Friday, February 5, 2010

On the structure of hind wings

Above: Anchiornis model from NatGeo. Something is wrong with those foot feathers...

With all the recent hubbub over Anchiornis and its coloration, have come several pretty cool life reconstructions showing the animal as it would, supposedly, look in life. I posted one in my blog yesterday on the topic, and NatGeo has a cool 3D model on their site, here (and figured above).

One thing I've noticed about all these images is the very... standard rendering of the infamous hind wings (shared with Microraptor and Pedopenna). Much debate has surrounded these appendages, mainly relating to their ability to glide. A whole Nova special was made showing different attempts to reconstruct gliding posture, from biplane-like, to sprawled-out-sideways, to rear-swept canopy. And especially in that special, the same thing bugged me about the hind wings:

Every configuration required the primary foot feathers to emerge at a different angle from different parts of the limb!

It shouldn't be so hard to figure out how these feathers attached, which would help us suss out their potential uses. First of all, the reconstruction of Anchiornis above, and by extension reconstructions of Microraptor with the legs sprawled out, make little anatomical sense to me. Maybe I'm mistaken, but I think we should look to the example of how front wings are put together to inform on this, as developmentally the two are almost certainly related.

Above: Incorrect, pronated-handed Velociraptor skeletal by Frederik Spindler, licensed. Just where are those primary feathers supposed to be attaching? To the underside of the fingers?

Back in the days of pronated theropod hands, many similary ridiculous renderings of feathered theropods were made, many by me. Of course, we all know how wings 'look' when folded - primary and secondary feathers swept back along the side of the body. The problem is that, if the hand is incorrectly shown as pronated, this arrangement makes no sense. On a pronated hand, the feathers emerge laterally from the edge of the finger and hand, and from the lateral side of the ulna. Only when the hand is not pronated does the proper wing configuration emerge. If real wing anatomy were taken into consideration, artists of bunny-wrist theropods would show the primary feathers sicking out to the side of the animal, not flush with the body, when the arms are folded. Otherwise, the feathers would be attached to the palm of the hand, which is ridiculous.
Above: Probably incorrect rendering of Microraptor. Just how is this animal supposed to sit down?

So why isn't this also true of the foot? My guess: it is. For all intents and purposes, the foot is the same configuration as the hand, only it really is pronated. The toes and metatarsals are equivalent to the fingers and metacarpals. I'll put this to the experts out there, because it's something I've never seen addressed and makes no sense to me. If the hind wing feathers projected posteriorly from the metatarsals, how are these animals supposed to have sat, or brooded their eggs? We have several fossil examples of theropods sitting, sleeping, or brooding, all with metatarsals fully in contact with the ground. If there were very large, rigid feathers pointing straight down from the foot, that would be impossible. Furthermore, developmentally, it seems we would expect the legs to mirror the arms, with primary leg feathers articulating from the top of pedal digit 2 and metacarpal 2. Therefore, sticking straight out to the side when standing.

Above: An animal capable of sitting, with sticky-out hind wings. AMNH.

Now, maybe I missed something so please correct me if any of my reasoning is wrong. But in short, I believe the biplane or semi-biplane model is the only one anatomically feasible, barring some novel adaptation like feathers with a hinge at their base to allow them to deploy outward like an X-Wing from Star Wars. The pedal primaries have to have pointed out to the side, while the tibial primaries of the lower leg would still point backward, as in the model above. So all those recons purporting to show Anchiornis as it really was miss a pretty important point. Not to mention the naked, beak-like snout, doesn't anybody look at fossils anymore? ;)

Above: This pigeon with sticky-out, biplane configured foot feathers is apparently able to sit down.

Another note on this--maybe my whole reasoning about comparing the front to hind wings is off. Can't we then compare to modern hind winged animals? Like the one pictured above? Those foot feathers do not point backward and, as you can see, the animal is therefore able to sit. Imagine those feathers configured like the Anchiornis model at the top of this post. It just doesn't work.

Thursday, February 4, 2010

In other news... Anchiornis now in full color!

Above: Left - so close, yet so far away. Right: The actual colors of a dinosaur.

Oh man, I was SO CLOSE! And I've been planning a new digipainting of this guy too, but was hesitant to do so figuring the color would be worked out in a few months. You can all thank me for summoning this study with mind bullets. Thanks to Ed Young pointing this out on his blog!

Li et al. 2010. Plumage Color Patterns of an Extinct Dinosaur. Science http://dx.doi.org/10.1126/science.1186290

P.S. How flipping beautiful is this new specimen? Go to hell Berlin Archaeopteryx specimen ;)

Above: Awesomeness.

Banji the Hunted

Above: Possible extents of oviraptorid beaks, from Jansen's thesis.

Recently, some of my adventures on Wikipedia have addressed the question of oviraptorid beaks. Now, contain yourselves, I know this sounds a little too exciting. But it's really a very rarely addressed topic. It's obvious that oviraptorids had beaks of some kind, with their strong, toothless, pointy jaws that don't quite close right when unsheathed by rhamphotheca (the keratinous covering that forms the beak). But just how these beaks looked and to what extent they covered the jaws hasn't been studied in a lot of detail.

I asked about this a few times back on the old version of dinoforum, and some commentors there (especially oviraptorer Jaime Headden) were very helpful in pulling educated guesses based on modern analogues. Jaime's conclusion, IIRC, was that most oviraptorids have been constructed with beaks that are too large or, more specifically, cover too much of the skull and jaws. You can see the difference in two versions of my own ovi profile drawings, old one here with extensive beaks, new one with corrected beak after Headden here. Note also the more subtle difference--the beaks on the newer version meet flush, without the upper beak overlapping the other (what use would that be?) and the beak does not incorporate the nostril, which is the exception, not the rule in modern beaked animals like birds and turtles.

The first real scientific work on this topic has not yet hit the official literature, but an unpublished phD thesis by Stig Jansen addresses this topic and can be read here. Jansen also has a similar unpublished paper floating around for ornithomimid beaks, I'm sure you're Google-fu will turn it up.

Jansen's thesis comes to essentially the same conclusions as Headden did, and presents two potential extremes for the extent of oviraptorid beaks, illustrated by him above. Interestingly, one of the options has a keratin-less crest. Traditionally, crested oviraptorids or those with tall, pronounced skulls have been restored with horny, cassowary-like casques, though there was never any direct evidence for this, and they could just as well have been covered in skin or feathers.
Above: Skull reconstruction of Banji, note the striations on the crest. Fig. 2 from Xu & Han 2010.

Well, most of them, but maybe not a newly described genus of oviraptorid (named Banji long, or "dragon with striped crest"... this is, incidentally, another case of bad grammar in a binomial. The specific name is supposed to be an adjective, not a noun, which makes for very awkward translations like Mei long, or "dragon, comma, soundly sleeping". Also, it sounds like Benji.). Anyway, Banji reserves some unique vertical striations on its crest. The implications of this feature aren't discussed in the paper, but the first thing that jumps to my mind is that these may form the bases of more pronounced striations in an overlying keratinous crest. As Jansen shows, underlying bony features of a bak often subtly reflect larger features of the keratin, like the keratinous pseudo-teeth on the beaks of some oviraptorids. A future paper will describe Banji and its implications in more detail, so we'll see if this ends up supported by actual study of the fossils, but right now it's a very interesting possibility that may argue for the more extensive beak suggested by Jansen's thesis.