Re: Late Jurassic salamander from Liaoning, China (free pdf)

["David Marjanovic" <david.marjanovic@gmx.at>]

...Wow. This is the first lissamphibian ever, dead or alive, to have:

-- canals on any bones for the lateral-line organ. Aquatic lissamphibians 
today, including the larvae of metamorphosing taxa, have the organ, but it 
doesn't leave any traces on the bones, in stark contrast to the plesiomorphic 
condition for gnathostomes.
-- branchial denticles. The paper equates them with the gill rakers found in 
extant aquatic salamanders, but that's wrong: gill rakers are cartilaginous 
projections from the cartilaginous ceratobranchials, branchial denticles are 
teeth set on small dermal bone plates (which are preserved in *Beiyanerpeton*) 
which overlie the (usually) cartilaginous ceratobranchials.*

Further, I seem to have misremembered what my source said about the palatine: 
of all previously known adult salamanders, only the sirenids are known to 
retain it, while the others apparently lose it in ontogeny even if neotenic. 
Too bad I don't have that source here. Similarly, *Dicamptodon* has a toothless 
coronoid both before and after metamorphosis.

Also, I'm cited! :-) Though only because I took the other position in the 
confusion about the age of the Daohugou beds (and thus *Chunerpeton*), 
considering them likely as young as Valanginian in 2007. I don't take a 
position anymore; I'm just confused. :-) And I'm happy they managed to spell my 
name.

The supplementary information cites Tet Zoo (if only for a nomenclatural issue) 
and a few papers on the age of Daohugou that I should read.

All characters were unordered. Several are potentially continuous (like "small 
(0), medium (1), large (1)" and therefore should have been ordered. At least 
one should have been given a more complex stepmatrix. There are correlated 
characters, most obviously 77 and 78.

The article and the supplementary information are available as a single pdf 
here: 
http://www.pnas.org/content/early/2012/03/07/1009828109.full.pdf+html?with-ds=yes

* The branchial denticles of branchiosaurid temnospondyls were isolated, not 
set on bone plates. I think that's the only case of a tooth without a bone 
outside of Chondrichthyes. Anyway, *Beiyanerpeton* fairly strongly suggests 
that the lack of such plates is not a synapomorphy of branchiosaurids and 
lissamphibians, though I don't know if anyone has ever explicitly suggested it 
was. -- Plesiomorphically, there are 4 pairs of rows of denticle-bearing 
platelets on the 4 pairs of ceratobranchials; in branchiosaurids, there are 6 
pairs of rows of denticles (the middle two pairs of ceratobranchials each 
carried 2 rows, so that each gill slit had 2 rows of denticles on each side); 
*Beiyanerpeton* apparently preserves only 3 rows of platelets on each side, but 
I'm not quite sure if that's an artefact of preparation, and the paper doesn't 
mention this issue.