Re: [dinosaur] New Mongolia dinosaur finds + ceratosaurs + mesosaurs + more

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

> Gesendet:ÂSamstag, 30. Juni 2018 um 21:21 Uhr
> Von:Â"Ben Creisler" <bcreisler@gmail.com>
>
> For now, this is only an abstract, full open access paper to be posted later:
>
> Mark J. MacDougall,ÂSean P. Modesto, Neil Brocklehurst, Antoine 
> VerriÃre,ÂRobert R. Reisz and JÃrg FrÃbisch (2018)
> Response: A Reassessment of the Taxonomic Position of Mesosaurs, and a 
> Surprising Phylogeny of Early Amniotes.
>
> Frontiers in Earth Science (accepted paper)
> doi: 10.3389/feart.2018.00099
>
> [...] Over the last three decades several studies regarding the phylogenetic 
> placement of mesosaurs have helped to reinforce their position as the most 
> basal parareptile clade.

It's true that the phylogenetic analysis Laurin & PiÃeiro did was a very 
preliminary proof-of-concept type of thing that can definitely be greatly 
improved; and it's very reassuring to see that MacDougall et al. have begun to 
do that, seeing as Antoine has a large number of previously neglected but 
splendidly preserved complete mesosaur skeletons at his disposal. However, the 
sentence I'm quoting from their abstract is misleading. It gives the impression 
that there's been a series of phylogenetic analyses of Amniota "[o]ver the last 
three decades". In reality, there haven't been any since 1995.

With two exceptions, there have only been analyses of only synapsids, of only 
"parareptiles" (sometimes including mesosaurs) or of only 
"eureptiles"/romeriidans, with the other two groups either wholly excluded or 
only included as composite OTUs whose monophyly was not and could not be 
tested. One exception is the analysis by Reisz, Laurin and yours truly (2010, 
accompanying the redescription of *Apsisaurus*), which contained synapsids and 
romeriidans but no "parareptiles" whatsoever. The other was performed by 
Spindler et al. (2018, accompanying the description of *Ascendonanus*) but not 
shown in their publication, supplementary information included, because it led 
to such unusual results that the authors preferred to throw all three 
"parareptiles", one of the three romeriidans and both of the "diadectomorphs" 
out of their otherwise synapsid-only matrix before running a new analysis and 
showing its results as their fig. 30; this is explicitly stated on their p. 337.

As of right now, there is no good morphology-based phylogenetic analysis of 
Amniota, let alone "several studies" of that kind, any more than there is of 
Placentalia or Neornithes, 'good' even just by the simplest, most obvious 
criterion of taxon sampling; and nothing less than that can test "the 
phylogenetic placement of mesosaurs".