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

Re: [dinosaur] Rise of tetrapods after Carboniferous rainforest collapse (free pdf)

> Emma M. Dunne, Roger A. Close, David J. Button, Neil Brocklehurst, Daniel D. 
> Cashmore, Graeme T. Lloyd & Richard J. Butler (2018)
> Diversity change during the rise of tetrapods and the impact of the 
> 'Carboniferous rainforest collapse'.
> Proceedings of the Royal Society B 2018 285 20172730
> DOI: 10.1098/rspb.2017.2730

This paper spells NÃÅany "NyraÃy". Between the seven authors, the editor and 
the reviewers, nobody thought to look it up; the letter à doesn't even exist 
in Czech.

But that fact merely has entertainment value. What _might_ affect the actual 
scientific results is the topology of the tree in the supplementary 
information. Inevitably, that tree is a handmade supertree. Evitably, it has... 
potential issues. For example:

â Of the cited papers, three that date from 2001, 2007 and 2017 bear on the 
position of AÃstopoda. All aÃstopods are placed where the 2001 and 2007 
papers have them, except for *Lethiscus*, which is placed (more or less) in the 
radically different place where the 2017 paper found it and *Coloraderpeton*. 
*Coloraderpeton*, however, remains in the old place. The 2017 paper argued, of 
course, that all aÃstopods belong in that place; aÃstopod monophyly has not 
been questioned there or anywhere else.
â Different papers have found the colosteids in slightly different positions. 
The tree shows *Colosteus* in one of those places, and *Greererpeton* in the 
other; I'm not aware of any publication that would question the membership of 
*Greererpeton* in Colosteidae, within which it even seems to be the 
sister-group of *Colosteus*. (No other colosteids were sampled.)
â *"Archegosaurus" dyscriton* (i.e. *Memonomenos*) is in a defensible place, 
but I have no idea how the real *Archegosaurus* (*A. decheni*) ended up within 
*Trimerorhachis*, a very different beast as aquatic temnospondyls go. It should 
be in the same faraway polytomy as *Memonomenos*.
â Many of the species in the scope of the taxon sample have been lumped or 
otherwise renamed in the last hundred years. The tree lacks most of the 
lumping, so that many species are represented two or three times. For example, 
the three species of *Oestocephalus* were lumped in 1998; "*Cheliderpeton 
lellbachae*" is a barely published specimen that was said in 2009, in the 
redescription of the (then) four species of *Sclerocephalus*, to be probably 
synonymous with one of those four species and to require "re"description; and 
both *Lysorophus* species are nomina dubia â most of the material referred to 
them belongs to *Brachydectes elongatus*, as shown in a paper from 1991 and 
another from 2016. (...except that the type specimen of *B. elongatus* may not 
be diagnostic either; but there clearly are two species of *B.* in any case.) 
*L.* and *B.* aren't even sister-groups in the new tree; I have no idea how 
they ended up apart, because that's never been proposed.
â If you scroll far enough to the right, you'll see a completely novel 
amniote topology:
(Sauropsida (Parareptilia (Mesosauridae, Synapsida))))
The current consensus is:
(Synapsida (other sauropsids (Mesosauridae, other parareptiles)))
Now, the current consensus hasn't been tested in twenty years, and there are 
vague hints in recent literature that all sorts of things might be wrong with 
it. But not only has the new topology never been proposed before, it's also not 
argued for, or even mentioned, in the new paper...

Again, I haven't tested whether fixing all that stuff would change the results. 
But it's certainly _possible_ that it would, and I can't help but express 
surprise that between the seven authors, the editor and the reviewers, nobody 
noticed how off that tree is. It looks as if, of all these people, one author 
made the tree, and nobody else ever looked at it before it was published.

I repeat my usual plea (ceterum censeo...) to reviewers to read the 
supplementary information.