Oh, what the hell: given that we’ve already covered a new Wealden theropod, and have looked a bit at the palaeobiology of Majungasaurus within the week, I may as well resist my urge not to do more dinosaurs. In other words, I may as well cover Limusaurus as well, despite my previous protestations. It is, after all, a pretty incredible animal. Limusaurus is a small, long-legged dinosaur with short, gracile forelimbs, tiny hands, a slender neck and tail, a short, deep skull, and a slender lower jaw with a down-curved tip. It is toothless, and beak tissue is preserved around its jaw margins. This description makes it sound something like a sort of generalised small ornithischian (bar the toothlessness), but the incredible thing is that this is a ceratosaur, the first from the Jurassic of Asia (Xu et al. 2009) [the holotype is shown below. Scale bar = 50 mm. The coloured inset shows bone histology within the fibula: growth rings show that the specimen was about 5 years old at death. The yellow arrows point to a crocodyliform lying right next to the dinosaur!].
The big deal for me is that this dinosaur really is something radically new: all the ceratosaurs known previously are carnivores (or so it’s been assumed), typically with big heads, big teeth and all the other features you associate with predatory theropods. Yet again, the fossil record surprises us by showing that we can still discover entirely new, totally unexpected lineages.
I have to note to begin with that ‘ceratosaur’ is now being used in a more restrictive sense than the Gauthieresque version you might be familiar with: Ceratosauria does not include coelophysids or dilophosaurids in recent phylogenies, but is used only for the ceratosaurid + abelisauroid clade (or, more specifically, for all theropods closer to Ceratosaurus nasicornis than to birds).
One frivolous observation I have on Xu et al.’s paper is that the animal’s name – Limusaurus inextricabilis – is – in my totally irrelevant and entirely idiosyncratic personal opinion – inexcusably lame. Meaning something like ‘mire lizard, impossible to extricate’, it refers to the fact that the type specimen* seems to have died after getting trapped in mud. Oh, come on. A toothless, beaked, probably herbivorous, Asian ceratosaur with a fused sternal plate (!), surreal stumpy little hands, gastroliths, super-long feet… and they give it a name that tells us that
one two individuals died after getting stuck in mud? Sigh.
* One of two specimens that are known, both are from the Oxfordian Shishugou Formation of the Junggar Basin, Xinjiang.
Its very long, slender hindlimbs indicate that it was cursorial. While long legs and cursorial abilities are not at all unusual for theropods, note that this is not necessarily the case for non-coelurosaurs. Indeed, ceratosaurs typically have proportionally short, stocky hindlimbs compared to coelurosaurs: look at Majungasaurus and the small abelisaur Masiakasaurus [shown here, from Carrano et al. (2002)], for example. While I’ve said that Limusaurus is ‘radically new’, you’ll note that this is only partly true, as one of the most interesting things about this animal is that, superficially, it’s very similar to certain Cretaceous coelurosaurs (notably ornithomimosaurs), and to the non-dinosaurian shuvosaurids of the Triassic. None of these groups are closely related, yet they converged on highly similar bauplans: shuvosaurids and Limusaurus are especially similar in having large mandibular fenestrae and abbreviated forelimbs (Xu et al. 2009).
All the fuss over those weird little hands
It’s the very weird little hands of Limusaurus [shown here] that are getting all the attention. It seems that four metacarpals were present: digit I (the thumb, or pollex) lacked phalanges (only a small metacarpal remained), digits II and III had large, robust metacarpals and three phalanges each, and digit IV was short, and perhaps with a tiny phalanx. While it’s generally been thought that non-avian theropods exhibit what’s been termed lateral digit reduction (LDR) – that is, they lost digits V and IV (and III in tyrannosauroids and a few others) – it’s pretty clear that Limusaurus exhibited bilateral digit reduction (BDR): that is, that it lost digits from both the lateral and medial sides of its hand.
As you’ll surely know, embryologists have often (though not always) argued that birds exhibit BDR, such that their tridactyl hands represent digits II, III and IV rather than the I, II and III thought universal among coelurosaurian theropods. Those who contend that birds cannot be theropods have latched on to this as an integral bit of their case: Alan Feduccia in particular has repeatedly said that bird hands and theropod hands are fundamentally different, and that this degree of difference bars theropods from avian ancestry (Burke & Feduccia 1997, Feduccia 1999, 2001, 2002, 2003, Feduccia & Nowicki 2002) [developing ostrich hands from Feduccia & Nowicki (2002) shown below]. Yeah, as if one feature – no matter how profound or major – can somehow outweigh tens of others: what excellent science. The hypothesis (note: hypothesis) that bird hands represent digits II-IV rests mostly on the fact that the primary axis of condensation (the first digit precursor to appear in the embryonic hand) corresponds to digit IV: because bird embryos grow two fingers medial to this axis, these two must be digits III and II (incidentally, this is contested by some embryologists and is not universally accepted. To keep things as simple as possible, we’ll ignore that for now).
Despite what Feduccia and his ‘birds are not dinosaurs’ colleagues state, the morphological evidence showing that birds really are theropod dinosaurs is overwhelmingly good, so if birds and other theropods really do have different digit patterns in the hand, something unusual must have occurred during evolution. One idea is that a frame shift occurred: that is, that the condensation axes that originally produced topographical digits II-IV became modified during later development, such that the digits that grew in these places came to resemble topographical digits I-III instead of II-IV (Wagner & Gauthier 1999). If the frame shift hypothesis is valid, then – somewhere in theropod evolution – the ‘true’ digit I was lost, and ‘true’ digit II became digit I. However, evidence from Hox genes indicates that the condensation axis for embryonic digit I receives a Hox signal normally associated with…. topographical digit I, thereby showing that the bird ‘thumb’ really IS the thumb (Vargas & Fallon 2005, Vargas et al. 2008).
Here’s where we come back to Limusaurus. Xu et al. (2009) contend that, having lost topographical digit I, this theropod was in the process of turning its topographical digit II into a developmental digit I. In other words, that its hand provides evidence for the frame shift hypothesis. Because Limusaurus is basal to Tetanurae (the theropod clade that includes birds and all of the more bird-like theropods), then the digits I-III we see in all tridactyl tetanurans are, after all, topographical digits II-IV, and the frame shift occurred well prior to the origin of birds. If true, this would provide a tidy explanation of the supposed discrepancy that exists between the embryological data and the inferences that palaeontologists have made from fossils. This is the main take-home message from Xu et al. (2009), and it’s the bit of the paper that everyone is talking about.
By this way, this debate is pretty involved, so well done if you’ve followed it so far.
However… how sure are we that Limusaurus was really caught in the act of ‘frame shifting’? There’s no doubt that its digit I was absent, but was its topographical digit II really morphing into a ‘new thumb’? I’m not convinced. While metacarpal II in Limusaurus does possess a few subtle features normally seen on metacarpal I (such as a dorsolateral flange), we may just as well be seeing a very weird, reduced little hand where all the digits were becoming short and stumpy.
We already know that ceratosaurs did some very freaky things with their hands. One of the best known ceratosaurs, Carnotaurus, does not have long, allosaur-like clawed fingers as usually shown in artwork and movies: instead, it had short, apparently claw-less digits I-III, while digit IV was absent, with only a posteriorly projecting metacarpal remaining. In its close relative Aucasaurus, digit I was absent and only represented by a conical metacarpal, metacarpals II and III had just one or two phalanges each, and metacarpal IV was also conical and, apparently, devoid of phalanges [Aucasaurus forelimb shown here, from Coria et al. (2002)]. Because Carnotaurus and Aucasaurus are deeply nested within the ceratosaur clade Abelisauridae, it’s never been suggested that their bizarre, reduced little hands might be at all relevant to what happened in the avian hand. Why, then, is Limusaurus regarded as being so informative?
Part of the answer lies in the phylogenetic position recovered for this dinosaur by Xu et al. (2009): they state that it occupies ‘a very basal position within Ceratosauria’ (p. 941), and hence they imply that its hand anatomy might tell us something about the ancestral condition for tetanurans. However, one of the first things that struck me about Limusaurus is how gracile its metatarsus is. This is reminiscent of what’s seen in Elaphrosaurus (the ceratosaur regarded by Xu et al. (2009) as the closest relative of Limusaurus), but it’s also the condition present in noasaurid abelisauroids (like Noasaurus and Masiakasaurus). Andrea Cau reports at Theropoda that he finds Limusaurus to be a noasaurid, specifically one close to the much larger African taxon Deltadromeus (this would be amazing if true, as it raises the possibility that Deltadromeus was an edentulous herbivore/omnivore too – wow!). If Limusaurus is deeply nested within Ceratosauria, rather than down at its base, its peculiar hand morphology might – one could argue – be less significant in terms of big-picture implications. More work is needed to sort this out. Xu et al. (2009) analyse a lot of character data, and, while they do report finding Limusaurus to group with at least some noasaurids in at least some analyses (this is in the 101-page-long supplementary data, not in the published paper), they conclude that Limusaurus is close to Elaphrosaurus, and that both are outside of a clade that includes Ceratosaurus and abelisauroids (noasaurids + abelisaurids). This position is only very weakly supported (it hangs on one character!) and, as noted above, might be wrong.
Don’t get me wrong: I think it’s certainly conceivable that Limusaurus says what the authors think it might… that a frame shift occurred pretty early in theropods, and that the topographical second digit came to function as a homologue of the first digit. However, the hand of Limusaurus is so weird, and the phylogenetic position of this theropod within Ceratosauria is so contestable, that I think this is open to question. And, as expressed here – I hope – I really think that the more interesting details about this animal are being mostly overlooked. A toothless, beaked ceratosaur, convegent on ornithomimosaurs and shuvosaurids, from the Jurassic of Asia… wow. Just, wow…
That really will be it on dinosaurs for a while now, I promise. Limusaurus has also been covered by Dave Hone (I should hope so, he’s on the authorship), Pharyngula, at Panda’s Thumb, by Carl Zimmer, and by a million others I’m sure. This is why I don’t like blogging about new discoveries. Anyway, for previous articles on Mesozoic theropods see…
- Troodontids and owls: oh, the irony (part II)
- My most favouritest dinosaurs: ceratosaurids
- Long and Schouten’s Feathered Dinosaurs, a review
- Epidexipteryx: bizarre little strap-feathered maniraptoran
- A month in dinosaurs (and pterosaurs): 1, therizinosauroid fuzz
- A month in dinosaurs (and pterosaurs): 2, of alvarezsaurids and avialians
- Myth of the six-foot super-owl
- 100 years of Tyrannosaurus rex
- Oh no, not another new Wealden theropod!
- The aquatic Majungasaurus, or not
Refs – –
Burke, A. C. & Feduccia, A. 1997. Development patterns and the identification of homologies in the avian hand. Science 278, 666-668.
Carrano, M. T., Sampson, S. D. & Forster, C. A. 2002. The osteology of Masiakasaurus knopfleri, a small abelisauroid (Dinosauria: Theropoda) from the Late Cretaceous of Madagascar. Journal of Vertebrate Paleontology 22, 510-534.
Coria, R. A., Chiappe, L. M. & Dingus, L. 2002. A new close relative of Carnotaurus sastrei Bonaparte 1985 (Theropoda: Abelisauridae) from the Late Cretaceous of Patagonia. Journal of Vertebrate Paleontology 22, 460-465.
Feduccia, A. 1999. 1,2,3 = 2,3,4: accomodating the cladogram. Proceedings of the National Academy of Sciences 96, 4740-4742.
– . 2001. Digit homology of birds and dinosaurs: accomodating the cladogram. Trends in Ecology & Evolution 16, 285-286.
– . 2002. Birds are dinosaurs: simple answer to a complex problem. The Auk 119, 1187-1201.
– . 2003. Bird origins: problem solved, but the debate continues… Trends in Ecology and Evolution 18, 9-10.
– . & Nowicki, J. 2002. The hand of birds revealed by early ostrich embryos. Naturwissenschaften 89, 391-393.
Vargas, A. O. & Fallon, J. F. 2005. Birds have dinosaur wings: the molecular evidence. Journal of Experimental Zoology (Mol Dev Evol) 304B, 86-90.
– ., Kohlsdorf, T., Fallon, J. F., VandenBrooks, J. & Wagner, G. P. 2008. The evolution of HoxD-11 expression in the bird wing: insights from Alligator mississippiensis. PLoS ONE 3(10): e3325 doi:10.1371/journal.pone.0003325
Wagner, G. P. & Gauthier, J. A. 1999. 1,2,3 = 2,3,4: a solution to the problem of the homology of the digits in the avian hand. Proceedings of the National Academy of Sciences 96, 5111-5116.
Xu X, Clark JM, Mo J, Choiniere J, Forster CA, Erickson GM, Hone DW, Sullivan C, Eberth DA, Nesbitt S, Zhao Q, Hernandez R, Jia CK, Han FL, & Guo Y (2009). A Jurassic ceratosaur from China helps clarify avian digital homologies. Nature, 459 (7249), 940-4 PMID: 19536256