The sickle-shaped “killing claws” of dinosaurs like Deinonychus and Velociraptor have captured the imagination for decades. They were held aloft from the second toe, and were far bigger than the neighbouring claws. In Jurassic Park, Alan Grant tells an annoying child that the dinosaurs used their claws to disembowel their prey with slashing motions. That seems unlikely – they didn’t have a suitable cutting edge. Others have suggested that they were used for climbing onto larger prey.
But neither idea made sense to Denver Fowler from Montana State University, who has put forward a very different idea about how these animals used their infamous claws. He compared the feet of extinct dinosaurs like Deinonychus to those of living dinosaurs like eagles, hawks and other birds of prey. Both groups are known as “raptors” and Fowler thinks that they share more than their nicknames.
In his vision, which he calls the “ripper” model, Deinonychus killed small and medium-sized prey in a similar style to a hawk or eagle dispatching on a rabbit. Deinonychus leapt onto its target and pinned it down with its full body weight. The large sickle-shaped claws dug into its victim, gripping tightly to prevent it from escaping. Then, Deinonychus leant down and tore into it with its jaws. The killer claws were neither knives nor climbing hooks; they were more like anchors.
It’s a simple idea, but a potentially important one, for it casts Deinonychus’s entire body into a new light. Fowler thinks that it flapped its large feathered arms to keep its balance while killing a struggling victim. And its feet, which were adapted for grasping prey, would have given its descendants the right shape for perching on branches. Fowler says, “It really helps to make sense of the weird anatomy of these little carnivorous dinosaurs.”
Fowler’s interest began years ago when someone mentioned to him that some birds of prey have a slightly larger claw on their second toe, just like Deinonychus and its entire family – the dromaeosaurids. He soon got his chance to check that. He was going through local museum collections with colleague John Scanella, when he stumbled across a box full of bird of prey feet.
By analysing his collection, and comparing them to videos of their owners in action, Fowler showed that the shape and size of a bird of prey’s feet provides important clues about how it kills. Some use their talons as precision stabbing weapons; others use them to suffocate their prey; and yet others use them as cages.
When Fowler brought Deinonychus into the mix, he found that its feet most closely resemble those of eagles and hawks (the accipitrids) than to other birds of prey. When he looked at specific features of dromaeosaurid feet, the similarities to birds of prey stood out even further. “We saw the same parallel adaptations in the birds of prey and the fossils,” he says.
Deinonychus has a foot adapted for grasping. Its metatarsus, the bone between the ankle and toes, was surprisingly short and stocky, like that of an owl. It would have given the foot a strong grip at the expense of running speed. The fourth toe could have improved the grip even more by moving round to oppose the first one. And the joints in the foot were very mobile, all the better for resisting the struggles of prey.
This idea of Deinonychus sitting on top of small prey seems at odds with classic picture of this predator working in packs to bring down larger quarry. But again, modern birds show how the same grasping motions might have worked against big targets. Golden eagles can kill reindeer. They dig their talons deep into their victim’s back, holding on while the struggling reindeer widens its own wounds. Deinonychus might have used the same strategy to kill larger prey.
Tom Holtz Jr, who studies predatory dinosaurs, says, “Prey-riding is also common in the Galapagos hawk – there’s classic footage of them taking down marine iguanas much bigger than they are. They pin them down, and flap away as the iguanas take them for a ride.” Philip Currie from the University of Alberta also mentions the famous Mongolian “fighting dinosaurs” – a Velociraptor found in pitched battle with a Protoceratops. “It confirms that dromaeosaurids were seeking prey animals of their own approximate body size.”
Not all dromaeosaurids had a grasping foot. Earlier ones like Sinornithosaurus had a relatively long metatarsus. So did the group’s closest relatives – another lot of small hunters called the troodontids. A long metatarsus would have given them a longer stride and a faster sprint. Fowler thinks that dromaeosaurids and troodontids both evolved from fast-running ancestors. From there, they went down different paths. Troodontids took their fleet-footed adaptations even further, and chased down their prey. Meanwhile, the dromaeosaurids sacrificed speed for grasping strength, and evolved into ambush hunters with strong but slower feet.
“The paper provides a very nice explanation for aspects of dromaeosaurid anatomy that have always bothered me and others,” says Currie. “Deinonychus, Velociraptor and their relatives are often referred to as fast and agile, but their hindlimb proportions and peculiarities of their hips are inconsistent with such an interpretation.”
Fowler thinks that his ripper model explains other odd features of dromaeosaurids beyond their feet. For example, their jaws weren’t very strong or sturdy, but you don’t need an especially strong bite or robust jaws if your prey is fully restrained by massive talons. (Hawks and eagles too have relatively weak bites for meat-eating birds).
The ripper model could also explain the weird arms of dromaeosaurids. Their hand bones suggest that the claws could exert a lot of force. However, their arms didn’t have a wide range of motion and they bore long flight feathers that might have got in the way. They look like arms that were adapted for flapping rather than for clawing or grasping.
Fowler suggests that they could have used these odd limbs in two ways. First, they could have circled them around their prey, covering it in a cloak of feathers to either hide it from rivals or trap it even further. Modern birds of prey do this – it’s called “mantling”. From within the mantle, the dinosaurs could have used small movements of their claws to pull back prey that had escaped from the feet.
Second, by flapping their arms, dromaeosaurids could have kept their balance while trying to dismember the prey between their legs. Modern raptors do exactly this – they use their wings and tail to get their body weight pressing down on top of their prey, and then to keep themselves there.
If Fowler is right, his model has important implications for the evolution of flight. The dromaeosaurids would have been very nearly ready for life in the trees. Their grasping feet, with opposable toes, could easily have adapted to grip branches as well as prey. Their flapping arms, used to balance themselves, could have adapted to help them fly. These animals were positively pre-adapted for life in the trees. Perhaps the graceful wings and perching feet of a blue tit got their start with bloody murder on the ground.
It’s an intriguing idea, which gets over the thorny question of “What use is half a wing?” Even wings that couldn’t hold dromaeosaurids aloft would have been useful for “stability flapping”. But Fowler admits that there are problems with the idea. “I’m not sure we’ve quite got there yet,” he says. “We can get to the point where you have all this vigorous flapping but taking that to a mode of locomotion – what’s the evolutionary pressure for that? We’re working on developing that part of the model.”
“As many of these things that involve behaviour and fossils, it probably extrapolates somewhat beyond the evidence but it’s hard to do anything with fossils that doesn’t suffer from that,” says Peter Makovicky, who studies raptors (the prehistoric ones). “It’s better tested than some of the other ideas that have been proposed,” he says.
Reference: Fowler, Freedman, Scannella, & Kambic. 2011. The predatory ecology 1 of Deinonychus and the origin of flapping in birds. PLoS ONE.
Images: restoration by Nobu Tomura; feet by Didier Descouens
December 14th, 2011 at 5:17 pm
If I remember correctly, the “slashing” hypothesis of claw use came from some of the early finds of Deinonychus. I don’t think Creighton invented the hypothesis for Jurassic Park.
December 14th, 2011 at 6:33 pm
I didn’t suggest that – just used JP as an example.
December 14th, 2011 at 7:00 pm
Is it possible that flight evolved from the raptors ambushing their prey by dropping on them from tree branches? Would this plus competiton for food be a possible ‘evolutionary pressure’ or does the term mean something more precise than that? If proto wings were used for balancing it is not a huge step (sic) to directing the drop and from there on to actual manouvering in the air and flight.
December 14th, 2011 at 7:41 pm
Given that earlier, smaller dromaeosaurs like _Sinornithosaurus_ and _Microraptor_ had less specialized feet and they show clear arboreal adaptations, isn’t it as likely–if not more likely–that larger, later dromaeosaurs co-opted their feet and wings for prey capture? That would again bring up the “flightless dromaeosaurs” hypothesis, but in a different context.
I saw this talk at SPV 2010 (I think) and was intriuged then. I really like this idea–it does explain a lot about big dromaeosaurs. I wonder if the really big guys like _Utahraptor_ and _Achillobataar_ practiced the same kind of “mantling” behavior.
December 14th, 2011 at 8:23 pm
@Zach: Sinornithosaurus and Microraptor do not show any unambiguous arboreal adaptations (this was also recently published by Dececchi & Larson, 2011: which is referenced in our paper). In our paper we discuss how grasping adaptations may be useful for an arboreal lifestyle, but they may be more parsimoniously explained as predatory adaptations. Further, these basal dromaeosaurs have a subarctometatarsalian metatarsus which is a cursorial adaptation (something else we discuss in the paper). It is difficult to see how this would be selected for in an arboreal animal.
A correction: basal dromaeosaurs have grasping proportions of the toes, it’s just that they have a long metatarsus, so they wouldn’t have had as much leverage (grip strength).
Thanks for the post Ed. Really good.
December 14th, 2011 at 8:40 pm
Zach, the authors suggest exactly that: that this is an exaptation of a more widespread grasping foot for a particular predator mode.
December 14th, 2011 at 8:51 pm
What beasties do the two feet in the second image belong to?
Oops! Never mind. A flyover shows they’re both Deinonychus.
December 14th, 2011 at 9:51 pm
Actually, the authors suggest both possibilities: either a tree-grasping first or a hunting first. They favor the hunting first, but I disagree with some of their reasons for rejecting it.
December 15th, 2011 at 4:54 am
I’m not sure it’s possible to segregate the origins of tree-grasping and hunting technique in this instance.
Hunting is basal to the group and the specific killing techniques of any given carnivore will be adapted to the environment – in modern taxa members of the same species can use different techniques under different conditions. It may simply be that ambush hunting by a fairly generalist predatory deinonychid in a storeyed habitat led to the coevolution of adaptations for tree grasping and a ‘ripping’ killing technique.
Heck, it may be that the ‘ripping’ killing method would work better with an ambush from an elevated position, increasing the impact force. This would provide a good reason for aerodynamic feather development, as a mechanism for controlling the trajectory of the ambush leap. But can we ever know?
December 15th, 2011 at 2:37 pm
Okay, now I’m a little confused. What came first–grasping feet for branch-hugging or grasping feet for prey capture?
December 15th, 2011 at 5:15 pm
Ooh, I love this. Thank you Ed, thank you Denver, thank you all! My thought when reading was: With the toes specialized for hunting, could the hand claws have been adapted to help with climbing or — stretching it — to grasping onto branches when gliding?
December 16th, 2011 at 1:31 pm
With this ‘dive-bomber’ hypothesis in play, I suppose it’s only a matter of time before somebody names a new dromaeosaur “Stukasaurus” or “Helldiveria”??
December 16th, 2011 at 10:33 pm
Zerosaurus!!!
December 16th, 2011 at 10:43 pm
If these things are ambush predators. Could feathers help them with that? Both lengthening their jump and distracting the prey by making a larger creature suddenly being present and changing the color scheme of the environment. I remember seeing a YouTube video of a cuttlefish doing that to an octopus.
December 16th, 2011 at 11:01 pm
If these things are ambush predators. Could feathers help them with that? Both lengthening their jump and distracting the prey by making a larger creature suddenly being present and changing the color scheme of the environment. I remember seeing a YouTube video of a cuttlefish doing that to an octopus. Could these proto wings help a raptor climb the ribs of a sauropod? Could a pack of these kill a cerotop by jumping and fluttering past the horns and frills, to inflict weakening wounds? Were they the first bronco riders?
December 16th, 2011 at 11:08 pm
I thought I was updating my second post with my third post. Damn iPhone interface. Sorry for the inconvenience.
December 16th, 2011 at 11:17 pm
Could the frill of a cerotops be designed to keep a raptor, or pack of raptors, off the neck, rather than to keep a tyrannosaur from biting into it
December 18th, 2011 at 1:56 pm
I didn’t yet read the entire paper but can also this suggest a hunting strategy from the trees onto to prey by dromaeosaurs with suitable size to arboreal life-style?
December 21st, 2011 at 10:36 am
this answers the creationist myth that a half formed wing (eye in the original) is useless
December 31st, 2011 at 2:56 pm
So Jurassic Park is wrong–Bummer! So is are those great scenes in the latest Godzilla remake where the baby godzillas/raptors are chasing the humans around in the coliseum.
I guess it was bound to be as our understanding of the dinos evolves…
January 15th, 2012 at 2:45 pm
A fossil of velociraptor showed it using it’s large claw to stab the neck of a protoceratops,(died whilst doing it), I figure deinonychus would have done the same, though this makes sense,
the use of the sickle claw could vary between the different dromeasaurids or be used two or more ways by a species even.