One thing worth bringing up on the wrists front never quite made it into the paper (or even the supplementary material) in quite the way I would have liked. It gets a mention but not the emphasis required at least in terms of palaeoart but also in terms of the actual biology of the animal in hand (since there is only limited space and you can’t really go off on a tangent for a few hundred words even if you want to).
Microraptor, as we know, has very long arm feathers, far longer than anything else we looked at in fact. Those keeping up will also note that it has really rather strongly rooted feathers and thus, what you see in the holotype is what you get. Take another look at that and you can see that the big flight feathers on the hand are stick out at a fair angle from the hand, but still act as a kind of angled extension to the arm. They were also probably rooted firmly in place and could not move much, even if the animal wanted to move them. But what does this mean?
Well, if you take a look at most palaeoart of Microraptor and even academic reconstructions and drawings of the animal in published papers, people seem to have been cheating. Not in a dishonest sense I hasten to add, more that, whether they meant to or not, the image just does not match the anatomy of the specimen. I’ve even checked a few of these and the feathers are either not at the right angle to the hand (which would be fixed), or are much shorter than as are preserved. The feathers are so long, and indeed the arms are so long that, [important note looming] assuming the arms are held such that the palms face inwards (the ‘clapper’ position) the feathers will be stuck in the ground. They can’t not be, they are just too long and can’t be moved out of the way. Microraptor has a serious feather issue.
Even with a flexed wrist, Microraptor's feathers would be stuck in the ground. Even if the arm were held out straight and parallel to the ground, the problem would remain. Image from Sullivan et al., 2010.
There are of course solutions to this problem and we illustrate and describe them. First of all, if you pronate the hands a little (i.e. rotate the hands inwards towards a ‘slapper’ position) then of course the feathers will now point at least partially to the side rather than down, and while this makes them stick out awkwardly, they won’t drag on the ground. Secondly, if you raise the arm right back so that the humerus is parallel to the spine, and then crook the elbow and / or the wrist a lot, you can get the feathers to point backwards or even up. The arm is in an odd looking position and might even get in the way of the legs (though of course birds, with an ever more abducted wrist, pull this off quite neatly) but again the feathers are at least clear of the ground.
Bringing the humerus back to be subparallel to the spine and then bending the elbow and the wrist finally brings the feathers free of the ground. Image from Sullivan et al., 2010.
Now Microraptor has some really big feather and clearly led a rather unusual life. Most, if not all of these things would be unnecessary for pretty much every other feathered dinosaur, but the general point is important. Feathers (whatever their purpose) were very important for the animals that had them, and as we can see from the fossils, they evolved to have lots of them and big one, and were kept in good condition. You don’t see dromaeosaurs or oviraptorosaurs with frayed feather tips and ragged edges, they looked after them and this includes not scragging them on vegetation or abrading them on the ground. Whatever else Microraptor was doing, it was looking after its feathers. This means it did not hold its arms in a way that would have led to profound damage to the feathers (or may just have even been physically impossible if those feather shafts were especially inflexible), and thus one of these, admittedly unusual, postures must have been adopted. You can’t manipulate the anatomy and you can’t contradict the evidence, so anyone thinking of drawing Microraptor in an old-style ‘Velociraptor’ pose with the hands held in front of the animal below the head needs to think again.
Biologically this is also interesting and potentially important. After all if the hands of Microraptor (or any similarly feathered dinosaur) tried to bring its hands to its mouth while holding something or to strike at prey in front of it, then it won’t be able to without dragging its feathers across the floor. This puts rather a limit on what the arms can actually do, especially when brought forwards. This isn’t necessarily and issue for the animal, such a penalty would clearly be a trade off against having those enormous flight feathers. But it does mean that if we are to successfully consider and test ideas about how the hands and arms of maniraptorans evolved and were used, we have to take such issues into account and not consider the bones in isolation – those feathers were important too.
(A few extra caveats / details. Yes, feather shafts are at least a bit flexible, and could probably be crushed against the ground without breaking them or doing too much damage, and yes lots of preening would help keep them in good shape. However, doing this every time the animal wanted to bring the arms forwards would in the long term risk damaging them, especially if done often, and preening also brings its own costs in terms of time and effort. On balance, it is far easier to just keep the feathers off the ground the bend shafts, ‘unzip’ barbs, and break tips that it is to go through that rigmarole and risk constantly. It’s also possible that the feathers were less firmly fixed / more motile than in modern birds, but they do have to take the forces of flight [presumably] and we do see things like quill knobs that argues for a very secure and fixed attachment).
I guess I should add, in the best traditions of SV-POW! and Tet Zoo, massive ‘Musings bonus points’ will be available to anyone who can find a ‘correct’ Microraptor online or produce one themselves.
Dave Hone's Archosaur Musings 
This is really fascinating stuff! I’ve never drawn a grounded Microraptor before (although I have done a pipecleaner one: http://albertonykus.deviantart.com/art/Microraptor-123006115 ), but I’ll keep this in mind. (Not that my flying ones are any better – their wings are lifted too high.)
Here’s another hypothesis: It didn’t usually stand on the ground but in trees?
Well, yes. I was kinda hedging around that, but without really saying it, but that is one obvious extension of this. Still though, they would have ended up on the ground sooner or later and would have to have done something with those arms.
This thought occurred to me as well, then again ‘sticky out’ feathers would likely get in the way of climbers as well.
Two solutions:
(1) In terrestrial mode, it ran around only in places where there were two parasagittal grooves on the ground in which to put and protect the feathers.
(2) It always ran around with its arms abducted, in an evolutionarily-advantageous, “airing out the pits” posture (sure to make it more attractive to the opposite sex).
I look forward to seeing those in print Jerry.
I’d draw you one Dave, but I’m a bit busy on er….something else.
Yes, and for that I am most grateful!
On a serious note, wasn’t Greg Paul drawing raptors with hands crooked right back that way since year dot?
I know there was a bit of debate a while back that resulted in therapods all getting depicted with the happy clappy pose for arms, did this get overstated in therapod images?
Looking at the old ‘Pred dinos of the world’ which is all I have to hand, some are flexed far too much and some far too few (compared to our results) and many non-maniraptorans seem to be slappers and not clappers. So definite credit for getting them flexed, but lots are out in terms of degree or orientation, though of course a great deal of this was done before much had ever been said about wrists, so this is not a criticism, merely hindsight.
Very interesting. I wonder if this has anything to do with the apparent lack of tertial flight feathers in many non-ornithurine maniraptorans?
Sorry, only just spotted this comment, obviously there’s been quite a run here and I’ve been out of the office most of the day. Short, answer though (sorry) not sure, I’ll have to think about it and go see some specimens. Feel free to speculate.
Hello! I’m a lurker and artist, and enjoy your posts thoroughly. I haven’t had the opportunity to read your wrists paper (yet), but these two posts have spurned me to ask a few questions:
1. You write that one option may be to rotate the hands into a “slapper” position (pronation) to keep the feathers off the ground by lifting them horizontally. I am under the impression that pronation was not really an option for non-avian theropods, just as with birds. Therefore, do you mean that the arms themselves would rotate and thus the feathers would stick out horizontally? If not, what do you mean? Could theropods actually pronate their wrists?
2. Could Microraptor have folded (abducted) its wrists more fully than you show in the illustrations above?
3. Is there any possibility that Microraptor flapped to power its flight? Could it launch from the ground, or was it a glider/parachuter only (as I had previously understood it to be)?
Thanks for all your help! I’ve been planning out an *M. gui* illustration and these posts have already been influential. I’ll make some noise when it’s on its way to being done…
Hey there, good questions, but I’m busy so apologies for being brief.
1. They don’t have to pronate the wrists, they can pronate the arm. This was, as I have said before, really quite limited, but even a little bit of pronation would of course have quite an effect on very long feathers.
2. None, that’s based on our reconstructions. Admittedly that’s actually largely based on Deinonychus and Phil Senter’s paper, so there could be more in there, but as you can see it would have to be quite a lot greater to get the feathers up.
3. Personally I think it’s a strong possibility, if only though the huge sternum which was prtobably housing some decent musculature to pull the arms down. Could it fly like birds, or even Archaeopteryx? No, but it might have been able to do a couple of half hearted flaps to gain a little bit of thrust / control. That though is tending rapidly towards vague (and none too well informed) specualtion so I’ll levae it there before i dig myself into a huge hole.
Hello! I’m a lurker and artist, and enjoy your posts thoroughly. I haven’t had the opportunity to read your wrists paper (yet), but these two posts have spurned me to ask a few questions:
1. You write that one option may be to rotate the hands into a “slapper” position (pronation) to keep the feathers off the ground by lifting them horizontally/laterally. I am under the impression that pronation was not really an option for non-avian theropods, as with birds. Therefore, do you mean that the shoulders would rotate and thus the feathers would stick out horizontally? If not, what do you mean? Could theropods actually pronate their wrists?
2. Could Microraptor have folded (abducted) its wrists more fully than you show in the illustrations above?
3. Is there any possibility that Microraptor flapped to power its flight? Could it launch from the ground, or was it a glider/parachuter only (as I had previously understood it to be)?
Thanks for all your help! I’ve been planning out an *M. gui* illustration and these posts have already been influential. I’ll make some noise when it’s on its way to being done…
Sorry, I have to approve all comments from new commenters, hence this not appearing before now.
I suppose the portion of the argument I would take issue with is- “They were also probably rooted firmly in place and could not move much, even if the animal wanted to move them.” Smooth muscle bundles attach to the sides of the follicles in living birds, often in extensive networks that can move groups of feathers at once. In fact, there is an interremigial ligement which bounds the primaries and secondaries so that they can be folded in unison. It’s true that the outermost primaries in modern birds are not very mobile (unlike the inner primaries and secondaries), but remember that in modern birds these attach to the manual phalanges. Deinonychosaurs like Microraptor had much better developed phalangeal joints, so could have flexed their fingers to further modify their primary feather angle. Another factor is that since Microraptor lacked quill knobs, its remiges might have been less strongly attached than those of recent birds, giving its feathers more mobility regardless of the bones’ position.
“It’s true that the outermost primaries in modern birds are not very mobile (unlike the inner primaries and secondaries),”
Even more than that, they are virtually immobile on the ones I have looked at, which is what I’m suggest (suggesting, I freely admit I could be wrong). However, while it’s possible that these were more felxible / lacked a big ligament to help them there, there are two reasons I think this is not the case.
First off, if Microraptor is gliding (and I think pretty much everyone is happy that it is doing this, with the arms at least) then those primaries would be much less effective if they could move a lot since drag / lift forces would likely bend / twist them to the point that they are not producing any good aeronautical effect / control rendering them little more that a breaking device which is not what a glider wants. Secondly, in all the specimens I have seen, unlike body contour feathers, the position of the feathers is consistent (this is also true of things like Caudipteryx as well actually and a bunch of Liaoning birds) which implies that the primary / wing feathers were more rigid / strongly rooted / less motile than body feathers.
I don’t think the finger argument works since if they curled the fingers (as if making a fist) then the featherswould still be forced inwards towards the body or the ground, so that won’t help them get out of the way, and some of these primaries from the wrist would still reach the ground in any case, so the problem remains.
Equally, quill knobs don’t turn up on plenty of extant birds, so that’s not necessarily a reason to think they are essential for strongly rooted feathers, and in ant case, the bones of Microratpor are not that great (in all the specimens I’ve seen) so they could be present but not preserved.
So while i do take your point, I think my conclusions are pretty safe at the moment, though some decent modelling and more work on modern feathers and some empirical measuring of feather angles wrt bones in the fossils would all help shore this up. I’m working on it, but for once wanted to write about it well in advance of me having even started a paper on it.
Hmm. Wang’s (2007) thesis suggests that in the pileated woodpecker, the outer primaries can be folded to about 50 degrees compared to the carpometacarpus. Of course, that’s based on my estimation of the carpometacarpus’ position in the wing, and the author’s manipulation of a dead specimen. I certainly agree Microraptor’s primaries were less mobile than most of its other feathers, though how much less mobile they were and just how much immobility is needed for aerial locomotion is another matter. As for the finger flexion, I was thinking the metacarpal-based primaries would be more flexible (so not still reach the ground). You’re right it wouldn’t solve the problem without pronation, but with a pronated arm it would at least stop them from projecting laterally. I agree that more empirical data is needed on modern birds first though.
In the meantime, I wonder if Microraptor’s extremely long hindlimbs are partially there to keep the remiges from dragging, and if its distally restricted tail fan (compared to Archaeopteryx, Jinfengopteryx, etc.) was partially to give the primaries space if they were normally held in your posterior position.
Well that’s also a single taxon too, were any others mentioned in this paper? I do want to get my hands on some dead birds, but I have my share of experience with pigeons and chickens and various other domestics and the odd bird of prey and I recall them being pretty fixed.
As for the long legs, while not shown in the reconstruction above we did play with the legs and even kind of standing it on tip-toe you still can’t get the feathers clear of the ground without bringing the humerus up or pronation. All of this does indeed need to be explored more, but based on the (currently) available evidence and certainly the massive finger feathers, I think that it’s fair to say the ‘default’ for now should be that the arm needs and must go back, and not come forward unless there’s some pronation.
By the way, is it possible the derived troodonts shortened their forelimbs in order to avoid this problem?
It’s possible I suppose, though I doubt it. Not that evolution takes the easiest route, but it would be easier and perhaps better to just put the feathers on at more of an angle or to evolve some motility for them than to shorten the arms. Certainly flying animals would not want to reduce the arms, so the feather angle / position would be less critical for them.
I may be mistaken somehow, but aren’t derived troodonts thought to be cursorial and non volant, and their forelimbs short compared to most dromaeosaurids and flying birds?
Non-volant yes, or at least probably (I can at least concieve of something like Anchiornis parachuting, and that would raise the possibilty in other taxa). Should have been more clear there.
Still, as I say the issue here (which of course incredibly hard to secodn guess evolution over 100 mya) is one of what is most likely to happen and I think there are more plausible scnarios to reduce this problem than by shortening the arms.
My speculations: When Microraptors hopped on the ground, climbed branches and flew they used the same ‘galloping’ muscles. The hyperextended toe claws held them upright while perched with the tail and associated fan of barbed firm feathers anchoring vertically to the bark like hooked velcro or snake belly scales. (I’m assuming the tail vertebrae had very limited lateral flexion (no tail wagging) but good dorsiflexion.) The wings had similar barbed feathers which allowed good climbing up but not down (thus “forcing” flight/gliding. They probably slept perched upright, heads turned back like geese. So regarding wings and feathers, IMO flight was secondary to climbing, display tertiary.
“The hyperextended toe claws held them upright while perched with the tail and associated fan of barbed firm feathers anchoring vertically to the bark like hooked velcro or snake belly scales.”
Which feathers? The vast majortiy, even on Microratpro are still realtively simple proto-feather like filamnets and not long, stiff, vaned feathers.
“So regarding wings and feathers, IMO flight was secondary to climbing, display tertiary.“
But that implies things like Dilong, Compsognathus, and Beipiaosaurs evolved feathers to help climb, which is unrealistic. Even if you want to try and limit that to a pennibrachium, what about Caudipteryx? Or for that matter the (very probably) arboreal scansoripterigids which have not rigid feathers at all (aside form the unusual plumes on Epidexipteryx).
I think you are trying to fit everything for the evolution of feathers to Microraptor alone and it doesn’t match the pattern of evolution we see or correspond to the taxa which have feathers and their probable lifestyles.
“Well that’s also a single taxon too, were any others mentioned in this paper? I do want to get my hands on some dead birds, but I have my share of experience with pigeons and chickens and various other domestics and the odd bird of prey and I recall them being pretty fixed.”
No, the thesis was only on woodpeckers. Interspecific variation is certainly an important factor though. I wonder, do Shenzhouraptor or Sapeornis have the same problem? They both have quite long wings as well.
Quite possibly though I’ve not looked at them, on indeed anything else like that yet. I really should get stuck into Archaeopteryx if nothing else….
IMO theropod etc. dinos had feathers, derived from ambushing arboreal ancestor, which due to large size (pigeon-turkey) could not sleep inside crevices/holes in trees (cf arboreal salamanders/lizards), so slept on branches (perch claws, crepuscular eyes, omni-carnivore diet, longer feathers & hooked feather tail), and selected for thermoregulatory camouflaged silent downy feathers, with stiff flight feathers gradually evolving for longer directed leap-flight-ambush. They killed lizards not by ground chasing but by leaping down from above, hawk-like or puma-like. Possibly hard calcified eggs selected against bone/teeth and for gizzard stones & increased lightweight keratin beak/claws/feathers, reducing paired organs while enlarging air sacs.
(insert before my last previous sentence)
Once a toe evolutionarily moved side-wards and reversed (great perching-flying advantage), the teeth, hyperextended toe and long bony tail were lost in direct arboreal descendants (birds), while dinos became more terrestrial (bigger, teethy, wagging tail and shrank hyperext. toe). Only after dinos went extinct did some birds lose primary arboreal traits and became big nonmigratory runners. Arboreal water birds (divers, waders, paddlers) were probably the first big migratory birds.
Sorry, this is really hard to follow. As I said above though you seem to be conflating all theropods into soem kind of single hypothetical transition and you are not taking account of the variation seen in feather types and expression, body size, diet, habitat type etc. You can’t just cram all theropods or all feathered theropods into a single lifestyle / habitiat to explain flight/ arboreality. There are terrestrial and arboreal forms, those with filamentous feathers and those with vaned feathers, arboreal and terrestrial forms, predators and herbivores and everything from Anchiornis in size to Beipiaosaursus and quite possibly even T.rex. This also has to fit the trends we can observe (like lengthening arms, felxing wrist, changing feathers patterns, etc) both phylogenetic and evolutionary.
Just saying “feathered dinosaurs were X and so must have evolved by Y” doesn’t work. What are these cerpuscualr non-feathered arboreal ancestors supposed to be? You need to think this through and put out a hypothesis that is consistent with the available evidence (and there is quite a bit of it).
It folded its arms back, thus anticipating birds, and ate with head only, as a snake does.
Interesting question. Approaching the problem as an artist rather than an anatomist, my instinct would be to just increase the abduction angle of the wrist until the feathers clear the ground (so maybe 120 degrees instead of 60 degrees). I suppose you are confident enough with your results to not allow that.
I’ve noticed that people also tend to restore Mei with a much higher degree of manual abduction than your study supports. For example, Julius Csotonyi’s restoration of Mei assumes that the position of the manus in the skeleton is in a lifelike articulation- I guess you would say that it is not?
Well we are quite confident of the results, though in terms of exact wrist flexion as opposed to exact angle of the radiale etc. (there are multiple factors here) it’s a combination of our work and others. It seems unlikely they could flex 50-100% more than we think. (Take a look at the Berlin Archaeopteryx for example).
As for Mei, I’d have to go and take alook at the specimen, but in general I note quote a few people have drawn paravians with hyperflexed bird-like wrists when really while they are on the way, can’t go that far.
Is it ok to draw Microraptor standing like this?
(Doesn’t work with the wrist fully flexed, of course)
I’d really like to see someone reconstruct a walk cycle of Microraptor! How does it avoid dragging its metatarsal feathers on the ground when it lifts its feet?
That works as a posture for standing, but not for locomotion – I doubt it could run like that. And of course if those feathers are still, they are even more likely to catch on something and even trip the animal when pointing forwards. It could get into that posture I suspect, but I doubt it could do much but stand (in terms of keeping the feathers clear).
I’m a college animator and I’ve been putting quite a bit of time into this species for a piece I’m working out. I really appreciate this article! I too noticed that many drawings were completely wacky (particularly gliding). Is it possible that they might crawl on the ground, as they would on trees? I know that’s decidedly uncharacteristic of their near-relatives, but it was something I’d considered might be possible if they only grounded themselves rarely. How would it climb, anyway? Would the legs come out to the side a bit like a frog or would it stick its rear-end out? I figured it would be the former because of the long legs, but would that interfere with the wings? Some resources would help me a lot.
Well all the evidence points to them having an upright and parasaggital posture (legs under the body, arms tucked in). A large part of the gliding question has always been how far out can the legs move (as if doing the splits to the sides) to bring the leg feathers up into the plane of the arm feathers and on this people profoundly disagree (also see my recent paper on the position of the leg feathers).
Also the arms are unsuitable for crawling as the can’t pronate (rotate such that the palms point down) so if it was to crawl it would be walking on the edges of it’s fingers (which in this case are covered in feathers) so not a great form of locomotion.
Climbing position must be a bit speculative, but that usually depicted (and I think probably about right) is with the arms grasping the tree and the legs lined up behind. Watch a cat climb a big tree and they do the same thing, support with the hands and power with the legs. IT fits with what we know of how the arms and legs position in dromaeosaurs so that’s probably your best bet.
Is there any indication that might suggest it would have, or might have preferred, to launch from a specific position such as upright or upside down from the trunk? This might be too in-depth, but I figured it was worth a shot.
Well few animals can move head-down vertical surfaces as this requires a specialised ankle (as in squirrels and fossa for example) to rotate the joint and maintain a grip and Microraptor definitely does not have that, so I imagine it’d have been head-up almost the whole time it was in trees.
Sorry to dig this blog post out of the graveyards of history, but I wanted to ask a question and then make a point about the “cheats” in skeletal reconstructions.
Question: Extant birds are able to fold their primaries back some, rather than have them permanently in the same plane, and those quills viewed in side view (or via dissection) also lie right up against the bones (e.g. http://preview.tinyurl.com/feather-insertion ) so what exactly is the evidence that the “strongly-rooted” feathers of Microraptor are less mobile than those of extant birds?
Comment: Academic skeletal reconstructions of dinosaurs…especially the forelimbs of bipedal dinosaurs, almost universally draw them in a flat 2d plane hanging down from the body. Even in the taxa where this is possible (a surprising number of them) it isn’t meant to imply it’s the normal one. Generally the elbows would stick out of the plane at the viewer and the wrists would tuck somewhat back away from the viewer. But without doing a front or top view, that sort of foreshortening is counterintuitive to most viewers (not to mention time-consuming to produce…especially via pen and ink, which is how most people do them) so it’s basically visual short-hand.
I’m not arguing that this hasn’t caused mis-perceptions in paleoart…it almost certainly has and your Microraptor wing issue is an excellent example, but it’s sort of like saying that strat column illustrations are flawed because there isn’t really a cliff that looks like them.
That isn’t to say the situation shouldn’t improve, but until either 3d tools or else time and budget catch up to the point where it can be accomplished it probably won’t change rapidly.
That said, in an attempt to gain some musing points I rushed this out: http://skeletaldrawing.com/skeletal_anatomy/Microraptor_wings.jpg
Of course even though the feathers clear the ground in that diagram, they would still be in the plane that the hind limbs occupy during locomotion. The most likely solution to my mind isn’t rotation at the wrist, but rather the entire arm pivoting out at the shoulder instead.
In answer to the question Scott, the short version is ‘none’. I hope it’s clear that this is a hypothesis. However, I was unable to find all but the barest evidence that primaries could move much at all in the literature. Even if they can, this is still not necessarily a problem for my hypothesis since a) the result is so extreme for Microraptor that these feathers would have to shift a very long way before they because clear of the ground and b) one might expect early feathers to be less derived and therefore less motile than in modern birds. They might well not have developed the muscualture and articualtions necessary to provide strong anchoring and motility.
As to the comment, yes indeed I take the point. And I did at least mention the possibility of moving the arms around (if via pronation or odd elbow flexion) to get around this problem. I agree that the cruicial issue is likely to be time to try and get everything *just* right. If noting else this post was (in an oblique way) a bit of a dig at those who simply recycle things and assume they are accurate or who don’t pay attention to the details (i.e. note the feathers are long, but not how long or where and how they articualte and at what angle). I’ve seen a ton of identical looking Microraptors online and they all seem to repeat the same errors abotu feather length and articaultion.
Fair enough. I know that movement (in the same plane as the long axis of the limb) has been shown to occur for some time (e.g. http://jp.physoc.org/content/30/3-4/221.full.pdf for a description of the muscles and nerves that are involved) and if you contrast the arc distance occupied by the primaries when the wing is extended and when it is folded at rest it’s clear that some movement must occur to let the feathers to stack up.
That said, it doesn’t demonstrate whether the distal-most pinions actually fold back, or if the more proximal primaries fold towards the distal ones (in which case the problem you describe would still be in effect). And despite a rather energetic attempt to cull through the literature, I couldn’t find anything either.
Looking at living birds with their wrings folded up, as well as x-rays (e.g. http://www.midcoast.com/~nlund/X-Ray_Chickadee.jpg ) it looks to me like the most distal primaries can fold back 30-40 degrees from the long axis of digit 2; I’ll grant you that the linked image is of a dead bird and may not be representative, but it seems to jive with what can be observed in living birds with their wings folded.
I admit this is not terribly precise, and even if correct it doesn’t mean that Microraptor could do it (it is consistent with the feather positions of several fossil taxa, including some specimens of Archaeopteryx).
Anyways, fascinating post! I thank you for entertaining my johhny-come-lately timing as well.
A good discussion is always a good discussion. All I would conclude here is that I could find no good evidence anywhere that Microraptor (or indeed modern birds) could move their primaries extensively and thus (from the point of what is really notes in an appendix) it was fiar to conclude that they moved little. And hence to this post. This could be very wrong, but the evidence is currently profoundly lacking so I think it best to be conservative right now and stick to the fossils, which as noted above, seem not to be the case (in terms of size or angle) in a great many cases.
Dr. Hone, this is a fascinating topic of discussion. I just read the 2010 paper and I noticed that you did not mention examining any Microraptor skeletal material. Your conclusions for Dromaeosaur wrist anatomy seems to be drawn from Senter’s analyses of Deinonychus and Bambiraptor. Can you be certain that Microraptor, an animal that probably engaged in aerodynamic behaviors, had the same range of wrist motion as Deinonychus?
It seems to me that if Microraptor could fold its wrist more tightly than larger Dromaeosaurs then this whole discussion may be moot.
Perhaps you’ve already discussed this and I missed it in the comments and replies above.
Again, though, this is entry inspires me to research the matter further and then to contribute some information to the debate.
Thank You,
-Jason
Well Deinonychus as examined directly by Corwin Sullivan (table 1 in the paper). Offhand i can’t recall why we didn’t include Microraptor though I’d imagine it’s becuase the relevant bones were simply not visible at the required orientation to get the information needed. Obviously at the IVPP we had access to numerous specimens, but that doesn’t always mean you get what you want.
Indeed this could be moot if t had more flexion, but that assumes that this was likely. However, this would need to fold it’s wrist a huge amount more than we see for other related taxa. We’re talking about a massive increase in flexibilty and that seems unlikely. Note that even some early actual flying birds don’t have much more flexion that other dromaeosaurs, so I’d be quite amazed if Microrapor had something like double the felxibility we see when even the birds didn’t achieve this for quite some time. You can always invoke “well we don’t know but *maybe*….” but in this case it’s not much of an argument given the evidence we have.
Well, perhaps it’s a better argument if we consider evidence drawn from Microraptor itself, rather than other taxa. Sure, it is probably most parsimonious to assume that all the members of a clade share a feature if you have no other information. However, luckily, we have a lot of other information.
I am looking at the figures in the Xu et al. 2003 Nature monograph on M. gui right now. Figure 2i shows the left hand with an angle of almost exactly 90 degrees, while Figure 2f shows the right hand with angle between the 2nd metacarpal and ulna of around 115 degrees. Both are more than three times the abduction you report for Deinonychus.
Perhaps you suspect that these wrists are preserved in a disjointed position, but there are more specimens.
Fig. 23, CAGS 20-7-004, of Hwang et al. 2002 (AMNH Novitates) shows even more abduction in the wrists of M. zhaoianus. I get about 125 degrees. In this specimen the semilunate carpal is pretty clearly still articulated.
In the type of M. gui The primary feathers are preserved at about 135 degrees to the ulna. If you drew the primaries at that angle on your diagram above they would be well clear of the ground.
The first hypothesis is that that Microraptor had the same wrist anatomy as its relatives, and that it therefore had unfoldable wings and some unimaginable lifestyle. Another hypothesis is that Microraptor had forelimbs that were quite specialized relative to Deinonychus, possibly in conjunction with aerodynamic behaviors that were not shared with Deinonychus. In this hypothesis the primary feathers and wrists co-evolved to specialize for better folding and proportionately longer primaries. The latter hypothesis seems supported by the preserved positions in the actual fossils of Microraptor, while the former is supported by evidence drawn from other taxa.
Again, this topic is most stimulating. I would love to get the opinions of Mick Ellison (AMNH), who has studied the wings of Microraptor extensively.
Thank you, Dr. Hone, for hosting this discussion. It has truly inspired me to go back to work on Microraptor.
Well the holotype of M. gui certainly appears to have greater flexion you are right, but I’m not really in a position to assess if that’s genuine or not. I do have soem photos of this but the wrists are, frankly, a mess. And while obviously I wrote this blog post, it was Corwin Sullivan who did all the measuring and analysing of angles on this paper so you’re better off asking him about this (and I’ll point him in this direction). Either way though, going from the top figure, (with a more ‘natural’) arm position, even a 90 or so degree flexion wouldn’t keep the feathers clear by the looks of things. Even if they did it would not be by much and certainly would not be a well tucked up position with the feathers obviously clear of any debris on the ground etc. In short, yes, looking at the photos I have flexion might be much higher thant I state here (and in my comment above) but I’m really not convinced it detracts too much from our overall point.
As for the feather angles, that I did measure myself for this images done here. Now that was over a year ago so I can’t be entirely sure, but as I recall, I measured the longest feathers in the hand and took their angle. The angles of various feathers do vary quite a bit in the arm and naturally i was interested in the biggest distal primaries as these would be most prone to dragging on the ground. And that was the angle i used for this. I’d also add these are really quite hard to measure accurately except on the specimen itself as the bases especially are rather cryptic and it’s hard to follow a single feather all the way to the bone and get the angles right.
This discussion has brought up so many interesting issues that we could research and discuss at length
1) the state of the feather musculature and connective tissues in dromaeosaurs.
2) the effect that flexing the wing finger would have on the primaries
3) the relative strengths of range of motion analyses vs. other lines of evidence
4) the responsibility that paleoartists have to be informed about all of these issues and an almost endless array of other ones.
This one topic could almost be its own website or discussion group.
I wasn’t clear about feather angles. I mean the angle of the primaries – on the manus – relative to the long axis of the ulna. I propose this as a measurement of total degree of folding in the wing, including the angle of the primaries themselves plus the abduction of the wrist, as preserved. The result is 135 degrees.
Put more simply, in the type of M. gui the primaries of the right wing are a lot closer to being parallel with the dorsal vertebral column, rather than closer to perpendicular, as your diagram indicates. If they were, say, 25 or 30 degrees off parallel to the spinal column, wouldn’t they be clear of the ground?
I’d be interested to see you add the angles preserved in the fossils (and please don’t forget M. zhaoianus, which has a well – preserved wrist) to your diagram to see if the feathers are clear of the ground or not. With your permission I could borrow your diagram and add possible positions from the fossils I’ve mentioned.
But there’s no question that paleoartists, including me, have probably made Microraptor’s wing folding too much like crown group birds. Start with the much more ventral glenoid, add the super – long primaries, and consider the super – long manus, and you have something that shouldn’t look like a hawk, with wings tightly stowed on the back and primary tips nestled atop the tail.
Sorry for the redundant posts, I was trying to fix my typos.
Just quickly, the Dave fossil, NGMC 91, probably Sinornithosaurus, has wrists at about 90 degrees also. The left hand is more than 90.
I’ve now deleted the excess ones, hope that’s OK.
If I follow your reasoning correctly, we must assume cheetahs don’t exist because its relatives have some very different structures for running. If cheetahs were extinct and we had never seen one alive, would we have to believe it could not run like it does because tigers, lions, leopards, and so on don’t have its adaptations? All this is supposing a fossil specimen might not have all the wrist and foot elements, for example.
I’d say that’s a bad assumption.
Is there any reason to suppose animals like Microraptor could not flex their wrists enough for the feathers to clear the ground based on the skeletons themselves, rather than comparing them to birds and other feathered dinosaurs? Or am I misunderstanding you?
Well i don’t know if you’re misunderstanding me because I don’t quite follow your argument.
We know the relatives of cheetah’s DO run (obviously) they have jsut made it a rather more extreme case of cursoriality, but the flexion of the wrist is, as we understand it, critical to flight as as measured is not good enough in Microraptor etc. The difference is not the presence or absence of wrist flexion but the degree in the same way that cheetahs have better cursiorial adaptations but other cats can still run.
You are again saying that the flexion of Microraptor’s wrist as measured is not good enough in Microraptor, yet you have admitted above that you did not measure it in Microraptor. My measurements show that the flexion IS good enough if we consider the preserved position in virtually all of the actual Microraptor fossils.
have you had time to measure if the feathers are clear of the ground if you use the wrist flexion preserved in actual Microraptor fossils?
You’re conflating two different arguments here (in as much as you are now adding to a different thread). Yes, I did admit that we did not measure Microraptor directly and it’s probably higher than stated. However, as noted above, having played with this briefly on paper, I don’t think it alters my original statement that the feathers would still be so long as to be a problem with the ground.
In terms of Bryan’s comment above, Microraptor still probably has a wrist flexion less that of birds (I’ve not done the detailed measurments), so while it may be higher than originally stated here, that doesn’t mean it could fly. Nor does it affect his general argument which, as far as I can see because I’m not entirely sure, seem to be about a continuous variable, and not the issue of a near binary one (true flight or not).
My apologies, I did conflate the two different arguments.
My own investigation of wing folding in my Microraptor model, using a range of flexion angles taken from actual specimens, suggests that Microraptor could keep its primaries clear of the ground. I will make the photographs available shortly. Perhaps you could also make your new paper estimates available and we could compare.
Thank you for your thought – provoking blog.
Right, gotcha. Good to have that cleared up.
I’m not actually working on anyhting else on this right now, there’s no ‘second’ paper planned, but I’d still be intrigued to see what you have.
I think the best place to start would be by demonstrating that deinonychosaurs could fold their wings.
Have you ever seen the type specimen of Mei long in person? I’m sure you’ll recall that it is preserved with the snout tucked under the wing. It is also a crucial fossil because it is almost undistorted and preserved in three dimensions. Luckily more specimens have recently come to light as well. I have a photo that is a three-quarter view of Mei long. Can I e-mail it to you and maybe you could post it here? Both wings are highly folded and preserved in or just barely out of articulation.
Yes I have seen it and indeed photographed it. It was in my office for quite some time.
And the head is tucked under the wing, but mostly it’s tucked under the elbow. Going back over my photos at least, both hands appear to be missing, so it says nothing about wrist folding, only that is has a well folded elbow which is hardly news. In any case I’d be careful about interpreting that specimen, there are bits missing and other that have obviously disarticulated, so while the whole this is in a pretty natural pose, i would not start taking measurements of individual angles and treating them as accurate.
Hmm, no, both hands are there, but you can only see them in ventral view. They are gripping the knees and feet actually. They are labelled lm and rm in Figure 1b of the monograph. They are complete and preserved in articulation.
The specimen is not disarticulated nor dorsoventrally flattened, contra Senter, 2006. The only part of the left arm or shoulder that is missing is a middle section of the radius, but the forearm is preserved in articulation nonetheless.
The glenoids face laterally, and that’s crucial because it allows real elevation of the elbow above the glenoid. Note also that the elbows are far more flexed than in your Microraptor work.
You have cautioned me to not start taking measurements from this specimen, and I have not. In return, I would suggest that it is crucial that all hypotheses about function should be grounded in close examination of the fossils. My difference with you on the subject of this blog is that you have stated quite confidently that Microraptor cannot keep its primary feathers from sticking into the ground, and that they can’t not be. I think you have based this bold, categorical, conclusion on your carpal joint paper and the work of Senter. Yet there are contrary examples that indicate a greater range of motion, as Senter (2006) noted. Thus I would simply ask that you keep an open mind on this subject, and I will endeavor to do the same.
“The specimen is not disarticulated nor dorsoventrally flattened ”
Well it is disarticualted. Because I’ve seen it quite recently to look at a different problem and it’s quite obvious. The skull is obviously disarticualted (there is matrix between the bones and they have separated out), and there are a number of dorsal vertebrae missing, and the base of the neck looks badly disjointed and unnatural so either something went wrong or there are more verts missing there too. So clearly this is not a ‘perfect’ specimen and things have moved or been altered. So as I said, it is obviously *generally* in a near natural position, but I don’t think you can confidently take detailed measurements from it of something like a joint angle, as it must have settled / moved to a degree between death and preservation (as seen by the invasion of rock into the skull etc.).
“My difference with you on the subject of this blog is that you have stated quite confidently that Microraptor cannot keep its primary feathers from sticking into the ground, and that they can’t not be. ”
Yes. And I seem to have admitted several times now that I did do the reconstruction above on our *general* calculation of Microraptor and that this might be in error. [See my earlier comment “Well the holotype of M. gui certainly appears to have greater flexion you are right, but I’m not really in a position to assess if that’s genuine or not.”].
I also said that I have not done the detailed calculations on the material of Microraptor to determine exact angles. However, eyeballing things, I strongly suspect my *overall* comment / assertion is probably still right. [See my earlier comment “Either way though, going from the top figure, (with a more ‘natural’) arm position, even a 90 or so degree flexion wouldn’t keep the feathers clear by the looks of things.”] .
I’m quite willing to be wrong. But I’ve not done the work and you’ve yet to provide with with anything more than “the angle is probably more than you said” which I agree with. So either of could be right, you think you are and I think I am. That’s a simple unproven disagreement. I seem to just be treading water here and restating my position again and again.
I am being open minded (hence conceeding I might well be wrong and that things need to be checked.). But as I have not checked the details of the wrist of Sinornithosaurus or Microraptor (and nor has anyone else), and I’ve not redone my figures (and nor has anyone else), I really don’t see that I am being in any way unreasonable by standing by my personal assessment that even increasing the angle quite a bit wouldn’t necessarily get the feathers well clear of the ground.