Yesterday I covered the basic introduction to my new paper about a Velociraptor specimen with an azhdarchid element preserved in it’s gut. Today I want to move on from the basics (what is there) to what this potentially means and how this is inferred. Most of my recent research is based around theropod ecology and behaviour (like this, this and this for example) and specimens like this one can provide new information and evidence for how these animals were acting. The obvious question here is why is this inferred as scavenging and not predation? As usual with such questions going so far back in time, it’s hard to be definitive, but this is the better supported inference.
First off there is the relative sizes of the animals. While it’s not unknown for predators to tackle other predatory animals, or relatively big prey it’s certainly not normal. Lions don’t routinely hunt leopards or bears go after wolves. This is relevant here since azhdarchids were most likely active predators themselves and so a potentially dangerous animal to attempt to kill. Moreover, the azhdarchid in question was most likely 9 kg in weight with a 3 m wingspan (and could have been considerably larger), while the Velociraptor was a sub-adult of around 13 kg. In short if this was a predation it was no mean feat – perhaps the equivalent of a small coyote bringing down a big eagle. Sure it’s possible, but it’s not unreasonable to think this was really very unlikely. It’s more likely this was a young carnivore scavenging on the carcass of a dead pterosaur, as indeed was inferred for a similar previous specimen from Canada.
Even if we assume that it was a kill, other things don’t add up well to support this. Theropods don’t tend to consume large amounts of bone like this. They might consume relatively large items (like a whole small prey item) but not large chunks of bone like this. And it is a pretty big chunk of bone, probably the same length as the skull of the dromaeosaur. Moreover, we also know that theropods can be really quite delicate feeders, including other velociraptorines. The tendency seems to be to scrape meat free of the bones, now chew up and swallow whole ones (like modern birds of prey, they’ll swallow a mouse, but will pull chunks off of rabbit or sheep). Carcass consumption patterns by modern vertebrates also show that whole big bones like that don’t tend to be swallowed. Finally, the pterosaur weighted at least half and potentially more than the dromaeosaur. Given their apparent skill at stripping a carcass of meat I don’t think I dromaeosaur would be swallowing whole bones (and ones that would be pneumatic, not filled with marrow) when much of it’s own weight was sitting there in muscle and viscera.
In short, predators don’t normally predate other predators. Predators (including theropods) don’t usually seek out large prey. Predators (including theropods) don’t usually consume large bones of large prey unless they are a bone specialist or there’s nothing left. Even when there’s not much meat left, theropods tend to scape this free to eat rather than swallow bones. Sure all of these could hit the ‘least likely’ option and it’s a who-knows-what to 1 chance that a small dromaeosaur took on a big azhdarchoid, killed it and started swallowing big bones. But it’s far more reasonable to infer that it scavenged the last bit of a carcass it chanced across.
We are then left with scavenging as the most likely explanation as to why this animal was swallowing whole bones. Interestingly, we do also see shed teeth being a common feature of dromaeosaur (and indeed theropod in general) feeding yet here every tooth in the skull is intact. That is admittedly merely a soupscon of evidence for scavenging, but one might well expect a tooth or two to be lost during a fight with such a big adversary. or even biting through bones to swallow them again suggesting it just picked up and swallowed what it could find without much or any oral processing.
Uncoloured version of Velociraptor feeding. Courtesy of, and copyright to Brett Booth.
Moving on from this issue then, what does this tell us about the ecology of dromaeosaurs? Well to degree, not much we didn’t know already. There’s already evidence for both predation and scavenging in the dromaeosaurs, and indeed already evidence they were eating pterosaurs. Even so, more evidence is always good, and it does at least reinforce the existing evidence we have. It also therefore takes us a little further away (sadly) from the idea that dromaeosaurs were some kind of hyper-carnivorous super-predator that spent their time knocking down huge prey items with all their claws and teeth. I say sadly, because it’s a great idea and a wonderfully romantic notion, but sadly these animals were every bit as opportunistic as other carnivores and clearly were not beyond taking the odd, or indeed regular, free meal through scavenging. Indeed given the number of specimens we now have supporting a scavenging interpretation, this does seem to have been a pretty common part of their behavioural repertoire as carnivores.
Dave Hone's Archosaur Musings 
Thank you for this two-part series! It makes sense to me, I don’t know of many living carnivores that will turn up their nose at carrion unless it’s too far rotted for their metabolism. Scavenging is part of the usual predatory repertoire. The way you showed how you inferred it was great.
Thanks too for the beautiful artwork by Brett Booth. I enjoyed seeing both the colored and uncolored versions. That’s one of the nicest velociraptor pictures I’ve seen in quite a while.
On the latter note, I was delighted Brett was interested. He was kind enough to go a rush, but very good, job on the last Velociraptor feeding paper and I thought it would be cool to get him in again and sort of pair-up the two illustrations of scavenging. He was very keen which speaks volumes for his interest and kindness to do this work. And well, the results are awesome.
A nice find, and it’s wonderful you take the time to elaborate the study this well for the popular audience after sure spending a good number of hours writing the actual paper.
As a less likely and unproveable, but maybe plausible scenario: could it have been that a group of adult raptors had taken the pterosaur down and ate the best parts, while our subadult had to wait and eat the bony scraps? Yeah, I know, no evidence of pack-hunting in non-avian dinosaurs. It’s just such a romantic notion, as you put it.
And truly wonderful artwork. I like the black & white version even better than the colored one. It has more life in it, as infinished artwork tends to have.
As to pack hunting, it’s not impossible of course. But aside from the general issues (there is no really good evidence really of pack hunting for any given species) I’d still favour scavenging in this case becuase of the issues of feeding traces and the general rarity of predators tackling big prey / predators, even in groups. Definitive evidence for packs of dromaeosaurs working together would certainly swing the evidence away from this interpretation a bit, but for my money, not enough to make hunting more likely than not.
There’s been a lot of discussion about this paper over the last couple days on the DML. It all boils down to the evidence for predation being just as likely as scavenging.
First we have the size of the pterosaur. Mark Witton (http://dml.cmnh.org/2012Mar/msg00039.html) says “The pterosaur material described in the paper is a scrap of bone – 75 x 12 mm – so we have no idea what part of the body it represents. How do we know how large the pterosaur was if we don’t even know what bone we’ve got? The paper tells us it was likely exceeded 100 mm in length: why should we assume that? Assuming it represents a limb bone of some kind, conservative wingspan estimates based on its proportions are 2 – 3.6 m. A 2 m span azhdarchid may weigh as little as 4-5 kg, less than half of the Velociraptor in this instance. The authors of this paper favour a 3 m span for the animal, but do not say why this is the preferred span estimate. The relative masses of the two animals discussed in the paper is a critical part of the scavenging story, so it needs to be more substantially verified to give credence to their idea. If this paper favoured the lower (and more conservative) estimate, the body sizes of these two animals would be far more disparate and the scavenging argument weakened considerably. Alternatively, that scrap could represent a part of a larger bone from a giant pterosaur limb girdle or cranium, which would make the scavenging notion much stronger. We just don’t know: it’s a _scrap of bone_, so we need to be realistic about what stories it can tell us.”
So the size of the prey was highly uncertain, but even your measurement puts it at 70% the predator’s mass. You say killing such prey is “really very unlikely”, but where’s the data showing this is uncommon in modern predators? Danver Fowler notes “Birds of prey will fairly commonly take prey that are subequal in size to themselves (which would be the case with a 13kg Velociraptor vs a 9kg pterosaur). Peregrines take a lot of pigeons, hawks take rabbits; accipitrines take all manner of large birds their own size, or sometimes slightly larger.” Wikipedia states 20-90 kg African leopards “focus their hunting activity on locally abundant medium-sized ungulate species in the 20 to 80 kg (44 to 180 lb) range.”
Second, you claim “While it’s not unknown for predators to tackle other predatory animals … it’s certainly not normal”, but again you present no data, and indeed don’t state this in the paper itself. I’m no ecologist, but this seems absurd on its face, as it would leave predators to only die of disease, starvation, accident, etc.. And indeed, using leopards again for convenience, they’re killed by lions, tigers, crocodiles and such. Carnivores are not necessarily more dangerous than herbivores- do you really think an azhdarchid was more dangerous prey than Protoceratops?
Third, you claim theropods don’t usually eat large chunks of bone. This fragment is 75 mm long, so given the Velociraptor’s 192-194 mm long femur, was not “probably the same length as the skull of the dromaeosaur.” The Fighting Dinosaurs specimen of Velociraptor has a skull ~111% of femoral length, so the “scavenging” individual would have a skull ~214 mm long, which is 2.85 times longer than the fragment. This is a small fragment compared to e.g. the dromaeosaurid tibia Sinocalliopteryx’s holotype ate (155 mm for a specimen with a skull ~270 mm long). Or there’s Scipionyx’s holotype that a lepidosaur foot half its femur length. While you say “Carcass consumption patterns by modern vertebrates also show that whole big bones like that don’t tend to be swallowed”, you provide no references to to any studies, and I’d be surprised if crocodilians, Komodo dragons and such don’t tear off limbs and swallow the bones they contain whole.
Finally, you say “Given their apparent skill at stripping a carcass of meat I don’t think I dromaeosaur would be swallowing whole bones (and ones that would be pneumatic, not filled with marrow) when much of it’s own weight was sitting there in muscle and viscera.” But how do you know this Velociraptor DIDN’T have a belly full of meat in addition to the bone fragment? It’s not like any flesh would be preserved in the fossil.
So basically, the paper presents a lot of assertions without the data to back them up. There aren’t even any references to modern animal feeding habits in the bibliography. Maybe it did scavenge the bone, but until you show modern predators a) usually don’t kill animals half their mass or more, b) don’t often kill predators (taking into account carnivore/herbivore population differences), c) usually don’t swallow bones that measure a third of skull length, we can’t support either hypothesis with any confidence. The tooth loss point is also moot until someone has studies of crocodilian tooth loss per kill, and even there the different morphology and dismembering behavior of crocs would make any comparison questionable. I’d also like to know exactly which scientific references portray dromaeosaurids as being less prone to scavenging than other theropods/carnivores. The paper references Ostrom (1990), but he merely describes their ability to kill without bringing killing/scavenging frequencies into it.
I also have been curious about some valid points brought on the DML. It would be quite neat if you had the time to answer some of these, David.
Ok there’s obviously a lot here and so I’ll try to be brief.
1. On the size. As the paper does note, the bone is a bare minimum of 75 mm. 100 mm is considered a more likely minimum however as there is no tapering to the bone that we can see or any evidence of shapes like DP crests, articular ends or whatever. There’s not many bones in a pterosaur that are that long and straight with no changes to the shaft etc. Maybe the tibia, but even a 75 mm tibia is quite big. There are more elements that are likely to give a minimum figure of a 3 m wingspan that would be 2. Even at 2 m, and a 3 kg pterosaur, you still have a decent amount of muscles etc. and that does not get past the overall general issue of what we know of theropod (and including velociraptoribne feeding) vs bone swallowing. i.e. you can drop the size and mass a fair bit and I would still argue it doesn’t really change the overall arguments.
2. Yes birds and other predators can take big prey. I’m saying *in general* most don’t. And that stuff is well cited in the Hone & Rauhut paper. Also I’d argue that using birds as a model for this is perhaps a bit dodgy given how they can be quite light for their size. In neither case though does this affect the issue of the fact that azhdarchids were most likely predators. Even a 2 m wingspan azhdarchid would I’d argue, be a challenge for a less than 2 m dromaeosaur. While obviously some animals do tackle odd and dangerous prey (mongooses on cobras) it’s not common. Certainly I can’t think of any animal that makes a habit of it – the chance of injury is greatly increased and well, predators are generally quite rare. I really don’t think you need do some form of ecological analysis to demonstrate that predators are rare components of ecosystems and that taking on another predator your own size with claws and teeth might lead to more injury than tackling the young of a herbivore.
3. OK yes I got the skull length vs bone length wrong. But that’s only on here and NOT in the paper. Bad estimate, my bad. Yes there are a couple of examples of bone being swallowed by theropods (Sinoclliopteyx I would call swallowing a large prey item, not quite the same thing) but these are rare (again, see Hone & Rauhut where I have a whole discussion of this issue). There are far more examples of scrapes than swallows for example. As for big bone piece swallowing without oral processing I’m not aware of any species that does this apart from those well suited to that task (like hyena) and these engage in lots of oral processing. Even n things like komodo dragons do swallow big chunks of bone, these are not animals that we know of that do or are even capable of delicate scrape feeding etc as described for various theropods. Thus I don’t think it’s a fair analogy / doesn’t deal with the evidence we do have for their feeding.
4. No, I don’t know it ‘didn’t have a belly full of meat’. But nor do I accept that if it did, it would then start trying to swallow a large bone. Again, that’s no a behaviour commonly associated with dromaeosaurs or indeed theropods *in general*. This is an isolated shaft of a pterosaur bone, not the kind of thing that generally has either marrow in it, or lots of muscles attached (the ends or processes might be a different issue). Would it really fill up on muscle and viscera and then start to swallow a bone like this? I think it unlikely. Why no oral processing or scape marks? That’s a far more common behaviour.
“So basically, the paper presents a lot of assertions without the data to back them up. There aren’t even any references to modern animal feeding habits in the bibliography”
Not true. As I do cite my own Hone & Rauhut paper and Hone et al on Velociraptorine feeding and Hone & Watabe of trannosaur feeding extensively. I do this rather than cite original papers on these subjects as frankly all of those papers were somewhat reviews (the former especially so) of theropod feeding ideas and include lots of stuff of extant taxa feeding etc. If you have written a thorough review I don’t see why there is an issue citing that review rather than all the primary data that wen into them. If you have read those papers they do, I would argue (obviously) provide a framework for such interpretations which I follow here.
For example, sample paragrpah plus citation on extant taxa hunting preferences from H&R 2010: Predators select for the young,
the old and the weak or injured (Palmqvist et al.
1996), or at least their presence greatly enhances the
ability of predators to hunt successfully (Temple
1987). Modern and recent mammalian predators
(Kru¨ger et al. 1999; Husseman et al. 2003; Steele
2004) and birds (Donazar & Ceballos 1989; Boshoff
et al. 1994; Rohner & Krebs 1996) certainly prefer
these kinds of sub-optimal fitness prey. Small prey of
a given species is also typically preferred over larger
individuals by such diverse predators as leopards
(Hayward et al. 2006), hunting dogs (Fuller & Kat
1993), bluefish (Scharf et al. 1998) and several invertebrates
(Barbeau & Scheibling 1994) or simply small
prey in general (e.g. Webb et al. 1991; Gotmark &
Post 1996; Turesson et al. 2002).
I think that deals with all the substantive points and I think suggest my position is not as weak as you suggest and nor that I have not doen the research or reading as you suggest.
Thanks for the detailed reply.
1. You agree the pterosaur’s size is somewhat ambiguous, so far so good. You say it’s not that important due to bone swallowing, which is discussed below…
2. For the size of prey, you indicate references in Hone and Rauhut (2010), of which you reposted here. Since we don’t know the age, fitness or health of the azhdarchid, those preferences don’t help us with the current specimen. Similarly, the azhdarchid could be a small individual of its species for all we know. For prey size in general, your last three listed references (on crocs, hawks and pike) are unfortunately all behind paywalls. Not that the pike study is very relevent, since pike can’t eat prey larger than their mouth regardless. And looking at Radloff and Du Toit (2004- http://courses.biology.utah.edu/goller/7406/Goller7406/duToitPdfs/Largepredators_2004.pdf), figure 1 shows that prey between 50-100% of predator size (as in the current situation) makes up 35% of cheetahs’ kills, and 30% of leopards’. Indeed, the mean prey mass listed in table 2 for cheetahs and leopards is 60-70% of predator mass, which almost exactly matches your proposed ratio between the azhdarchid and dromaeosaurid.
I think we need more than our intuition that carnivores are more dangerous to predate than herbivores. Again I ask if Protoceratops with its strong shearing jaws and robust build was really less dangerous than an azhdarchid with slender hollow bones, vulnerable wings and jaws that “were ill suited for demanding feeding techniques or for subduing large, struggling prey, and were better adapted for handling relatively small or immobile food items” as Witton and Naish (2008) put it. More empirically, Radloff and Du Toit show that one of the most common prey items of savanna carnivores are warthogs- omnivores with dangerous tusks.
3. For the frequency of bone swallowing in theropods, I think the Hone and Rauhut paper is right when it says “even in largely articulated specimens, the ribcage is often broken open and prey bone fragments might have been scattered outside the ribcage. Thus, they may simply have gone unnoticed or unreported in the past as unrecognizable and undiagnostic bone fragments, if collected at all, are often not studied in any detail during the study of a fossil.” It goes on to say that theropods “would have had long digestion times due to the length of their gut, and there is evidence that at least in the stomach there was a highly destructive acid environment.” The conclusion would seem to be that as currently sampled at least, the fossil record cannot be expected to show how often theropods ate large bones. That they didn’t break large bones often with their teeth like mammals doesn’t mean they wouldn’t have dismembered prey like crocs and Komodo dragons. You claim Komodo dragons aren’t comparable since they can’t scrape-feed, but surely theropods could have both scraped and ingested bone depending on the circumstance.
4. The entire conversation of whether Velociraptor would eat a bone fragment in addition to a good amount of flesh just has too many assumptions and unknowns to begin to evaluate scientifically. How discerning would a dromaeosaurid be (keeping in mind it’s less intelligent than birds or mammals), where was the bone fragment located, how did it break in the first place, was the dromaeosaurid in a rush due to other predators? Unanswerable questions. I could just as easily argue that if it was eating the bone out of desperation from a carcass, as your picture shows, that it would have probably eaten more than just one bone fragment. But that involves just as many assumptions as your proposal does.
“How discerning would a dromaeosaurid be (keeping in mind it’s less intelligent than birds or mammals),”
I wouldn’t say that, given Walsh & Milner 2011. For the sake of saving space, see the following quote for a shorter/less technical version of what Walsh & Milner said in said paper. Otherwise, good debate.
Quoting Gardom & Milner ( http://www.amazon.com/Natural-History-Museum-Book-Dinosaurs/dp/184442183X/ref=sr_1_1?s=books&ie=UTF8&qid=1325974924&sr=1-1 ): “A further clue to the brain power of dromaeosaurs comes from Bambiraptor, a diminutive sub-adult dromaeosaur from the Upper Cretaceous of Montana, USA that was less than a metre long and may have weighed about 2 kg. Bambiraptor had one of the largest-known dinosaur brains relative to its body size. The areas of the brain that deal with agility, co- ordination, intelligence and sight were enlarged and very bird-like. This suggests several life styles might have been possible – pack-hunting where co- operation and communication between individuals would have been vital as we saw in Chapter 6, or tree-climbing, as appears to have been the case in some other small dromaeosaurs described in Chapter 10.”
“I wouldn’t say that, given Walsh & Milner 2011”
Sure dromaeosaurids were intelligent for non-bird dinosaurs, but I’m just saying that some of the statements (like Velociraptor wouldn’t swallow a bone fragment if there were lots of meat available, or Velociraptor would scrape a bone before it was desperate enough to swallow it) imply a sort of rational process that humans might follow. But for an animal with less intelligence than a chicken, surely it’s at least as probable this Velociraptor was just tearing convenient pieces of flesh off and getting them into its stomach before competition showed up, without much thought as to which particular bit was most nutritious or what the most efficient strategy of stripping meat would be. We just don’t know, and this makes statements like Hone’s “Would it really fill up on muscle and viscera and then start to swallow a bone like this?” impossible to answer and thus irrelevent to the question at hand.
2. “Since we don’t know the age, fitness or health of the azhdarchid, those preferences don’t help us with the current specimen”. True but we never know that for anything like this. Again though it comes down to what is mostly likely. I think the cheetah / leopard examples are potentially misleading. Behaviour is hugely varied and it’s not going to be that hard to find a couple of exceptions to any given rule. The issue is, do most predators ever regularly go after animals close to their own size? And that appears to be a no. And I do still maintain that predators don’t tend to go after other predators – remember that most of them favour juvenile prey (H&R again) which will have underdeveloped jaws / tusks / horns etc. And suffer all manner of behavioural and ecological penalties that make them vulnerable to predators (a large part of why they are targeted).
3. Yes but I say bone fragments and thus wrt oral processing (as seen in e.g. t. rex coprolite) and not whole bones (or big chunks of them). And I think you have this the wrong way around. Even after oral processing of a juvenile bone (i..e less ossified, smaller), these pass through even the long gut of a t. rex still intact enough for them to be identified. Is that’s the case even for this condition which should maximise bone digestion then in fact if theropods were eating bones regularly we should find them in the gut way more often than we do. Unless of course, they’re not eating them often and instead doing realtively delciate feeding jobs. Cats lick bones clean, dogs gnaw them, big raptors pluck the meat off. Komodos and crocs swallow big chunks (thereopods were not beyond this, like sinocalliopteryx or Microraptor) but again there;’s a difference between swallowing a leg and swallowing a tibia.
4. I don’t think this is anything like as speculative as you suggest. Hone & Watabe suggest that actually they were quite accomplished feeders as shown by the different techniques employed on different parts of the bones. Animals might not be analytical or think through things, but they are also optimised by evolution to be logical. Even komodos seem to follow the same carcass-consumption patterns of mammals of going for the hindquarters and belly first and things like the feet and head last. If they can do something that complex, i don’t think it’s a struggle to argue that an animal would not start by eating bones or even eat them after a big meal of meat etc. It’s not an optimal strategy.
Do we need more studies on these kinds of things? Sure. I’ve even spoken to a couple of guys working in this field and suggested more things that could be observed / studied to add to the data pool for interpreting fossils. But based on what we do have, i still think all of this falls on the *most likely* side of scavenging rather than predation.
In short, it’s perfectly plausible that this was predation, i don’t say otherwise. I don’t think its the most likely explanation as still for me all of these possibilites lie of the lesser side of what’s most likely. Less likely it went for a big prey item, less likely it went after another predator etc.
2. “Behaviour is hugely varied and it’s not going to be that hard to find a couple of exceptions to any given rule. The issue is, do most predators ever regularly go after animals close to their own size? And that appears to be a no.” This still lacks supporting data as a generalization. Even if your three references from Hone and Rauhut 2010 do actually show pikes, crocodiles and hawks only rarely eat prey greater than half their mass (which I still don’t know, since they’re paywalled), that’s three examples compared to my two examples of leopards and cheetahs. Note lions and wild dogs also usually go for larger prey according to Radloff and Du Toit, but I left those out since they hunt cooperatively. Then again, maybe Velociraptor did too. It’s certainly been suggested for Deinonychus. Specifically regarding dromaeosaurids, the Fighting Dinosaurs Velociraptor seems to have a prey item larger than itself, and Fowler et al. (2011) suggested dromaeosaurids were adapted to tackling large prey. Maybe large cats are exceptions, maybe the Fighting Dinosaurs Protoceratops was dead or trapped as Osmolska 1993 proposes (but maybe the azhdarchid was trapped too, again who can say), and maybe Fowler et al.’s Raptor Prey Restraint model is wrong. The bottom line is that with such variation and suggestive details, I don’t think we can use “Velociraptor probably mostly hunted small animals” as a proposition to support any hypothesis.
As for you maintaining predators usually don’t go after predators or adults, I’ve yet to see any reason the azhdarchid was an adult or refutation of Witton and Naish’s point that their beak wouldn’t be effective against large animals. Azhdarchids were predators in the sense that storks and cranes are, so this would be like an average female bobcat (6.8 kg) killing a big white stork (4.4 kg). And indeed, bobcats (average 6.8-9.6 kg) are the major predator of whooping cranes, which are 6-7 kg. So again we have modern predation data that strongly matches this specimen.
3. To an extent, I think this issue is a red herring, since regardless of how common it was for theropods to eat bone, in none of our specimens do we know whether they did this when predating or when scavenging. Even if we agreed it was very rare, it could be done at those rare times to get minerals, and that’s just as easily taken advantage of in killed prey as scavenged prey. So until you establish that eating bone in theropods most likely happens due to desperate scavenging as opposed to other reasons like mineral gain or accidental uptake, it really doesn’t matter how rare it was. That’s something that could be studied in modern carnivores that rarely eat bone, of course.
4. I guess we’ll just have to disagree on the degree of speculation here. Sure Komodo dragons have instincts to go for certain parts of prey over others, but how often do they make exceptions, what if other factors like competition with other carnivores is taken into account, etc.. I’m not at all confident that assuming the bone fragment was obvious to the dromaeosaurid and it realized that it was suboptimal to eat, that Velociraptor would follow a certain eating order rigorously enough and under enough varying conditions for us to say that eating the bone fragment was an improbable outcome given a predated carcass.
“But for an animal with less intelligence than a chicken,”
Correct me if I’m wrong, but I was under the impression that chickens (unlike dromaeosaurids, as implied by Gardom, Walsh & Milner based on brain structure) have neither the intelligence for pack hunting nor for arboreal gliding.
“surely it’s at least as probable this Velociraptor was just tearing convenient pieces of flesh off and getting them into its stomach before competition showed up, without much thought as to which particular bit was most nutritious or what the most efficient strategy of stripping meat would be.”
If anything, that sounds more like croc behavior than bird behavior ( http://www.youtube.com/watch?v=o-3zfBn2JcM&feature=plcp&context=C30c0d2cUDOEgsToPDskLnJrihleUfPkJzMJSC1KtA ).
All modern bird brains, including those of chickens, have much larger cerebra (=telencephalons) than Archaeopteryx or dromaeosaurids like Bambiraptor. Milner and Walsh (2009) describe this in detail. Regardless of what The Natural History Museum Book of Dinosaurs says, we can’t tell if Bambiraptor had the intelligence for pack hunting, and indeed some people think dromaeosaurids lacked it and used a simpler method of disorganized group hunting like Komodo dragons. As for arboreal gliding, chickens can fly and have a much more sophisticated wing than any dromaeosaurid, so I don’t see why you’d think dromaeosaurids would be superior in their aerial calculations. Note too many less intelligent animals can glide, like Draco and flying frogs. I’m no neurologist, but I bet that’s far more the job of the cerebellum than the cerebrum.
And yes, my point with suggesting that feeding behavior for dromaeosaurids was their intelligence seems to be between that of crocs and birds. Thus we shouldn’t assume they’d behave as intelligently as birds and mammals, and indeed might resemble crocodilians more in some behaviors.
“and indeed some people think dromaeosaurids lacked it and used a simpler method of disorganized group hunting like Komodo dragons.”
Assuming you’re referring to Roach & Brinkman, I’d be surprised if anyone who knew better still took their 2007 paper seriously, given its many problems.
“Note too many less intelligent animals can glide, like Draco and flying frogs. I’m no neurologist, but I bet that’s far more the job of the cerebellum than the cerebrum.”
I would think that arboreal gliding mammals would be better (though not perfect) analogs, given that said mammals (like microraptorines, but unlike arboreal gliding herps) are highly-active endotherms. As for the relevance of the cerebrum, see Walsh & Milner 2011 (which also implied that pack hunting was probably w/in the cognitive ability of “Cretaceous nonavian maniraptoran theropods”).
“And yes, my point with suggesting that feeding behavior for dromaeosaurids was their intelligence seems to be between that of crocs and birds. Thus we shouldn’t assume they’d behave as intelligently as birds and mammals, and indeed might resemble crocodilians more in some behaviors.”
If I’m understanding you correctly, you’re saying that dromaeosaurids (despite having more bird-like brains) had more croc-like behavior.
“Assuming you’re referring to Roach & Brinkman, I’d be surprised if anyone who knew better still took their 2007 paper seriously, given its many problems.”
Where are these problems published or posted online (not on JP fanboards I assume 😉 )?
“I would think that arboreal gliding mammals would be better (though not perfect) analogs, given that said mammals (like microraptorines, but unlike arboreal gliding herps) are highly-active endotherms. As for the relevance of the cerebrum, see Walsh & Milner 2011 (which also implied that pack hunting was probably w/in the cognitive ability of “Cretaceous nonavian maniraptoran theropods”).”
Why would endothermy or activity have any relevence? In any case, according to Walsh and Milner, “at present there seems to be no consistent relationship between flying ability and either brain size or morphology.” Thus there’s no known way to say from its brain that Bambiraptor “had the intelligence for arboreal gliding”, contra your claim.
They also never imply dromaeosaurids had the intelligence to cooperatively hunt. Instead, they say cerebrum size is connected to behaviors like pack hunting in modern birds. But we don’t know if the cerebral expansion in coelurosaurs allowed such behaviors, or if the cerebral expansion in Neornithes did. And indeed, as Jura points out below, even talking about the problem this way is probably much too simplistic and hypothetical given our poor understanding of how intelligence and ability relates to brain morphology.
“If I’m understanding you correctly, you’re saying that dromaeosaurids (despite having more bird-like brains) had more croc-like behavior.”
No, you’re still not getting my point. Dromaeosaurid brains were (roughly) intermediate between crocs and birds. You’re only thinking of it one way- that they were more bird-like than crocs. But while true, you can also say that they were more croc-like than birds. So just like some morphological characters were more similar to birds (feathers) and some were more similar to crocs (teeth), we’d expect some behaviors to be more similar to birds, and some to be more similar to crocs. Which category methodical eating falls in, we don’t know, and as above “even talking about the problem this way is probably much too simplistic.”
“Where are these problems published or posted online (not on JP fanboards I assume )?”
Anyone who ACTUALLY looks into R&B 2007’s claims can see the paper’s many problems for him-/herself. However, for the sake of argument, I’ll give an example of 1 of the paper’s worst offenses: It claims that, among birds, “truly coop-erative, coordinated hunting behavior is seen in only a few species of diurnal raptors”; Besides the issue of semantics (Last I checked, “many” > “a few”: http://www.jstor.org/discover/10.2307/1312102?uid=3739696&uid=2&uid=4&uid=3739256&sid=55909845433 ), it ignores the existence of several papers showing otherwise (Off the top of my head, Hendricks & Schlang 1998, Frye & Gerhardt 2001, Bowman 2003, Hannah 2005). In Hone’s words, “someone either wasn’t citing them deliberately (very poor practice) or hadn’t read them (poor scholarship) neither of which is a good thing to be.”
“Why would endothermy or activity have any relevence?”
Do you really not see the relevance of physiology to ecology & behavior?
“In any case, according to Walsh and Milner, “at present there seems to be no consistent relationship between flying ability and either brain size or morphology.” Thus there’s no known way to say from its brain that Bambiraptor “had the intelligence for arboreal gliding”, contra your claim.”
I never said anything about flight, which is clearly different from arboreal gliding (I know Milner said “tree-climbing” & “an arboreal lifestyle”, but I figured I should specify since we were talking about dromaeosaurids in particular). Besides that, the relevance of the cerebrum to said lifestyle is clearly discussed on pg. 298.
“They also never imply dromaeosaurids had the intelligence to cooperatively hunt.”
Actually, they do (E.g. “One explanation for the expansion of the telencephalon might relate to drivers toward enhancement of cognition”; “Although undoubtedly enhanced in those modern groups (especially cor- vids and parrots) compared with early birds and their immediate ancestors, such improvements in cognition would presumably represent a competitive advantage over less encephalized coeval species”; Pack hunting was 1 of the “drivers”/”improvements” discussed on pg. 298).
“Dromaeosaurid brains were (roughly) intermediate between crocs and birds. You’re only thinking of it one way- that they were more bird-like than crocs. But while true, you can also say that they were more croc-like than birds. So just like some morphological characters were more similar to birds (feathers) and some were more similar to crocs (teeth), we’d expect some behaviors to be more similar to birds, and some to be more similar to crocs. Which category methodical eating falls in, we don’t know, and as above “even talking about the problem this way is probably much too simplistic.””
I don’t see it: AFAICT, the only things that make dromaeosaurid brains more croc-like than those of living modern birds are larger olfactory bulbs & the like, in which case you’d have to say the same about extinct modern birds (in reference to Zelenitsky et al. 2011). Even your tooth example is no good, given that dromaeosaurids apparently used their teeth more like raptor beaks (for ripping small pieces of subdued prey) than like croc teeth (for holding onto struggling prey), & that’s besides the evidence for “methodical eating” discussed by Hone & Rogers.
The truth is that we shouldn’t even be bothering with inferring any kind of intelligence for dinosaurs. As it stands right now we have but crappy data for intelligence in extant animals (most of which focus on mammals and birds). The association of brain size with intelligence has long been questioned, especially since mammals seem to require much more brain space to do the same tasks as lizards, sharks, birds and other animals (likely related to centralization of many tasks), thus making mammalian learning a poor standard for non-mammals. A fine example of this can be read in Leal and Powell’s 2011 paper on learning in anoles. Despite having a brain smaller than a penny, their test animals were able to learn at rates as high, or higher than certain endotherms (including doves, sparrows and crows).
Given this kind of poor resolution for extant animals I would not even venture to guess what extinct animals were incapable of doing. Was the _Velociraptor_ in this case actively discriminating between the bone and other tissue? We don’t know. As it stands we currently have no way of knowing, so it’s best just not to even consider it unless it is going to have some kind of major outcome on the final interpretation (which it doesn’t seem to).
As a natural skeptic of all things velociraptor, I have to say: I find your reasoning for how the bone got in the stomach quite reasonable.
Purely in the interests of self-promotion, I give you PVP: Predator vs predator. Predators are often predated upon.
As was going to mention that Darren but actually had trouble finding the original article. Though again the issue for me is not that predators don’t kill and eat others (they do) but the frequency. Leopards and hyena and hunting dogs make up part of a lions diet but it’s the exception, not the rule.
Hi Dave. As discussed in the blog article I linked to, intraguild predation is ubiquitous and common across quite a few studied species, and (as someone who knows a little bit about theropods and pterosaurs) I don’t see why, theoretically, a dromaeosaur couldn’t be in the habit of routinely killing smallish (wingspan 2-3 m) azhdarchids. But what the hell – I haven’t even read your paper yet.
Well it’s a question of degree and likelyhood of course. Like I say, i don’t have a problem with it occurring (i.e. lots of species engage in it) but I remain unconvinced that this would be ‘routine’ for many species. Just being careful with semantics to avoid confusion there’s a difference between something happening ‘regularly’ and ‘often’ (regular could be once a year, often could be once a day, or both) and for me ‘routine’ implies this was totally normal and happened all the time without a hitch.
Some perhaps would make as frequent a habit as they might of killing another predator for food (though of course this itself would likely be a rare event as few ecosystems are loaded down with large populations of predators) but again I’d say that, on average, there are now (and were then) few species that did this very often i.e. actively engaged in trying to take out large other predators specifically as a food source.
While I’d certainly want to go back and the stuff you cite in that blog post, I’m basically just saying that I think it’s a bit of a leap to assume that this was what was going one here. And yes of course, obviously there are multiple issues tied up in the inference I’m making (like trace marks, what we infer as the habits of dromaeosaurs from other fossils etc.) and not jsut the one of predators vs predators.
Given how the threads above have intertwined while I’ve been away I’ll continue down here:
2. Well I’m not trying to organise a ‘who has the most citaitons’ battle. Thats pointless. However, while no i don’t have them to hand, the citaitons in H&R were representative, not the only things i could find. there are mosr stuides out there on prey sizes and reivews which talk about juveniles / small species being preferred. While yes the fighting dinosaurs are out there, we do also have two Micrirpator specimens with gut contents of small animals suggesting some dromaeosaurs did go after small prey. So there is data for both sides on that front at least.
3. I don’t think it is a red herring to look at regualr patterns of bone consumption at all. i think it’s a good source of data.
4. Well again i’d simply say see point 3. I do think such consumption is unusual in general for living taxa and I do think (following H&R) it’s not normal for theropods.
Ultimately though i have to confess I’m really not sure *quite* why this is generating so much discontent. I think everyone is happy with dromaeosaurs being predatory (as seen in Microraptor for example) and would scavenge (as seen in the Hone et al. on protoceratops). While they inferred scavenging, we do at least have a reference for velociraptorines trying to eat azhdarchids (Currie & Jacobsen) and indisputably this is a Velociraptor that ate an azhdarchid.
So we already know they were predators and scavengers and that they ate pterosaurs. As such, lets assume I’m totally wrong and all my evidence is totally busted (which obviously I disagree with and moreover I think at worst people would argue that it’s ambiguous as to whether or not this is scavenging rather than it definately wasn’t), does it actually really change in the grand scheme of things how we view the carnivorous habits of these animals? Predator or scavenger it’s eaten a pterosaur and it’s swallowed a fairly large bone. That may not be much news, but it is news and, I think incontrovertible. It merely makes one example rather controversial, and there are enough of those out there which are contested or have never really been very well justified.
And my own reply comes late after a busy week…
2. It’s not intended as a citation battle, it’s that prey size is such a large part of your argument, yet you just expect us to believe you when you say you’ve surveyed the literature and it says predators rarely kill animals over half their mass. Argument from authority doesn’t cut it (and I repeat again that prey age is useless here since we don’t know how old the azdarchid was; ditto for prey individual size within its species). Who had the most citations would actually be incredibly usetful, since then you could say “here’s data from 100 animals similar in size to Velociraptor and only 5 kill prey over half their size more than 5% of the time”. Then I’d believe you.
3 and 4. Then you’ll have to establish that eating bone in theropods most likely happens due to desperate scavenging as opposed to other reasons like mineral gain or accidental uptake.
As for the amount of discontent, it’s partially because this study is hot on the heels of O’Connor et al.’s Microraptor feeding paper. That study also presented ideas without backing them up with data (arboreal birds aren’t usually predated on the ground; partially articulated animals aren’t eaten that way if scavenged; etc.). And thus both papers trumpeted results which weren’t supported by the evidence. As a scientist, that worries me.
2. Prey size is indeed a part of the argument. Though I’m not sure I’d call it a large part. Weighting up which of the lines of evidence I advocate is the more important is naturally hard to assess. “yet you just expect us to believe you when you say you’ve surveyed the literature and it says predators rarely kill animals over half their mass.” I’m not trying to fool people, nor is this argument form authority. I have actually read a chunk of the literature and the fact that I simply haven’t regurgitated a list of references on demand does not mean I’m lying / making this up / exaggerating.
““here’s data from 100 animals similar in size to Velociraptor and only 5 kill prey over half their size more than 5% of the time” Well that’s obviously not going to happen simply because even paper analyses things differently and it’s impossible to get something we might both agree on (what about 6% of the time for 3 of 50 examples? or 10% for only 2 of 100 etc.) and obviously it’s made more complex by the size of the azhdarchid.
Yes 2 m is an absolute minimum size, but it could potentially be a 10 m wingspan individual if the bone was a massive rib fragment (I doubt it but was suggested by one of the referees). It it was even just 4 or 5 m it would probably significantly exceed the dromaeosaur in size.
Here for example is a paper analysing multiple species of carnivores (http://onlinelibrary.wiley.com/doi/10.1111/j.0021-8790.2004.00817.x/full) where all but male lion and wild dog take most prey at less than their own mass, and the majority of that is less than half their mass. It’s further complicated in that hunting dogs being in packs can kinda cheat though they note that for impala art least most prey was of juveniles (so again we have have two intertwined issues here, age and size). I’d say that most of this data supports my *general* hypothesis that predators favour smaller and / or young prey. A 2-3 m azhdarchid at this size might be equivocal as to whether or not it would fall into the ‘likely prey’ spectrum relative to these examples, but any bigger and it certainly wouldn’t. Also worth noting in their introduction is that large predators tend to take larger prey and I don’t think anyone would count Velociraptor as a large predator even now, and certainly not in the Mesozoic. A different line of evidence but another one that would support the idea that it might favour smaller prey and avoid big stuff.
3&4. Well again, if you look at the gut content / coprolites for theropods you do see that either there’s lots of oral processing or that it’s bone you’d expect muscle to be attached to. This is the only case I can think of with neither (and again, bone consumption in general seems to be unusual). Hence a rare example. The only extant species I can think of that eats bone like this with any regularity is a couple of vultures. So I’d maintain that it’s a rare event. And mineral uptake could well be scavenging in any case if it’s eating part of an animal it didn’t kill.
“presented ideas without backing them up with data …And thus both papers trumpeted results which weren’t supported by the evidence” Well again, I disagree and please be more circumspect about your language. You can disagree about the strength of the evidence and it’s interpretations, but given that I have actually cited various sources for my hypothesis you are wrong to imply it’s so black ad white that it’s just not supported. And as noted above in my last response, yes this one case is open to debate (clearly) but it doesn’t change the picture that dromaeosaurs ate pterosaurs and were both predators and scavengers on occasion.
Sorry to join the discussion this late (been busy the past few days). The last entries seem to revolve about the question of likelyhood of predation on predators, but I haven’t seen that many arguments fore or against, except Darren’s blog post on intraguild predation (fore) and David’s gut feeling (against). However, intraguild predation seems (as far as I’ve read) to revolve around reducing competition, not as much as a feeding strategy. Did azhdarchids and velociraptors prey on the same type of animals? If so, does intraguild predation occur by juveniles (since the velociraptor in this case was a juvenile), or exclusively by adults?
Another thought, if predation on predators is not that common, I can imagine the main reason is that predators are more dangerous prey, given they have attributes to wound and kill other animals, so the risks of injuries are larger (that, and that predators are typically less common and therefore less available than non-predators). However, though azhdarchids are predators, they are still pterosaurs, and thus relatively fragile. I would think a velociraptor, perhaps ambushing, could with a single bite through the neck without the chance of getting injured itself. Comparing this to a coyote catching an eagle would imho not be a fair one, as an eagle seems a lot better equiped (hooked beak, huge claws) than a pterosaur and a velociraptor a lot better equiped (huge claws, powerful bite with nasty teeth, tough skin) than a coyote. If I had to make a comparison, it would be more along the lines of eagle (the velociraptor) vs. a marabou stork (the azhdarchid).
I know that no one but Dave will read this, but I wanted to share what I’ve learned since debating Micky anyway.
Short Version: I overestimated non-avian dino intelligence (although Micky still underestimated it) & underestimated fowl intelligence.
Long Version: Based on what I’ve since read (E.g. See the 1st 3 quotes), 1) Archaeopteryx were probably as intelligent as domestic chickens (& thus, fell “about mid- range on the intelligence scale of birds”), & 2) non-avian maniraptorans were probably either as intelligent as or more intelligent than Archaeopteryx (I’m thinking the latter, given the Bakker quote’s emphasis on the difference btwn wild & domestic fowl, intelligence-wise). I was originally thrown off by the facts that 1) non-avian maniraptorans probably had the intelligence for pack-hunting (See the Buchholtz quote), & 2) the only pack-hunting birds I knew of were diurnal raptors, corvids, & shrikes. However, I’ve since learned that ground hornbills (which, as indicated by the Firestone et al. quote, are turkey-like birds, behavior-wise) are also pack-hunting birds (See the Tudge quote).
Quoting Burish et al. (See “Abstract”: http://synapse.princeton.edu/burish_wang04_bbe.pdf ): “Vertebrate brains vary tremendously in size, but differ- ences in form are more subtle. To bring out functional contrasts that are independent of absolute size, we have normalized brain component sizes to whole brain vol- ume. The set of such volume fractions is the cerebrotype of a species. Using this approach in mammals we pre- viously identified specific associations between cerebro- type and behavioral specializations. Among primates, cerebrotypes are linked principally to enlargement of the cerebral cortex and are associated with increases in the complexity of social structure. Here we extend this analy- sis to include a second major vertebrate group, the birds. In birds the telencephalic volume fraction is strongly cor- related with social complexity. This correlation accounts for almost half of the observed variation in telencephalic size, more than any other behavioral specialization ex- amined, including the ability to learn song. A prominent exception to this pattern is owls, which are not social but still have very large forebrains. Interpolating the overall correlation for Archaeopteryx, an ancient bird, suggests that its social complexity was likely to have been on a par with modern domesticated chickens. Telencephalic vol- ume fraction outperforms residuals-based measures of brain size at separating birds by social structure. Telen- cephalic volume fraction may be an anatomical sub- strate for social complexity, and perhaps cognitive abili- ty, that can be generalized across a range of vertebrate brains, including dinosaurs.”
Quoting Riggs et al. ( http://www.amazon.com/Keeping-Chickens-For-Dummies-ebook/dp/B00852Z0BO/ref=sr_1_2?s=books&ie=UTF8&qid=1354688391&sr=1-2 ): “Some birds mimic the sounds of other birds and animals; few other animals mimic sounds. You can’t, however, teach chickens to talk as you can with some bird species, and they don’t mimic other animals, and so chickens probably fall about mid- range on the intelligence scale of birds.”
Quoting Walsh (See “Reptiles including nonavian dinosaurs and birds”: http://www.academia.edu/1061233/Directions_in_Palaeoneurology ): “It is now clear that while some theropod clades pos-sessed typically ‘reptile’-like brains (Giffin et al. 1988; Gif-fin 1989; Rogers 1999; Brochu 2000; Franzosa and Rowe 2005; Sanders and Smith 2005; Witmer and Ridgely 2009), at least some maniraptoran theropods had surpris-ingly bird-like brains (Kundra´t 2007; Balanoff et al. 2009; Norell et al. 2009). In these taxa, not only is the telen-cephalon significantly enlarged (sometimes to an extent greater than in Archaeopteryx), but flight-related regions such as the cerebellar flocculus, which are especially enlarged in birds, are also very well developed (Kundra´t 2007). This has fuelled the debate about whether these taxa, which often bear feathers, are in fact secondarily flightless birds rather than bird-like theropods (Witmer 2009; Kavanau 2010).”
Quoting Bakker ( http://blog.hmns.org/2012/06/bakker-blogs-the-kleptomania-continues-with-a-sid-vicious-julieraptor-dino-rustlers-part-ii/ ): “These mini-raptors were big-brained by dinosaur standards — as smart as a wild turkey (not the dumbed-down domestic version). Their eyes were huge — an adaptation for chasing nimble prey, like furry mammals and tree-climbing lizards. The extra-long arms and fingers gave the raptors three-dimensional abilities — they could scramble up trees quadrupedally, grabbing branches with claws on front and back paws. Long feathers on the arms and legs let the raptors glide from branch to branch like dino-flying squirrels.”
Quoting Buchholtz ( http://www.amazon.com/Complete-Dinosaur-Second-Life-Past/dp/0253357012/ref=sr_1_1?s=books&ie=UTF8&qid=1354904213&sr=1-1 ): “Endocasts of small coelurosaurs (Russell 1969, 1972; Colbert and Russell 1969; Kundrát 2007) display a strikingly different anatomical pattern (Fig. 10.6). They retain details of brain anatomy and roofing bone sutures on their surfaces, suggesting that the brain filled the braincase nearly completely. Brain flexures are minimal and olfactory bulbs are small, indicating that smell was not a dominant sense. Cerebral hemispheres are separable, convex, and expanded laterally and/or posteriorly (Kundrát 2007), suggesting an active intelligence. The large optic lobes are visible either dorsally or displaced laterally by the large cerebrum, as in living birds. Russell (1969) associated the large optic lobes with large eyes and binocular vision, and it is likely that sight was the dominant sense. Kundrát (2007) described an expanded cerebellum with presumptive cerebellar folia among the avianlike characters of the oviraptorid theropod Conchoraptor, inferring excellent balance and coordination.
Encephalization quotients of small coelurosaurs vary with predictions of body mass and percentage of braincase fill, but even conservatively, they are far higher than those of any other dinosaur group, overlapping those of living birds (Hopson 1977; Kundrát 2007). Larsson et al. (2000) estimated cerebral volumes by superimposing ellipsoids on endocasts with surficial indications of cerebral extent. Their data suggest at least three stages of increase of relative cerebral size to total brain size over a period of only 40 million years: of coelurosaurs over allosaurs, of Archaeopteryx over coelurosaurs, and of ornithurine birds over Archaeopteryx. The high encephalization values of small coelurosaurs indicate an active, complex, and social lifestyle that agrees well with their frequent interpretation as pack hunters.”
Quoting Firestone et al. ( http://www.amazon.com/Watching-Wildlife-Southern-Matthew-Firestone/dp/1741042100/ref=sr_1_8?s=books&ie=UTF8&qid=1354945753&sr=1-8 ): “The largest of Africa’s hornbills, the ground hornbill is almost entirely terrestrial, and is also the only Southern African bird that walks on the tips of its toes. Like a large turkey (and the ground hornbill is sometimes mistakenly called the turkey buzzard), small groups patrol grasslands and open woodland looking for prey.”
Quoting Tudge ( http://www.amazon.com/Bird-Natural-History-Birds-Where/dp/B007K4GP1W/ref=la_B000APQE3M_1_2?ie=UTF8&qid=1354947277&sr=1-2 ): “The sociality that is encouraged by the diet tends to spill over into all aspects of life. So it is that hornbills are fruit eaters and also, as we will see in Chapter 7, are outstandingly social breeders, with various kinds of social arrangements. But also among hornbills we see an interesting twistiwhere the innate sociality has in turn become adapted to a quite different kind of feeding. For among the biggest of all hornbills, and in various ways distinct from the rest, are the two species of ground-hornbills from Africa. Ground-hornbills are not mere fruit eaters: they are formidable predators. The beak is like an icepick. They can hack their way into a tortoise. The Northern species is among the biggest of all avian predators. The ancestors of ground-hornbills were presumably fruit eaters, and that, perhaps, is how they first evolved their sociality. Now, as predators, they hunt in packs. Typically they chase some hapless creature like a hare into a bush and then some act as beaters while others lie in wait and deliver the coup de grace. The packs are usually family groups. They can be seen as strategic predators like wolves or perhaps as problem families, terrorizing the neighborhood.”