some time in the near future.

    And, as you note, the usernames here at SciAm can’t be linked to anything – a major negative point…


    Link to this
  27. 27. Halbred 1:40 pm 07/3/2012

    @ohnosir: I too have had trouble accessing good synapsid information in one place, but I think that’s because there’s not an up-to-date volume covering synapsids as there is for dinosaurs (“The Dinosauria 2nd Ed.,” “The Complete Dinosaurs 2nd Ed.”) or even pterosaurs (“Prehistoric Flying Reptiles,” “The Pterosaurs from Deep Time,” and Mark Witton’s upcoming opus). This is a real shame, because there are some really strange synapsids, and the group’s history is fascinating.

    The usual excuse is that synapsids aren’t as “sexy” as dinosaurs, but that can’t be right–look at Eustemmenosuchus or Suminia!

    If there is a “catch-all” volume that I’ve somehow managed to completely overlook, folks, please let me know.

    Link to this
  28. 28. ohnosir 2:33 pm 07/3/2012

    Alas, I guess I’ll keep holding out for that synapsid article! But either way, I think a good critique of Reptile Evolution is well in order.

    Also, although you can link your profile to a social media site, there is absolutely no option anywhere that I can find to add a personal website to your profile, which is a pretty standard blogging practice to overlook on SciAm’s part.

    Link to this
  29. 29. David Marjanović 7:26 pm 07/9/2012

    This just isn’t credible. He now says his tree has over 500 taxa. That’s over 1000 nodes. Can we really believe that at least one of his 200 or so characters changes unambiguously on every one of those nodes? There just aren’t enough to go around.

    There may be, because many of his characters have very high numbers of states.

    However, I’m with everything Mickey has said so far.

    It’s notable that most previous studies, along with the 2012 Gauthier study, nest the highly derived burrowing snake, Leptotyphlops, at the base of the snake tree. Rather than being the most plesiomorphic snake, it is quite derived.

    As – sadly – usual, you’re completely misinterpreting how cladograms work. Time does not run from one tip to another, it runs from the root to all tips equally. Leptotyphlops isn’t any closer to the root than Boa or Natrix. It’s part of the sister-group of a clade that includes almost all other snakes; it has its own long independent evolution behind itself.

    Again: the tree by Gauthier et al. (2012) does not find Lepotyphlops at the base of the snake tree, it finds the base of the snake tree at the base of the snake tree; and it does not claim that Lepotyphlops retains the highest number of plesiomorphies, it doesn’t say anything about this topic at all.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Scientific American MIND iPad

Give a Gift & Get a Gift - Free!

Give a 1 year subscription as low as $14.99

Subscribe Now >>


Email this Article