Did dromaeosaur claws (both the sickle claws on the feet and the hand claws) have sharp edges for cutting, or were they more like hooks, sharp only at the tip?

Oh boy, you really know how to pick them Bryan. In short, we don't really know. This feels like a big 'palaeontologists don't know' session with your questions, but of course there are lots of tricky areas when it coems to reconstructing fossil animals. In the case of claws there is the underlying bone, and then a keratin sheath (like our fingernails) that grows over the top. The bones are often preserved, but the claws are rare. When they are found, it is becuase the fossil has been preserved unbser exceptional conditions, but it does mean the specimen is crushed flat. So while we might know the general extent and shpe of the claw, if it was sharp or blunt in impossible to tell.

As before there is support for both sides, so I'll briefly try to survey the evidence. Dromeaosaurs have long been thought to have had sharp 'killing' claws, designed to cut open prey in order to kill it. They have, as you note, the enlarged 'raptorial' claw on the foot that was thought to be especially effective for this. Lacking the strong jaws of mammalian predators that typically suffocate their prey, or the big bone-crushing jaws of large theropods or corocodiles that could deliver devestating bites, it was assumed thay would literally have to hack down their food to kill it, and thus needed sharp, cutting claws.

This concept was recently challenged with a paper where a robotic claw was made that could only puncture and not cut. It was therfore concluded that dromaeosaurids would have used these claws to hand on to their prey and then  However, I persoanlly have some rerious reservations about this study. They modelled the claw (that is, the keratin sheath shape, not the undelying bone) on that of modern predatory birds. These of course all have claws designed to puncturer prey as they swoop in on it, so it is hardly surprising that this is what the model did. If you find a loose handle and reconstruct it as a dowel, it is no surprise that it puctures and does not cut, but what about if you but a blade in the handle? Will it cut or puncture? They didn't test it, so we don't know, but I can guess ;-) Using mordern birds is fine in principle, but as they all have the same type or predatory behaviour, that we do not hink the dromaeosaurs shared, it is perhaps not the best model.

Personally, I would go for sharp killing claws. They had to bring their prey down somehow and I don't think that trying to hold onto your prey with your feet as a biped is a very good strategy, and I am unconvinced that a puncture claw could actually peice the very thick skin of many dinosaurs, which would make it either an ineffective or dangerous hunting strategy. Of course, I could be wrong!

Just to say that Dave is not alone in having serious reservations about the recent study suggesting that dromaeosaur claws were used as "crampons" rather than knives.  I don't want to say too much, but I get the feeling at least one response may be in the works.

At least, we know that the bony core of dromie claws were laterally compressed rather than circular or oval in cross-section -- so we can pretty safely conclude that the horny sheath was similarly compressed: blade-like rather than dowell-like.  Those of modern predatory birds are more nearly circular in cross-section.  The part we don't know is how sharp the ventral surface of dromie claws was.

Anyway, the bird-like claws of the "raptors" in Primeval, series 2, episode 1 and horribly wrong.  They made me feel queasy as I watched it :-)

One thing I forgot to add is that the bony part of the claw is often much longer than people realise. The bone tapers to a point as you might expect, but then beyong that is an exceptionally fine needle-like extension of the bone that can add about 40% more length to the bony part of the claw, and this is especially common in  dromarosasurids. This will obviously strongly influence the shape of the claw outside this too.

The projection is so thin and fine it is often lost during preservation or may be accidently destroyed when preparing the fossil. It is so narrow that it would be hard to tell if it was missing, and certainly we only see it in exceptionally preserved fossils, but we would predict it was present in many, if not all dromaeosaurids. It would make the claws longer, thinner and more curved. Whether it would have changed their 'style' from a puncturing to a cutting mode I can;'t really say, but again is soemthing else that must be taken into account and is often missed.

Mike Taylor wrote:

Just to say that Dave is not alone in having serious reservations about the recent study suggesting that dromaeosaur claws were used as "crampons" rather than knives.  I don't want to say too much, but I get the feeling at least one response may be in the works.

At least, we know that the bony core of dromie claws were laterally compressed rather than circular or oval in cross-section -- so we can pretty safely conclude that the horny sheath was similarly compressed: blade-like rather than dowell-like.  Those of modern predatory birds are more nearly circular in cross-section.



True, but cat claws are latterally compressed, and they are used for grabbing rather than cutting.

The dromeosaur work was done here in manchester, the reference is:
Manning, P.L., Payne, D., Pennicott, J., Barrett, P.M., and Ennos, R.A. (2005). "Dinosaur killer claws or climbing crampons?". Biology Letters 2: 110-112. doi:10.1098/rsbl.2005.0395

And the basic article is at the bottom here:
http://www.seaes.manchester.ac.uk/resea … laeo/news/

I'd take a bit of an issue with that Peter, I have had the misfortune to be scratched by a baby tiger and I can assure you those claws tear very well. If you watch lions and tigers fight for example you will see them slash with those claws and leave slashing wounds. They might still be better adapted for holding prey, but they do cut well.

Cats are impressive and efficient killers. They have a morphology that allows them to catch and kill prey with a single bite to the back of the neck, which separates the prey's vertebrae - breaking the neck (in prey smaller than themselves) or they have a throat/muzzle hold that strangles/suffocates larger prey. Their claws are adapted as climbing structures that are only really used on prey if it needs to be snagged to prevent escape or if the prey is very large and requires the cat to climb to its throat.   The claws are not used to kill.

It would be interesting to understand the need for a single large specialised claw on each foot if it is for climbing. Several small hooks are considerably better as climbing tools than a pair of massive hooks, and if large climbing hooks are to be used one would expect them on the forelimbs so the hindlimb claws could take the body weight for the disengagement of the hooks ready for the next step. Dromaeosaurid claws seem over-engineered to efficiently fulfill a climbing function - crampons relate to the size of the climber, not the size of the mountain.

It would be interesting to see if the bite force produced by dromaeosaurids would be sufficient to maintain a killing bite (on prey large enough to need to be climbed) - Manabu, any ideas?

I would predict dromaeosaurids killed by grabbing hold and doing what damage they could by stabbing, cutting and biting until blood loss weakened their prey enough to start feeding on it. This would be best achieved in the more vulnerable underparts - belly, throat etc. Imagine a wrestler with sickles on his boots and sharp teeth. Brutal.

Last edited by Paolo Viscardi (15th Apr 2008 07:27:01)

I know cats primarily hold with their claws and not cut, I was just making the point that they do actually cut rather well. This might not be their adaptive function, but it does demonstrate that laterally compressed claws can be used for more than just puncture/hold actions, unlike a well rounded claw.

The fighting dinosaurs hint that the Velociraptor was using its forelimbs to hold whilst it kicked its long claw into the Protoceratops. Whether the kick was intended to slash or stab is of interest in this example of preserved behaviour.

Pointed stabbing weapons are better at piercing armour and thick hide than bladed cutting weapons. The heavy hide of a herbivorous dinosaur would probably take a lot of effort to cut deeply enough to cause fatal wounds, but by stabbing and ripping back out with a long curved claw quite a lot of damage could be done relatively easily.

I agree with Dave that claws are perfectly capable of holding *and* cutting. I would suggest that they were capable of stabbing and cutting - but stabbing first, to initiate the effective opening of a fatal wound.

Either that or they were used to dig. The giant anteater has heterogeneous, large claws that it carries in a protected position. Maybe dromaeosaurids used them for digging out prey rather than termites.

I was just about to talk about digging claws! These are exceptionally long and very laterally compressed, but never strongly curved (well, nothing like a dromaeosaur, look at tamnaduras, moles and aardvarks) so I think we can rule out digging ;-), though of course it does demonstrate the 'cutting' effect of narrow claws.

I am none too convinced about stabbing to deliver dangerous wounds I must say, they might be deeper, but they are far less likely to hit something critical like and artery or a nerve cluster than a slash, which by definition can cause damage over a much greater area. I could go with a puncture and pull method of delivering damage though, and even superficial wounds can cuase a lot of damage - some major blood vessels / tendons etc. run close to the surface of the skin and would not require a deep wound to affect them.

Well, as to stabbing delivering dangerous wounds I am sure we could talk at length about the age-old slash vs stab debate.

Against an (even lightly) armoured opponent a stab *is* more effective than a slash. The amount of pressure exerted by a point of small area against a surface will be vastly greater than an edge could produce - that is basic physics (pressure = force/area).

In order to get to the vulnerable areas of an animal the skin (plus whatever additional integumentary covering that exists - be it keratin scales or bony scutes) must be breached. A point has the advantage of sliding off hard, high topography areas (scales/scutes) into stable lower topography areas (which would be the softer areas between scales/scutes) and it also has the advantage of being able to concentrate force into a tiny area at the tip, producing immense pressure.

A blade or edge has the disadvantage of sliding across hard, high topography areas onto more of the same - these high areas prevent the lower, more vulnerable, areas to be directly accessed. The edge also suffers by exerting force all along the contact area, thus reducing the pressure that can be applied. So the edge is exerting less pressure on a harder substrate, compared to a point that exerts far greater pressure on a softer substrate.

The strong curve to the dromaeosaurid claw would allow an efficient stabbing action powered by rotational movement. This shape would then allow either a rotational removal of the claw (to prevent snagging) or a linear removal that would result in an additional piercing action from the inside of the skin followed by progressive slicing by the proximal and distal ends of the claw. This would cause considerably deeper damage than a slicing action propogated from the outer surface prior to penetration (and let's face it the vast majority of major blood vessels and organs are deep inside an animal with only a few near the surface), it would also require less force, be less influenced by scales or scutes and it would not have its force reduced by the prey animal moving away from the point of attack - in fact the movement of the prey could improve the efficiency of the action.

I like puncture and pull - it works for me on a basic physical premise.

Last edited by Paolo Viscardi (15th Apr 2008 14:19:11)

Well I think you are overestiamting the skin of the average ornithischian, there may have had scales, but were more like those of small lizards than things like crocs, so would hardly have been a formidable barrier to be penentrated (on average, of course there were ankylosaurs and things knocking around!). Fossil skin impressions (and some real skin) sggest they were not that thick or armoured, (to specualte / extrapolate) no more so that a nice tough skinned deer or anterlope of a similar size.

I think we are talking a little at cross purposes with the stab / slash. Yes a deep (point) would is easier to exert and can reach more vital areas, but they still have to be accurate to ensure a fast kill (not a busted liver where the prey would die, but not for hours of days) but a slash across the tendons in the knee, or at the trachea (if difficult and dangerous to accomplish) would be instantly fatal / incapacitating. I guess it is a degree of personal preference (so to speak!) of which you feel might be more effective a tactic, though of course either is pretty speculative in terms of how it might be executed by a given dromaeosaur on a given potential prey species (cheetahs do hunt rather differently to lions or leopards, even if the actual killing by suffocation is generally the same).

Cats are very specialised in their hunting technique and it strikes me that there is little to suggest any similarity in the hunting techniques of dromaeosaurids to those of cats. With cursorial hunters small slashes in vulnerable areas (belly, anus, throat) and waiting for bleeding out works pretty well (hunting dogs, wolves, hyaenas often use this approach). As you say, the tendons in the leg can be cut - which is something dogs and hyaenas do as well. They don't need the prey to be dead, just immobile enough to allow feeding to commence. I don't see why a fast kill would be expected, especially given the cursorial nature of dromaeosaurids - they don't look like ambush predators that would be unable to keep pace with their injured prey over long distances. All modern predators that are adapted to long distance running kill their large prey slowly.

I would also say that a shallow slash has to be considerably *more* accurately placed to incapacitate or kill prey than a stab, and I would expect that several stabs would be made in the large target of the torso - that happens to be filled with vulnerable bits (look at the position of the hindlimb of Velociraptor in the fighting dinosaurs - it is consistent with a stab to the belly of Protoceratops - if it was a slash I doubt the claw would be at that orientation).

Of course, the prey size that is likely to have been taken is in question. We can see from the fighting dinosaurs that the Velociraptor was having a struggle to subdue the Protoceratops. Size of prey completely alters the technique used to kill it, making multi-use weaponry likely and therefore reducing the validity of hypotheses concentrating on a single adaptive function of the claw concerning prey acquisition.

Paolo Viscardi wrote:

therefore reducing the validity of hypotheses concentrating on a single adaptive function of the claw concerning prey acquisition.

I agree on that. I'm sure there is a combination of adaptive functions that dromaeosaurs had in order to kill/procure prey.

To answer an earlier question you threw at me, I don't think theropod dinosaurs in general (except things like T. rex perhaps) had much of a strong bite compared to their sizes. That is, they most likely had relatively low bite forces for their sizes. On the other hand, the teeth of dromaeosaurs are highly recurved, sometimes the neighbouring teeth even overlapping each other to form a larger "blade" (this is as observed in Bambiraptor). This is something seen to a lesser extent in most theropod tooth line morphology and also superficially very similar to that in the Komodo dragon. Auffenberg (1981, in The behavioural ecology of the Komodo monitor) mentions that "the outer form of the upper tooth row of the ora (Komodo dragon) is similar to that of a curved scalpel blade". Komodo dragons (and other monitors) have extremely low relative bite forces but they are capable of biting through adductor tendons of buffalo limbs. One buffalo was found dead (killed by a Komodo dragon) with the Achilles tendon bitten through and the belly ripped open (Auffenberg 1981). I can't remember the citation off the top of my head but varanids (and presumably the Komodo dragon) use slashing bites...?

So in essence, it is likely that dromaeosaurs used their entire tooth row as a single cutting edge much like a giant scalpel - the teeth acting like serrations on a steak knife with further serrations on the individual teeth as well. Coupled with the likeliness that they had low relative bite force, I imagine dromaeosaurs delivering (perhaps multiple?) slashing bites rather than a single fatal crushing bite.

In this case, it kind of makes sense that dromaeosaur claws were climbing crampons so that they could climb up to the neck or somewhere vulnerable and repeatedly bite the prey. Though, I'm not suggesting dromaeosaurs didn't use their claws to stab or slice. They may have done either as well...

Paulo, I think it's generally considered that dromaeosaurs (despite the name) are not well engineered as endurance runners, and that tyrannosaurids, to name ione group, are more cursorial.  It seems pretty likely to me that many or most dromies were ambush predators like cats rather than endurance chasers like dogs.

I wasn't suggesting dromaeosaurs hunted like cats!!! Just that one can see obvious similarities in gross and some specific morphology (of e.g. a cheetah and a lion) of related taxa (obviously) but this does not mean they were necesarily doing identical things. In other words, we could all be right - those claws could be used by some dromaeosaurs to puncture and others to slash, or combinations of them etc. A good 'generalist' claw will be useful for many things.

Mike Taylor wrote:

I think it's generally considered that dromaeosaurs (despite the name) are not well engineered as endurance runners, and that tyrannosaurids,..., are more cursorial.

What is this based on? From a quick and dirty analysis of hindlimb proportions (dataset from Jones, et al., 2000. Nature 406, 716-718 - see below for link) it strikes me that most dromaeosaurids were reasonably cursorial (i.e. short femur and longer tibia - much more markedly than in tyrannosaurs). The MIII length of dromaeosaurs is, admittedly, shorter than might be expected for a cursor, but how much of that related to the function of the claw and the musculature needed to use it? Overall, the relative leg length is longer in dromaeosaurids than in tyrannosaurids.


http://www.nature.com/nature/journal/v4 … 716ai1.doc

Last edited by Paolo Viscardi (16th Apr 2008 11:27:27)

I think what Mike is going at is that dromaeosaurs are not considered long-distance endurance runners - something necessary fo chasing prey over long distances like wolves - while T. rex is perhaps better adapted for that. It is true what you say about the limb proportions in dromaeosaurs, which is quite similar to those in ornithomimosaurs, so we can assume they were fast runners - but over what distance is a separate matter.

Manabu Sakamoto wrote:

I think what Mike is going at is that dromaeosaurs are not considered long-distance endurance runners - something necessary fo chasing prey over long distances like wolves

Again, I'd be interested to see why this is the case. I added some mammalian carnivore data (taken from
Spoor CF (1985): Body proportions in Hyaenidae. Anat. Anz. 160: 215-220) to the previous set I analysed (as a quick and dirty test) and found that Deinonychus and Velociraptor fell out with the same sort of hindlimb proportions as Canis latrans and C. lupus. Tyrannosaurus and Allosaurus were more similar to Pantera onca and Crocuta crocuta.

I still don't see why dromaeosaurids are not deemed long distance runners - the limb proportions fit as far as my Q&D analysis is concerned. I am happy to change my mind, but I'd like a sound biomechanical reason to do so.

You got me interested, so I emailed Tom Holtz, who is a palaeontologist specialising in tyrannosaurs, asking for clarification.

This just in from Tom: "Yes, you are missing absolute body size. In theropods (as with ungulates), tibiae and metatarsi grow at negative isometry. So young tyrannosaurids have the limb proportions of equivalent-sized ornithomimosaurs (and are far more gracile than equivalent-sized dromaeosaurs).

I have to run to teach in a moment, and will send you more info later."

I'll let you know what else turns up.

Just to add a bit more, dromaeosaurs have a number of features (notably the stiffened tail, but also the body posture of the arms and others) that are interpreted as being very useful for balance and or / turning. Of course a pursuit predator, operating at high speeds might need these kinds of adaptations to help it turn quickly if prey is dodging (something that is actually harder for a biped than a quadruped) and these suggest that dromaeosaurs were involved in rapid pursuit, rather than long distance chases. That might also be interpreted as fitting with their morphology elsewhere - if the object wass jsut to wear down the prey, they would not need the be equipped to deliver severe wounds, their enurance and the occasional nip or claw at the prey would be enough to tire it eventually, but being so equipped to deliver damamge would make sense to get as much injury caused as quickly as possible if they were to tire quickly for example, and risk the prey esacping in a fast pursuit.

Just a bit of specualtion, but it does fit the the other patterns here.

Ok, I see where the tyrannosaurid limb proportion scaling comes in to the comparative analysis with smaller (quadrupedal) carnivores - as I said, my analysis was *very* quick and dirty. But the crux to this is not just about comparing tyrannosaurids to dromaeosaurids - it is perfectly acceptable for both to be long distance cursors (with tyrannosaurs being better at it even).

I think that using terms like "pursuit" and "ambush" predators can be excessively pigeon-holing. It is possible to ambush, cause substantial wounds, back off and follow the weakened animal until it dies - good for hunting large or well armoured animals that are hard to kill outright. Komodo dragons do this sometimes, as do human hunters in some societies. Manoeuvrability is useful for getting in the first attack and then for finally subduing the beastie when it makes its last stand - so there is a good reason for adaptations to manoeuvrability.

I would be interested to know how far the tail is stiffened, does it include the sacrals? Are the sacrals fused?

On another note, I tend to think of ambush predators as being pretty robust, but dromaeosaurs strike me as being a bit on the gracile side. Certainly something like velociraptor seems too flimsy to overpower large prey without sustaining serious injury. They look better suited to killing small prey or delivering multiple swift attacks to something larger. Ambush hunters tend to be solitary (with the exception of lions and cheetahs), whilst most pack hunting animals pursue their prey over longer distances. How does this general trend fit with dromaeosaurs?

Of course this is all speculation, but it's really interesting!

Last edited by Paolo Viscardi (17th Apr 2008 09:21:01)

I'm not syaing they were ambush predators per se, just suggesting that speed and maneouverability might be their 'weapons' of choice, and thus being able to turn at high speed and use all their claws and teeth to deliver soem big wounds quickly might suit this hypothesised hunting style. It depends how they approcah the problem, I was thinking of a cheetah style quick pursuit (at very high speed that does not necessarily require a good stalk as a result) a  quick and severe attack inflicted before the speed wanes and the prey escapes through superior endurance, and then finsihing off a tired / injured animal afterwards.

Of course a co-ordinated, endurance approach as used by hyaenas, hunting dogs etc.  is perfecxtrly possible, but then why the huge array of weapony. If they were relying on endurance and / or large numbers, they only need to keep the prey moving and / or add the occasional small injury to tire / bleed it out. Why the mig manual claws and huge pedal claws just to nip at the ankles?

I'm not trying to pigeon hole them (or other extant predators) as one kind or another (some lions are solitary, others hunt togehter and they will hunt and eat anyhting from mice to elephants via ambush, teamwork, pursuit and even endurance) but most have a 'preferred' or 'best adapted' style and given the discussion here (now entering a new AAB record length) I think some kind of fast pursuit fits the (very) limited data we are specualting on.

Hmmm, could this be any less rigorous an anlsysis?

Probably not.

The following is from Carrano (1999) on limb proportions and relative speeds:

"Pendulum theory models the walking limb as an inverted pendulum that oscillates around the foot during the support phase (Gregory, 1912; Cavagna, Saibene & Margaria, 1964; Cavagna, Heglund & Taylor, 1977). Since the period of oscillation decreases as mass is increased further from the point of rotation, this model predicts lengthening of the distal limb in order to facilitate movement of muscle masses farther from the point of rotation. In other words, since the limb rotates over the foot during the support phase, it will operate as a more ef®cient inverted pendulum if muscles are concentrated at the hip joint, rather than more distally along the limb. Lightening of individual limb elements would also increase the relative proportion of total mass represented by the proximal musculature, thereby accomplishing the same goal. Additionally, more slender limb elements would create less inertia to be overcome by muscular exertion during the swing phase of each stride (Fedak, Heglund & Taylor, 1982).
As a result, the following features have been predicted in the limbs of animals designed to move with greater speed at less energetic cost (cursorial in the traditional sense): more slender individual limb elements, longer distal limb segments, and hip muscle insertions more proximally placed along the limb. A complementary set of features (more robust individual limb elements, shorter distal limb segments, and more distally placed muscle insertions) have been predicted to occur in the limbs of animals designed to move with lower speeds, but using more limb power per stride (graviportal in the traditional sense). `Cursorial' and `graviportal' are therefore endpoints, and not categories, that bound morphological-mechanical variation in limbs (Carrano, 1997)."

According to this paper, dromaeosaurs are "small-bodied intermediate" whereas tyrannosaurids are "cursorial" along with troodontids and ornithomimosaurs. However, Carrano thinks tyrannosaurs were not pursuit hunters - as in the conclusions of the same paper:

"In contrast to modern mammals, however, where large herbivores and carnivores of several lineages independently acquire cursorial adaptations, large carnivores and small herbivores are the most cursorial among the dinosaurs. No large dinosaurian herbivores appear to have been cursorial; most, in fact, show graviportal specializations associated with the acquisition of quadrupedal posture and a substantial increase in body size. This suggests that large theropods with cursorial adaptations were likely adapted to covering large home range areas and did not evolve as pursuit predators on their large, herbivorous counterparts."

Tom Holtz warns us trust the numbers of Jones et al. at our peril: they did not measure specimens, and many of their measurements are "inferred from line drawings by Paul", even when Greg was doing reconstructions putting in missing elements. For example, they give limb dimensions for Carnotaurus, for which the leg beneath the proximal tibia is missing!

Tom offers the following two graphs that plot femur length against metatarsal length:
http://www.miketaylor.org.uk/tmp/MtvsFem.jpg
and against tibia length.
http://www.miketaylor.org.uk/tmp/TibvsFem.jpg

As you'll see, the dromie line (lilac diamonds) is well below the tyrannosaur line (magenta squares) in the former, and pretty much coincident with it in the latter.

Yay!  Go, cursorial tyrannosaurids!

Just to add some unnecessary (or necessary actually) complication to this debate, it also at least partly depends on your definition of 'cursorial'. Most people (I think) tend to use it to mean 'capable of running', but of course there are huge differences between a slow jog and a sprint (both in terms of speed and the gait used) though either could be considered 'running'. I think it is generally accepted that the large tyrannosaurs (and probably other giant theropods) could not get into a sprinting gait (both feet off the ground at once) so we need to be careful about comparing taxa as both being 'cursorial' when it can cover a multitute of variation.

Not having a go at Mike, jsut post poiting out that when trying to compare a (probable) 8 ton Tyrannosaurs with a 35 kg Velociraptor, a word like 'cursorial' might not be the best one! :-)

David's right of course.  The most precise thing we can probably say is something like this: considering the allomtry of various limb segments across the various theropod lineages, we would expect a dromaeosaur ("raptor") that was as big as T. rex to be less well adapted for running.

Sorry to come back to this post, but it's really interesting!

Looking at the graphs it is clear that the dromaeosaurs have shorter MIII than tyrannosaurs - fair enough - but I did actually say in a previous post "The MIII length of dromaeosaurs is, admittedly, shorter than might be expected for a cursor, but how much of that related to the function of the claw and the musculature needed to use it?"

Tyrannosaurs are not usually thought of as doing anything but walking or running on their hindlimbs. Dromaeosaurs are obviously doing other things as well - hence the big claw (which started this record-breaking post off in the first place!). As a result there are complications in using simple comparative techniques to assess locomotion. The added mass of the claw and musculature needed to use it would be located at the very end to the limb, putting greater stresses on MIII during locomotion. I would suggest that having a longer tibia would allow greater stride length without compromising MIII safety limitations.

The graph of tibia vs femur shows the dromaeosaurids as having relatively long tibias in relation to other theropods of similar mass - conisitent with the expectations of MIII being constrained by the presence and function of the claw. Perhaps other features of ambush and pursuit predators could be considered to help address the question of hunting technique, which would then offer support for one of the claw use hypotheses out there (or in here)?