My Thoughts on Limusaurus' hand

[Michael Mortimer <mickey_mortimer111@msn.com>]

I finally had a chance to check out Xu et al.'s claims regarding Limusaurus and 
tetanurine hands being II-III-IV-V instead of I-II-III-IV, and my feelings 
mirror those of Andrea Cau on his Theropoda blog.
 
Xu et al. (2009) hypothesized Limusaurus may indicate metacarpals I-II-III-IV 
of tetanurines are homologous to metacarpals II-III-IV-V in other amniotes, 
based on several characters.
 
Digit I in Limusaurus and Aucasaurus are highly reduced, with no phalanges. Yet 
Ceratosaurus shows an articular surface for phalanx I-1, showing the condition 
in Limusaurus may be derived within ceratosaurs as opposed to a basal 
ceratosauroid (ceratosaur+tetanurine) state.
 
Metacarpal II is medially twisted in Limusaurus, Dilophosaurus and some 
Coelophysis specimens, similar to metacarpal I in other saurischians. 
Unfortunately, this is not easily determinable from most figures, so the 
condition in basal tetanurines is unknown.
 
While Xu et al. note metacarpal III lies ventral to metacarpal II in 
tetanurines, as metacarpal IV does to III in Limusaurus and coelophysoids, this 
is also true of metacarpal IV in tetanurines (e.g. Guanlong, as seen in its 
supplementary information; Tanycolagreus).
 
Similarly, while metacarpal I does not overlap II in non-tetanurines, this is 
seemingly also true in Xuanhanosaurus and Acrocanthosaurus (overlap is present 
in "Szechuanoraptor", Megaraptor, Torvosaurus, Allosaurus and Guanlong however).
 
There is a dorsolateral flange on metacarpal II of Dilophosaurus and Limusaurus 
which is similar to one on metacarpal I of some tetanurines (e.g. 
"Szechuanoraptor", Allosaurus, Guanlong). But Xuanhanosaurus and Megaraptor 
also have such a flange on metacarpal II, but not I, like basal theropods. Both 
flanges seem to exist in Acrocanthosaurus, while none exist in Aucasaurus and 
Torvosaurus.
 
Xu et al. state metacarpal II is more robust than I in non-tetanurine 
theropods, homologizing it to the robust metacarpal I in tetanurines, but the 
situation is more complex. It's clearly the size of the base which is 
important, since even Coelophysis and Dilophosaurus have metacarpal I shafts 
more robust than those of II. Yet basal tetanurines (e.g. Xuanhanasaurus, 
"Szechuanoraptor", Torvosaurus, Megaraptor, Acrocanthosaurus, Allosaurus) have 
metacarpal II more robust than I, while Aucasaurus and Herrerasaurus have the 
opposite condition. This is true in Xu et al.'s tetanurine example of Guanlong 
too, while even their example of Deinonychus has more proximal area and depth 
on metacarpal II, just less width.
 
Phalanx I-1 in tetanurines is said to be longer than phalanx I-1 in 
Herrerasaurus, Dilophosaurus and ceratosaurs, but phalanx I-1 is not preserved 
in any ceratosaur except for two questionably identified elements in 
Masiakasaurus. Furthermore, it is not as if any phalanx on digit II in 
Dilophosaurus or Herrerasaurus is notably more elongate than their phalanx I-1, 
and some basal tetanurines like Torvosaurus actually have an extremely short 
phalanx I-1.
 
Metacarpal II is longest in tetanurines, while III is longest in more basal 
theropods. Yet Limusaurus and Ceratosaurus resemble tetanurines in this (contra 
Xu et al.'s statements and measurements about the former), and Dilophosaurus 
and Megapnosaurus are polymorphic (e.g. II longer in the paratype of 
Dilophosaurus, III longer in the holotype).
 
There is a proximal dorsolateral process on metacarpal III in coelophysoids and 
Limusaurus, similar to one on metacarpal II in some basal tetanurines like 
Guanlong. Yet Guanlong also has a process on metacarpal III, which partly 
covers metacarpal IV, even though the latter is not illustrated in Xu et al.'s 
paper. Acrocanthosaurus and Allosaurus also have a processes on metacarpal III, 
while "Szechuanoraptor" lacks processes on metacarpals II or III. This process 
on metacarpal III of tetanurines could be homologous to the process on III in 
basal theropods as easily as it could the process on II.
 
Finally, Xu et al. state metacarpal III in tetanurines is short, slender and 
proximally triangular like metacarpal IV in basal theropods. Of course, 
metacarpal IV in tetanurines is also short and slender (even moreso than III), 
with those of Xuanhanosaurus and "Szechuanoraptor" resembling metacarpal IV in 
basal theropods more than their metacarpal III does. This is a case where the 
more reduced metacarpal III in derived tetanurines like Guanlong and 
Deinonychus (illustrated by Xu et al.) resembles basal theropod metacarpal IV 
more than metacarpal III in basal tetanurines (e.g. Xuanhanosaurus, 
"Szechuanoraptor", Torvosaurus, carnosaurs) do, with the thicker shaft and 
robust distal articulation in the latter taxa. As for their triangular proximal 
outline, metacarpal IV in Dilophosaurus and Limusaurus are more round than 
triangular, but "Szechuanoraptor" shows basal tetanurines have triangular 
metacarpal IV too in any case.
 
Xu et al. ran a phylogenetic analysis which determined that when characters 
states are ordered, the resulting tree assuming tetanurines have digits 
II-III-IV-V is six steps longer than if they are assumed to have digits 
I-II-III-IV. The length of the unordered trees is equal, but leaving characters 
unordered potentially leads to ridiculous "intermediate synapomorphies" like 
coelophysoids and Herrerasaurus being diagnosed by having a single phalanx on 
digit IV (not more or less) or taxon being diagnosed by having an intermediate 
ratio, as opposed to relatives with low and high ratios. Furthermore, neither 
Xuanhanosaurus, "Szechuanoraptor" or Megaraptor were included in their matrix, 
though these taxa show high homoplasy if II-III-IV-V is assumed, as noted 
above. I conclude that there is little evidence tetanurine hands lacked digit I.
 
Mickey Mortimer
The Theropod Database- http://home.comcast.net/~eoraptor/Home.html



[http://gfx1.hotmail.com/mail/w3/ltr/i_charity.gif] EMAILING FOR THE GREATER 
GOOD
Join me