Thursday, December 15, 2011

Deinonychus, a Cretaceous Rodeo Rider

"Gaaaar! I'm gonna hold you down and eat you alive! While flapping!" Photo by Dominique Pipet, via Flickr.

What did Deinonychus really do with its feet? John Ostrom initially painted the picture of a fleet-footed predator, chasing down prey animals and slashing at them with the enlarged claw on its second toe. To take it from Crichton, they were slitting bellies and dancing in spilled viscera. A few years ago, Phil Manning of the University of Manchester suggested they were more likely "climbing crampons," allowing them to cling to the panicky tenontosaurs they were attacking. In a new PLoS One paper, Denver Fowler, with Elizabeth Freedman, John Scannella, and Robert Kambic, puts forward another option: they were grasping tools for holding down smaller prey (or, in typically colorful science-speak, prey of "subequal body size").

To come to this conclusion, Fowler and team compared the feet of Deinonychus to other animals: a diverse group of extinct theropod genera as well as living birds. When compared to dinosaurs, the proportions of their feet differed strikingly from ornithomimids and alvarezsaurs, both groups whose leg proportions strongly suggest a cursorial lifestyle. They were runners. As Matt Martyniuk recently wrote in an excellent DinoGoss post, the sticky assumption that dromaeosaurs were particularly quick animals, chasing down prey, isn't really supported by the evidence. Noting that Ostrom's first ideas changed once he took full stock of the proportions of Deinonychus's metatarsus length to its tibia length, Matt writes, "Not only was Deinonychus not particularly fast, it probably could not have been nearly as fast as most other small theropods, including modern flightless birds, let alone cheetahs."

To test his hypothesis that Deinonychus was instead using its feet to hold onto prey as it killed it, Fowler compared its foot proportions to living birds who use their feet in diverse ways, including accipiters, owls, falcons, passerines, vultures, osprey, woodpeckers, and pheasants. Accipiters, AKA hawks and eagles, are adept at holding down prey, using second toes which have larger claws than the others. This second claw was the closest analog to the dromaeosaur "sickle-claw" yet found. The metatarsal bones were short and stocky like those of owls, suggesting that they were just as good at grasping small prey, if not built to constrict around prey and squeeze the life out of it the way accipiters can.

Also assisting Deinonychus in subduing and keeping a grip on its prey, Fowler suggests, is that feathered forelimbs could have been used the way accipiters use their wings, flapping to maintain stability on top of a struggling prey animal. He writes,
Even if Deinonychosauria were not capable of a full avian-like flapping ability, they may have been able to perform a rudimentary flight stroke during stability flapping. Similarly, long feathered tails are conspicuous in accipiters and aid in maneuverability and balance during stability flapping. Basal Paraves and Deinonychosauria possessed long bony tails which are shown to have been well feathered... and would have assisted balance during predation and stability flapping.
This suggests another use for flapping, feathered forelimbs other than flight. The dromaeosaur would ambush its prey, hold it down by plunging its enlarged second claw into it, holding it down by its own weight, flapping to keep its balance and maintain its grip. Those long forearm feathers and fiercely clawed hands would have also been useful to "mantle" their meal, protecting it from rivals and further preventing escape. Fowler writes,
An important part of our interpretation is that ground-based predation need not necessarily be conducted at high speed. It is commonplace for extant terrestrial predators to employ surprise ambush techniques; goshawks and other forest raptor species commonly hunt on the ground, employing ambush and maneuverability as strategies, rather than outright pursuit
Troodontids also figure into this, as Fowler notes their common presence in ecosystems alongside dromaeosaurs. Comparitively, the feet of troodontids were better suited for chasing down their prey. As dromaeosaurs went for larger prey, troodontids may have been specialized for hunting small mammals, for example, perhaps in darkness as suggested by their famously large eyes.

Noting that extant avian theropods such as Golden Eagles aren't afraid to go after much larger prey, Fowler writes that his research here doesn't necessarily preclude the long-inferred habit of dromaeosaurs to pick on animals their own size or larger. Had they worked up the gumption to have tenontosaur for supper, the "climbing crampon" quality of the second toe claw noted by Phil Manning would certainly come in handy for prey riding. In a sort of Cretaceous rodeo, the dromaeosaur would have leaped onto the back of a passing tenontosaur, holding on for dear life as it thrashed about until blood loss and exhaustion finally brought it down.

This research is rich with possibilities for future studies, and it will be fun to see what comes of it. There are the biomechanical implications of stability flapping to look at, as well as plenty of argument over how this hypothesis impacts the early evolution of flight. Lots of meat on this bone.

Also published in PLoS One yesterday: Scannella and Jack Horner continue to strike genera from the books, this time adding Nedoceratops to the list of ontogenetic phases of Triceratops. Another deals with how reproductive strategies affect body size in a variety of animals, including dinosaurs.


  1. Excellent post - good to see you bringing the quality content back after weeks of my toy/ugly robots nonsense and mean-spirited, not entirely self-aware enough mockery - but the model is pretty fugly...

  2. You need to find a job as a professional self-deprecator!

  3. Are we sure Alan Grant wasn't co-author of the study? Extract from the paper: "Victim is eaten alive or..."; Sam Neil in JP: "Point is, you're alive when they start to eat you".

    Joke aside I'm glad to see (yet)another plausible hypothesis for the function of the second digit in deinonychosaurs. It's implausible that such adaptation had one and only one function.

  4. And that was a big part of the book, too. Very memorable scene in which Dr. Wu gets it.

  5. I like how David balanced the quality of the article with the worst 'raptor' sculptyre in human history.

  6. @optimisticpainter I'm not sure if it's the worst ever. I've seen some very bad ones in my time.

    But yeah, it's very ugly.

  7. Actually, forget what I just said, and get a load of THIS.

  8. Ah, David, your penchant for cowboys & dinosaurs... :D

    @Marc: My word, those are frightful.

  9. I have mixed feelings about this paper.

    On the 1 hand, I like papers about what non-avian dinos (especially eudromaeosaurs) were like when alive based on comparisons w/living animals (especially eagles); I like the idea that grasping feet & stability flapping were pre-adaptations in paravians for WAIR/CFD, arboreality &/or big game hunting; & I like that I'm not the only 1 to notice the enlargement of eagle D-2 talons.

    On the other hand, I don't like that this paper ignores the blade-like morphology of eudromaeosaur D-2 claws (which suggests a slashing function); I don't like that this paper refers to "Manning et al. 2006" & "Roach & Brinkman 2007" for support despite their many problems (E.g. See the 25th comment: ) (E.g. See "Behavior": ); & I don't like that this paper ignores the probability of eudromaeosaur hands extending beyond the feathers (in reference to "Mantling and possible use of the forelimb"). It also doesn't help that 1 of the authors rubs me the wrong w/his DML posts (If I didn't know better, I'd say he's Horner's personal cheerleader).

  10. I almost forgot to mention the following.


    Am I the only 1 to notice the striking similarity btwn Carroll's Deinonychus ( ) & Rey's ( )?

    @David Orr

    Welcome back!

    @Mark Vincent

    & I thought Alan Groves' sculptures were bad.

  11. Love the comments here almost as much as the article. I think optimisticpainter's is my favorite.

    I have absolutely no expertise, and I know how some folks feel about Mr Fowler, so that may color mine or others' objections to the finding, but is this all that revelatory? I mean, I don't think anyone would ever have objected to the hypothesis that raptors subdued prey with their feet. His supporting hypotheses grab me even less, like flapping wings to stabilize one's self over struggling prey. Surely there are better reasons to retain wings, not to mention that wingless hunters today the same size as Deinonychus don't seem to have any trouble subduing prey without wings... Seems like if the prey's too big, two stubby-winged arms aren't going to do much good as far as holding it down. Some extra kg

    And as Hadiaz points out, accipiter claws are basically round in cross-section, not flat and blade/like. The propensity for deinonychosaur claws to "rack" sideways so easily must surely have been problematic-- dislocated toes, broken claws, etc. Only if the interior edge were very, very sharp would the utility of such a claw begin to mitigate these problems. This for me has always been the problem with dromaeosaurs, and why the function of their claw remains as inscrutable as ever.

    And Marc (and David, too)-- the "horrible dromaeosaur sculpture" one-upsmanship must stop. We can only enjoy them ironically so far before they make us physically ill;)

  12. Thanks for the excellent round of comments, everyone. A nice treat after a night of x-mas shopping. Thanks to Hadiaz and PH for those excellent observations on where the Fowler paper has some weaknesses. I hadn't thought to consider the *shape* of accipiter talons. This would be another spot that would be ripe for another piece of research.

    @Hadiaz - Good point about the Carroll drawing. The perspective on those big ol' arms is pretty comical. It is very Rey-ish. Just missing the lurid colors.

  13. So if I'm reading this aright, the very name Dromaeosaur is something of a misnomer?

  14. Ha ha, too true, Taranaich! And if you're persuaded by Fowler's arguments, it would almost be better to let dromies and troodontids trade names.

  15. We've known for sometime (since the 1990s) that the classic dromaeosaurids (the eudromaeosaurs nowadays) were absolutely NOT speed adapted animals, based on limb proportions. If it weren't for a certain major fictional franchise hyping the exact opposite, maybe it would have sunk in.

  16. @Holtz - We need more movies! It will be like the "fast zombie" vs "slow zombie" debate. That's how we'll settle it.

  17. @Hadiaz "I don't like that this paper ignores the probability of eudromaeosaur hands extending beyond the feathers"

    That would be quite an M.C. Escher like animal indeed. given that the manus including the metacarpal and at least the proximal phalanges are an integral part of the aviremigian wing to which the feathers attach, it would be logically impossible for the hand to extend beyond the wing--the wing itself being an extension of the hand.

  18. Thanks, I appreciate the blog post.

    In response to a couple points:
    The lateral compression that you see in dromaeosaur D-II unguals is diferent from accipitrids. We note about this (and Manning's work) in the supp info. What difference the lateral compression makes is not clear. I agree that lateral compression would make slashing easier than if they were broad, but I doubt if it would make much difference, and thus I still doubt that this would be the purpose of the claw. I suspect it is for impaling, and that the lateral compression makes it easier to impale (needs experiment). Ospreys also have laterally compressed claws, and they impale fish.

    I don't mind being called a cheerleader for Jack; he is my supervisor after all.

    I admit sometimes I get a bit highly charged about things. But then, I'm passionate about my science, and people have often been quite unreasonable about Jack's work. Jack has great ideas, and I never really understand why people object so strongly. Ontogeny is a new frontier in dinosaur paleobiology, and it has the potential to make sense of all sorts of issues.

    Anyway, thanks again for highlighting our work.

  19. @Denver The 'Fighting Dinosaurs' show a dromaeosaur using its 'sickle claw' in a stabbing/impaling fashion.

  20. I may be wrong, but I think this is the first time an author popped in to comment on a post about a new paper. Thanks for commenting, Denver!

  21. This comment has been removed by the author.

  22. @Matthew Martyniuk

    "That would be quite an M.C. Escher like animal indeed."

    How so? IIRC (If not, then please correct me), Qilong's Velociraptor ( ) has hands that extend beyond the feathers, yet the feathers are still long enough to brood w/. In any case, Senter basically said that while some dromaeosaur hands (E.g. Those of Microraptor) had long feathers, others (E.g. Those of Bambiraptor) extended beyond the feathers ( ).

    @Denver Fowler

    "The lateral compression that you see in dromaeosaur D-II unguals is diferent from accipitrids."

    1stly, many thanks for taking my concerns seriously.

    2ndly (as others have pointed out elsewhere), eudromaeosaur D-2 claws aren't just narrow, they're "narrowest ventrally" (See the 27th comment: ).

  23. @Hadiaz
    The claws do narrow ventrally, but the ventral edge is still rounded, much like Osprey unguals and claws. The fact that Ospreys have claws this shape is telling (they only impale). A survey of birds with narrow claws might be informative. IIRC woodpeckers have pretty narrow claws: again used for impaling (wood). Some owls have narrow claws (usually not all claws though; D-I IIRC).

    The physical features that would aid slashing (ie. a narrow claw) may also aid piercing (for impaling). Some physical experimentation may help with this question (repeating Manning's experiments maybe). Barrett's tetzoo comments are correct in asking how to test the slashing hypothesis. Do extant big cats rally slash with their claws? Exclusively? No modern birds slash with their foot claws, but the Seriema (see our sup) does hook-and-pull at flesh.

    The claws are obviously useful for different purposes. I wouldn't be surprised if dromeosaurs (et al) used the claws for pulling food apart, or helping climb (as we point out in the supp, all birds use claws to climb). The question is which function(s) is/are being actively selected for? We present accipitrids as the closest modern analogue, which also happens to be in a bird (ie. within the EPB; not a requirement, but it helps). Maybe we're wrong, but noone has considered this before, and it matches up with other aspects of pedal morphology and the evolutionary trends that we observe. Our paper is not supposed to be a conclusive answer / closed book. Rather, we present a new analogue, posit a few ideas that match up with the morphology, and invite others to investigate the hypotheses. Asking what the additional narrowness does functionally is a good question to try and test.

    On a related note, there is someone (not Burnham; I don't agree with his recent paper at all) currently working on the growth of the keratinous sheath and how it may be related to ungual shape . It might inform a bit on this subject. Hopefully that work is on its way to publication.

  24. @Denver Fowler

    I'm not saying that eudromaeosaur D-2 claws didn't have multiple functions: Like cat hind claws, they were probably used for hanging onto relatively large prey & climbing (in addition to slashing); Also, as I implied in my 1st comment, small basal paravians probably used their D-2 claws to pin relatively small prey & microraptorines/eudromaeosaurs later exapted their D-2 claws for arboreality/big game hunting (respectively). I'm just saying that 1 of their functions was probably slashing, given their striking resemblance to cat hind claws (As Naish pointed out, cat hind claws resemble eudromaeosaur D-2 claws even more than cat front claws), the cat-like hunting technique of eudromaeosaurs (Grapple-&-slash: ) & the preference of eudromaeosaurs for relatively large prey.

    BTW, aren't Osprey talons "nearly circular in cross section" ( ) as opposed to teardrop shaped in cross section like eudromaeosaur D-2 claws?

  25. One moment, please. I might have read it wrong but slashing motions for D-II claws are unlikely since eudromaeosaurs didn’t have a suitable cutting edge.

  26. @Hadiaz

    Osprey claws aren't circular. I think what the writer at is trying to enunciate is that typical falcon & hawk claws are arched (agreed) in cross section, whereas Ospreys' are rounded.

    Some pictures (not brilliant, but I am not at the museum, so I can't shoot photos of our own foot):,_Florida,_USA_-feet-8.jpg/800px-Pandion_haliaetus_-Sanibel_Island,_Florida,_USA_-feet-8.jpg,_museum_display-SPL.jpg

    Again, in this sense, dromaeosaur D-II claws are more like ospreys than hawks. I suspect hawks may have the arched cross section as they do more holding down of prey than piercing, whereas ospreys just pierce fish. this looks like I am contradicting myself, since we say that deinonychus is doing pinning. What I mean is that the D-II claw of dromaeosaurs is used for anchoring into the prey (hence it needs to pierce), but that the other digits are adapted to maintain hold, using the bodyweight to pin down the prey (ie. RPR).

    This would all need testing. Maybe some FEA of an arched cross section vs. a narrow one. Manning put a ?velociraptor manus ungual through FEA and they are similarly narrow (like most theropod manus unguals). Or just repeat the stabbing experiments.

    Incidentally, I appreciate this discussion. It's making me thing about claw morphology again.

  27. @henrique - Doesn't that depend on the keratinous sheath, though? Interested to see the research Denver refers to in yesterday's comment.

    @Denver - I also appreciate this; it's really my highest hope when I write a post.

  28. @Denver Fowler

    "Incidentally, I appreciate this discussion. It's making me thing about claw morphology again."

    Glad I could help.

  29. This comment has been removed by a blog administrator.

  30. diablo 3 gold is the game's primary currency, and is used to purchase skills and weapons that are much needed to do battle with demons from the underworld.Which makes it even more imperative to buy diablo 3 gold with which to get the very best gear.Star Wars: the Old Republic (which uses the in-game currency swtor credits) is a groundbreaking sci-fi MMO based on the world-famous series Star swtor credits now at cheap prices from, the most trusted source for SWTOR credits in the entire galaxy.


Trolls get baleted.