[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: [dinosaur] Chunerpeton (Jurassic salamander) osteology + tetrapod quadrupedal locomotion [...]



> Yu-Fen Rong, Davit Vasilyan, Li-Ping Dong & Yuan Wang (2020)
> Revision of Chunerpeton tianyiense (Lissamphibia, Caudata): is it a 
> cryptobranchid salamander?
> Palaeoworld (advance online publication)
> [...]

I'm happy that this paper is out! The description and comparison is top-notch 
and publishes a lot of, for the most part, previously unknown data that are 
very interesting for the phylogeny as well as for the evolution of the life 
history of salamanders.

That said, be aware that one of the first author's five institutions put both a 
publication requirement and a time limit on her doctoral thesis: she needed 
this paper to graduate, and she needed it fast.

In combination, these two constraints created _perverse incentives_. I have not 
tried to find out which institution it was, but whichever one it was should be 
_downright ashamed_ to force students to take shortcuts and to, in effect, 
pressure reviewers and editors to let that happen.

I am _not_ alleging that the authors, say, plagiarized, invented data, scooped 
anyone, or anything like that; it is clear that no such thing happened. The 
shortcut in this case was merely to withdraw the manuscript from one journal 
after several rounds of review and submit it to Palaeoworld. At the other 
journal, two of the reviewers had already recommended acceptance. I was the 
third. _I accept part of the blame_ because I didn't review as fast as I could 
(in the later rounds I was aware of the general situation) and because I 
overlooked some important things in my first two reviews. Between my insistence 
that the most glaring problems with the phylogenetic analysis and the 
presentation of its results should either be remedied or the whole phylogenetic 
analysis omitted, on the one hand, and the time constraint on the other, the 
manuscript was held up for too long, and now all those problems are still in 
the paper.

For example, characters 77 and 78 are duplicates of each other. They were also 
duplicates in the previous version of that matrix (2019), and the one before 
that (2016), and the original (2012) which had taken those two characters from 
two different sources. (All authors of all three of those papers are not 
involved in the present one.)

> One of those taxa, Chunerpeton tianyiense, has been considered as a crown or 
> stem member of the family Cryptobranchidae,

To the best of my knowledge, it has _never_ been considered a crown member. The 
claim was always that it was on the stem (and therefore, by the second author's 
own nomenclature, a member of Pancryptobrancha but not of Cryptobranchidae). I 
brought that up in my last review...

> significant for implying a long evolutionary history for cryptobranchids and 
> for calibrating the molecular clock of Caudata evolution.

I'm not aware of a molecular-dating study that used *Chunerpeton* to calibrate 
the origin of Cryptobranchidae, i.e. the split of the extant *Cryptobranchus* 
and *Andrias* from each other. I said so, too, and the authors didn't mention 
any. What *Chunerpeton* has been used for often is to calibrate the split 
between Pancryptobrancha and the total group of the extant Hynobiidae.

> Our phylogenetic analyses

Keep in mind that parts of the results are not reported in the paper. This is 
due to the use of majority-rule consensus trees. It is quite dangerous to 
represent the results of a phylogenetic analysis by a majrule tree. First, all 
most parsimonious trees are equally parsimonious; whether they agree with the 
majrule tree is _not_ a support measure in a parsimony analysis. Any node in 
the majrule tree that is not marked "100" has at least one fully equal 
alternative that is swept under the rug for no defensible reason. Second, the 
majrule tree is not necessarily identical to any single most parsimonious tree: 
nodes that occur together in the majrule tree may or may not occur together in 
any single MPT.

The emergency stop in Spaceballs should never be used, and neither should the 
majority-rule consensus. I said so in all my reviews.

> consistently place Chunerpeton as a stem Caudata outside of Cryptobranchidae 
> and crown salamanders. Exclusion of Chunerpeton from Cryptobranchidae will 
> require reconsideration of the origin time for Cryptobranchidae and 
> recalibration of the molecular clock for the whole caudatan tree.

Again, exclusion of *Chunerpeton* from Cryptobranchidae will require nothing at 
all, because it was never included. Exclusion of *Chunerpeton* from the 
salamander crown-group, which has been named Caudata in Phylonyms while nobody 
was looking, would indeed require such changes â but, because I was aware of 
the time constraint, I not only recommended specific changes to the matrix, I 
also implemented them, ran the analysis, and included that in my second-to-last 
review. *Chunerpeton* came out just barely inside Caudata. (That's after the 
molecular topology of the extant caudates was imposed as a constraint; without 
it, all sorts of nonsense happen.)

The good news is that the (I hope) newly-minted Dr. Rong will continue to work 
on salamander phylogeny in the near future and has begun to work on several 
improvements to that matrix. Better-tested trees are, in other words, 
forthcoming. Just don't put too much trust in the ones published in this paper 
just yet.

Personally, based on my impression of the new features reported in the (again: 
excellent!) descriptive part of the paper, I would not at all be surprised if 
*Chunerpeton* ended up on the stem after all. Salamander phylogeny, and the 
comparative osteology of fossil and especially extant salamanders, is pretty 
poorly understood â which is, of course, part of the reason why this paper 
represents a long-desired big step forward.

> Michael Berenbrink (2020)
> The role of myoglobin in the evolution of mammalian diving capacity â The 
> August Krogh principle applied in molecular and evolutionary physiology.
> Comparative Biochemistry and Physiology Part A: Molecular & Integrative 
> Physiology 252: 110843
> [...]
>
> Abstract
> After the Devonian tetrapod land invasion

...all that came out of the _Devonian_ land invasion was *Ichthyostega*, the 
Godzilla version of a mudskipper, still with internal gills. We (and the 
salamanders) are descended from one of probably two Early Carboniferous origins 
of amphibious tetrapods from aquatic ones.