Um hey, Scientific American? Bird knees bend the same way as everyone else.

Ok, time for a quick anatomy lesson: Despite what you may have heard, bird knees do not bend backward.  Nor, in fact, do the knees of any tetrapod perform this trick.  Given the role of the knee in locomotion, it's not even clear how such a reversal could evolve after the initial "knee bend" direction was settled upon several hundred million years ago.

Why bring the anatomical equivalent of a fairy tale?  Well, it's a fairly common misconception.  So common, in fact, that it was recently enshrined by none other than Scientific American.  So let's see if we can clear this up with some simple diagrams:

Really, it's not.

It's a surprisingly common mistake.  When looking at living birds many people fail to realize that part of the leg is hidden on a bird; the upper leg barely moves, and along with the knee it is actually buried up under the feathers of the wing and body.  Birds also have quite long ankles, leaving their ankle joints in roughly the position we'd expect the knees to be on a human.  Like this:

The key here is that people are plantigrade animals, while all theropods (including birds) are digitigrade.  That means that human ankles are flat on the ground, and in our case our knees are roughly in the middle of our legs.  In birds and other digitigrade animals (most dinosaurs and many mammals, like dogs, deer, and horses) it's only the toes that contact the ground.  The ankle joint is well up off the ground, and the knee is is actually in the upper 1/3rd of the leg.  And in birds the thigh is actually even a smaller part of the leg, and as mentioned above is also mostly hidden under feathers.  Here's a look comparing the same leg portions of a human (in two poses), a dog, and an extinct bird:

(Presbyornis copyright Scott Hartman, other skeletals modified after Charles Knight.)

In the diagram the thigh bones (femora) are all colored red, the shin bones and proximal ankle bones are colored blue, the "foot bones" (the distal tarsals and the metatarsals) are green, and the individual toe bones are yellow.  Notice that the femur of the extinct bird Presbyornis is small and very high, and the knee (the joint between the red and blue bones) is up where it would be hidden by the body and wings.  But please also notice...the knee is bending the same way as ours.  And the same way as everything else that has a spine and walks on land.

Which brings us back to the article posted by Scientific American.  I don't want to go too far with "gotcha" blogging, but Scientific American is generally one of the more highly regarded popularizers of science on the web and in print.  I took a look the other pieces published by the author, and she seems like a solid reporter that just happens to have made a mistake.  Journalists don't get science degrees (and even if they did that would only be one subject), so this should not be construed as an attack.

But at the same time, this is a serious error.  It's like reporting that September has 32 days in it, or that the Red Sox clinched a playoff spot this week.  Only worse, as it's perpetuating a myth that gets passed around as "common knowledge".  I attempted to bring this to the attention of the relevant parties shortly after it was reported, both on Google+, where SciAm blog editor Bora Zivkovic has been making effective use of the new social network, and the author's Twitter account (which is frequently used).

Despite near real-time feedback, days have gone by with no correction. Making a mistake is understandable, but failing to correct it is not.  Scientific American has written hundreds of articles on the state of science education, and has often been an effective advocate for ways to improve it.  But the authority they derive comes from their attention to scientific detail, so I hope we will now see a quick correction without further delay.

Talos: A troodontid with a leg up on the competition

It's my pleasure to introduce the newest member of the troodontid family:Talos sampsoni. Named for paleontologist Scott Samspon, Talos was described by Lindsay Zanno and others in the wonderful open source journal Plos ONE. Talos is the first troodontid to be named from the Kaiparowits Formation of Utah, making it about 76 million years old.  Cross-sections of the long bones suggest that the animal was between four and six years old, and while it hadn't stopped growing, it appeared to be reaching reproductive age at a smaller size than it's close relative Troodon.

Of note is that the specimen had a bone in its second toe that was injured and partially healed.  Since it appears to have been injured from violent trauma, it's consistent with with the idea that the "switchblade" toe was used in a way that could result in such an injury (presumably either attack or defense).  Also, since the rest of the foot shows no indication of the sort of limping or other adjustments you see from a prolonged foot injury, it also reinforces the idea that troodontids walked with the second toe off the ground (as shown above).

I was producing the above skeletal reconstruction for the Utah Museum of Natural History, who has constructed a new building filled with exciting displays that will opened Fall 2012.  It should be a state of the art exhibit that everyone should go see if they get the chance.  Dr. Zanno was working on describing Talos at the time, so I worked with her to incorporate the data into the skeletal.  There is quite a bit of troodontid material from the Kaiparowits, but only a fraction of it can be confidently assigned toTalos at this time.  Restricting ourselves to the type specimen looks something like this:

 You could be forgiven if your first reaction is "that's not very complete!".  But I was thrilled.  Why?  It's hard to remember sometimes, since Troodon is illustrated frequently, but the published material that Troodon is usually based on is really scrappy.  Here is a quick and dirty example of what was published in Dale Russell's 1969 paper:

To create the reconstruction (then called Stenonychosaurus) ofTroodon

that inspired him to create his infamous dinosauroid sculpture, Dale Russell had to combine all of the material from the Asian taxa Saurornithoides (as seen below) just to create a usable composite:

That made the size of the head and pelvis more clear, and provided more back and tail vertebrae to scale from, but look at how "dotted" the outlines of the limbs are - there still wasn't a good guide for scaling the limbs of any of these advanced troodontids.  A few unpublished specimens are out there, but none of those provided definitive limb proportions. Talos finally does that.

So in addition to the interesting conclusions on biogeography, ontogeny, and mode of life, Talos should be exciting to illustrators because it provides the first definitive glimpse of the limb proportions of advanced troodontids.  They aren't radically different from the proportions Dale Russell speculated on in the 1960s (although the forelimbs are shorter than some people have speculated since then), but the good news is that we now know for sure.  In fact at this point the only serious unknown is the neck, although there are plenty of more basal troodonts to base that on.

Running around like an Ornitholestes with it's head cut off!

If you watched Episode II of Dinosaur Revolution, you may have laughed at this Ornitholetes, who I'll refer to as Ichabod.  This may seem like an odd scene to pick for a scientific discussion, but I think it actually has something useful to teach.  Also, I'm partially responsible.  I should be clear, the story idea was not mine (that's above my pay grade), but it's something that was run by me, and I did not try to shoot it down (and still wouldn't).

Why?  I think it's reasonable.

No wait, let me explain: I'll grant you that there isn't much in the professional literature on the subject on the subject of "headless running" in animals, but from some criticisms I've read I think people maybe thinking about this the wrong way (i.e., wondering about the distribution of "headless-running" in birds like the behavior it's an advanced condition).

Vertebrates as a group have one of the more centralized nervous systems among animals (with some arthropods and especially some cephalopods as the other contestants in the "flexibility over redundancy sweepstakes"), but tetrapod nervous system evolution in general is a story of progressive centralization that (so far) culminates in mammals.  Even in humans, with our gobs of ridiculously calorie-hungry centralized gray matter, we still have autonomous reactions that don't require the brain to be involved (as anyone knows who's burned themselves and jerked their hand away before they felt the pain).  That said, we have gone a long way down the path of nervous system centralization, and if you cut a mammal's head off you may get some twitching but it won't run around; our limbs literally cannot coordinate themselves without the brain's involvement (although morbidly it does appear that the head itself retains some coordination afterwards, if medieval reports are true that heads react for up to a 15 seconds after a beheading).

This degree of centralization is the derived condition, not the primitive one.  So it seems unlikely that chickens are special here, except in as much as they more frequently get clean beheadings in the presence of human observers than most other birds (a quick Google search shows that turkey's exhibit this as well).  This should be true of lizards, crocs, etc too (diapsids as a whole).  So with enough experimental trials I'd fully expect an Ornnitholestes to do a good headless chicken impression.  Now, I'll grant you that this would require a pretty clean bite on the allosaur's part, and the odds of observing it in the wild would not be very high.  But the sequence was devised as be a surprising bit of humor in a scenario that was possible, not probable.

Given those parameters it seems reasonable enough to me.

Dinosaur Revolution Episodes III & IV on tonight on Discovery Science

Just a quick note, for those looking forward to the last two episodes.  They will be showing tonight on Discovery Science at 9 and 10pm (check local listings).

Reign of the Dinosaur Revolution

Copyright the Discovery Channel

One of the projects I've worked on the last two years was Dinosaur Revolution - although for most of that time it had the working name "Reign of the Dinosaurs" - which is still how I think of it in my head.  I'm sure I'll get over it.

Two of the episodes aired this last Sunday(September 4th) and the last two air this Sunday (Sep 11th).  (Note! The airing of episodes III  & IV have been postponed, as Discovery felt that the show wasn't appropriate for September 11th.  Thanks to Tom Holtz for point that out!)  I'm sure there will be other showings.  Anatomically-speaking, the dinosaurs will be the best to yet reach the screen (although there will be a few recent discoveries that couldn't get worked in on time).  The artists working to create these dinosaurs were about as good as it gets when you need to combine talent and knowledge of dinosaurs into one package.  The "story" in the episodes is more ambitious, attempting to hook viewers with no narration (although some minimal narration did get added late in production).  How well it works for you is probably a matter of opinion, although I enjoyed it.  Regardless, the show is well worth your time.

The reason I bring this up is over my long association with the project I've developed a new appreciation for how difficult it is to get anything on TV, let along maintain a specific scientific or artistic vision.  That might actually make for a good post or two.  But first I want to open this up for discussion.

So, who saw it, and what did you think?  Some of you commented on Google+ - feel free to cut and paste your response over here if you want (or else come up with something even snazzier to say).  I'll talk more about the experience, and the challenges, in the days to come.

Mike Habib's Great Flying Skeletals: Perspectives from Pterosaurs

Today we get the second guest post on pterosaurs.  In addition to his extensive background in biomechanics, Mike has started a blog with Justin Hall on biomechanics - you should check it out.  Now, one with the main programming...

Great Flying Skeletals: Perspectives from Pterosaurs

Scott has graciously invited me to do a little guest post from the perspective of a biomechanist.  While Skeletal Drawing focuses primarily on dinosaurs, I am going to depart a bit on this occasion and use pterosaurs to highlight some of the biomechanics issues that can occur with skeletal poses.

The Case of Running Pterosaurs

Many of you have probably seen pterosaurs reconstructed in the popular standard pose (particularly preferred by Greg Paul) where they are in mid-sprint, like so:

From here.  Copyright Gregory S. Paul.

It's an impressive pose for a flying animal.  A lot of people really like the look, and it lines up nicely with the same pose in birds, to which pterosaurs are often compared.  There's just one (big) problem with it: pterosaurs probably never sprinted around on their hindlimbs like the reconstructions show.

In the case of pterosaurs, the "standard running pose" is typically presented as a launch pose.  However, in the late 90's, Jim Cunningham made a strong case for quadrupedal launching in Quetzalcoatlus at a series of presentations for both engineers and biologists.  In 2008, I published a manuscript on a sizable comparative study I ran on bone structural strength estimates in the forelimb and hindlimb, which demonstrated that most pterosaurs probably launched quadrupedally rather than bipedally.

Now, I know you're thinking "Oh c'mon, Mike, you just don't like those bipedal running pterosaurs because they conflict with your personal results.  You're biased!"  I may be biased in some sense, but actually, that's not the problem.  I would not mind bipedal, sprinting pterosaurs if another study had used different data to support the idea.  But the reality is that no analysis has ever produced support for bipedal launching in pterosaurs.  In fact, so far as I am aware, my paper was the first attempt at testing between the two modes of launch.  There have not been a great number of biomechanical analyses run on pterosaurs, but there were a handful back in the 1970's and again in the early 2000's.  A few of these considered their performance during takeoff, and the authors all assumed a bipedal launch mechanism, as in birds.  The key word there is assumed - those studies asked the question "if pterosaurs launched like birds, then how would it work out?", but they never actually tested if a bipedal run was likely. I think the first lesson here is this: 

Don't reconstruct skeletal images in poses the animal was not known to reach, unless you are specifically trying to argue the plausibility in conjunction with the pose, with appropriate empirical data present.

Most viewers of a skeletal reconstruction will assume that the animal could (and did) the action shown by the skeletal pose.  More discriminating viewers may consider the issue more thoroughly, but either way this gets in the way of the point of a typical reconstruction. 

A typical skeletal is supposed to show off the anatomy.  If the paper you are illustrating happens to be arguing for a specific dynamic action, then it makes sense to show the animal in that pose.  If there is a dynamic pose that others have shown to be plausible, then that's fine, too - but not as a standard pose, because there will nearly always be some animal that you come across later that can't do it.  Nearly all terrestrial vertebrates can manage a slow walk, but only some can sprint - so choosing sprinting as your standard is risky.  Inevitably, something is going to end up sprinting in your illustrations that never did so in life. 

We can make an animal do anything we want in an illustration.  Scott made an Allosaurus do a handstand.  We could make Quetzalcoatlus launch by vaulting on its beak.  These extreme examples are obvious, but less extreme cases can be difficult to detect.  The ability to render a good illustration is powerful, because it can make the action or anatomy suggested by the image seem plausible, even if it's completely false or fabricated.  If you're an illustrator, use your powers wisely.

Where Does that Wing Go?

One really tricky issue with pterosaurs is the wings - we don't know for certain where the wings attached in most species, and even if we pick a particular attachment point, there are a range of potential resulting wing shapes (if you want to read more about this issue, check out the section on flight over at pterosaur.net).

(Image from David Hone's blog.)

Of course, the soft tissue extent need not affect a skeletal reconstruction, but the typical methodology for skeletal illustrations is to include a black body outline.  Usually this can made relatively conservative and follow typical muscle contours for vertebrates, but there is still a certain amount of conjecture there.  When there is a major soft tissue structure involved, like the wing of a pterosaur, this can get very tricky.  Any wing profile you show will be taken much the same as a pose: the viewer will assume you are explicitly supporting that particular wing shape.

One way to get around this issue is to leave off the wing membrane.  A simple black muscle contour can be drawn around the wing spar and then the wing itself can be left out.  This is, from my perspective, probably the best option if one is trying to simply show pterosaur skeletal anatomy in a neutral manner.  However, if your illustration is for wide audience, then be forewarned that you will probably need to make a note that the wing is left out somewhere in the caption, or else non-specialists will be very confused.

This same problem can arise with other taxa, of course, especially swimmers - flukes and fins tend to be largely soft tissue, and sometimes there is no exact match to a bony contour (look at the tails of cetaceans, for example).  In those cases, leaving off the fin or fluke might be especially confusing, so some sort of estimate might be required - but again, there is a danger of extrapolation.  I do not know what the best option in that case would be, though half-tone might be an option for showing speculative components.

Up and Away?

The last major challenge for pterosaur skeletal illustration is probably choosing whether to show them flying or walking.  As already discussed, bipedal sprinting is not likely, but pterosaurs certainly walked around (quadrupedally, as it turns out, based on the trackways), so one could easily show them walking, which would tend to match other skeletal reconstructions and work well for the standardization aspect.  However, it can be difficult to show off pterosaur anatomy well with a walking pose, because the folded wing gets in the way.  Some illustrators may also prefer to remind the viewer that the reconstructed critter was a flying animal.  As such, there are some good arguments for a flying pose.  As a biomechanist, I tend to be happy seeing either or both options, but this does bring one final consideration for skeletal standardization, which is that any given illustrator will inevitably need multiple standards.  You can't really show a mosasaur walking, for example.  As such, it may be that a separate standard is needed for each major mode of locomotion - flyers, swimmers, and walkers, as it were. 

There are additional concerns that could be raised for pterosaurs, and for biomechanics interests, but those three general issues (plausible posing, soft tissue suggestion, and multi-modal animals) are some of the factors that speak most quickly to biomechanists like myself.  Happy reading!  Thanks again to Scott for the guest slot.

--

Mike Habib

John Conway's Note from Pterosaur-land: End the Lateral Tyranny!

Today we have the first of two guest posts that look at skeletal reconstructions from a more pterosaur-centric point of view.  I feel these posts (and there will be some more guest posts on other subjects in the not-too-distant-future) are really important.  First, it helps break the stranglehold that dinosaurs have had on the subject matter thus far on SD Blog.  And second it diversifies the voices heard on the issues that impact the science behind reconstructing extinct animals.  And finally, they mean I can get by while doing less work.

First up is a message from John Conway, pterosaur skeletal drawing-er extraordinaire (all images belong to John), who also contributes to Pterosaur.net:

A Note from Pterosaur-land: End the Lateral Tyranny!

When I first started to draw skeletals, I still very much wanted to be Greg Paul when I grew up. But even so, I was not entirely happy with the sprinting pose he uses, I felt then, as I still do, that it’s too extreme, too hypothesis-laden, and unsuitable for many animals.

I tried walking poses and static poses, but never really settled on a standard. But its wasn’t until I started my first serious skeletals of pterosaurs that I really gave the matter of posing serious consideration. I gave it serious consideration because I had to: the running launch pose that Greg Paul was using is biologically inaccurate, and most walking poses for pterosaurs seriously obscure their anatomy.

Greg Paul notes that he ignores the slight foreshortening that would be apparent because of the bow in dinosaur legs. And that’s fine for dinosaurs, because the effect is tiny. Not so for pterosaurs, as their semi-erect stance would foreshorten the humerus (possibly) and the femur significantly, meaning lateral-view skeletals give misleading information about pterosaur proportions. I was still interested in a standards pose for all the animals I was interested in, but I just couldn’t think of one. So I gave up, and just went with something that would best elucidate pterosaur anatomy: a three-view skeletal with the limbs arranged in the only biologically plausible, non-foreshortened way I could think of.

As time goes by, I’ve come thing think I made a good decision. Every single time a sit down to do a skeletal, I find that the doing three or more views requires me to correct errors that I would have missed if I were just doing lateral or dorsal views. It works for just about any vertebrate (be it a sprawler, a strider, a hopper, or what ever)

However, I will admit that these skeletals aren’t exactly the most space efficient things in the world, and they may be a little dry for some contexts. However:

Gentlemen, We Can Repose Him, We Have the Technology!

Most people these days are using vector drawing programs like Illustrator to create skeletals (and if they aren’t, they should be). It is trivially easy to move these skeletals into just about any pose we like.

When I’m making a skeletal, usually it’s for my own reference, as a base for a life-reconstruction. For this purpose, multiple views are critical, posing much less so. Posing may be more important when we are trying to make a biomechanical point, or in popular works were we are trying to get across more information about the biology of the animal. Different contexts lend themselves to different poses, and perhaps we should be working with that, not against it.

So to end, I think it very doubtful that we will be able to come up with a standard pose that works for all vertebrates—and nor, perhaps, should we. Perhaps it is good enough to aim for clearly and accurately elucidating anatomy, posing depending on context.

-

John Conway