Re: [dinosaur] Ascendonanus, new arboreal varanopid synapsid from Permian of Germany; Microvaranops from South Africa

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

Oh. I forgot that this conference talk 
https://urldefense.proofpoint.com/v2/url?u=https-3A__www.youtube.com_watch-3Fv-3DfnqG2UOCF0s&d=DwIFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Ry_mO4IFaUmGof_Yl9MyZgecRCKHn5g4z1CYJgFW9SI&m=xr3CzqGjGzyyfgXs7Hzc25tk1uPFiYHa46aYILtY1ao&s=cm5PPoqxuOwz1naHYTmYff1ARrOaxw1s9xSaaKgb--E&e=
 , which used the matrix of Reisz et al. (2010) with one correction and one 
added terminal taxon, puts Varanopidae into Diapsida (adding the next step to 
the ancient confusion described in the new paper). So perhaps epidermal scales 
are a sauropsid autapomorphy after all. That said, the matrix is quite small, 
lacks any "parareptiles", and won't be used for the upcoming publication but 
replaced by a new one (D. Ford, pers. comm.). We're still living in interesting 
times.

A few passages in the new paper are of sufficiently general interest that I 
should address them here:

"Reisz et al. (2010) did not even cite Anderson and Reisz (2004)"

As the last "alius" of Reisz et al. (2010), I can explain what went on there. I 
did the phylogenetic analysis and originally added *Pyozia* Anderson & Reisz, 
2004, as well as several other varanopids that didn't make it into the paper; 
Robert Reisz told me to take them out again because Varanopidae was going to be 
revised.

"Sansom et al. (2010) demonstrated that data incompleteness causes stem-ward 
slippage"

I've seen other people make this misinterpretation as well. It is, of course, a 
misinterpretation. Stemward slippage results from missing data if and only if 
the older apomorphies are more easily preserved than the younger ones, as is 
the case in Cambrian chordates. It is not the case when all characters in a 
data matrix concern bones and teeth. In this conference talk 
https://urldefense.proofpoint.com/v2/url?u=https-3A__youtu.be_cptROtdSgNk&d=DwIFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Ry_mO4IFaUmGof_Yl9MyZgecRCKHn5g4z1CYJgFW9SI&m=xr3CzqGjGzyyfgXs7Hzc25tk1uPFiYHa46aYILtY1ao&s=YNGKAqvyeHNsCEaxxezZ5q7awUILJ6868s4jXW4G0S0&e=
 , Mark Purnell (the last author of Sansom et al. 2010) confirmed this.

"Since parsimony per se as well as the use of characters despite their 
dependent and uncountable nature is an act of abstraction, no substantial 
significance is expected from further steps."

That's why it's so important to make sure that all characters in a data matrix 
are independent from each other and therefore countable. Redundant characters 
distort trees.

"No outgroup was defined."
"Eight non-synapsid taxa were suggested to polarize the character history of 
varanopids: two basal Diadectomorpha (*Limnoscelis*, *Tseajaia*), three 
Eureptila (*Paleothyris*, *Protorothyris*, *Captorhinus*) and three early 
Parareptilia (*Milleropsis*, *Erpetonyx*, *Australothyris*). In order not to 
set anticipatory constraints, no defined outgroup was used."

That means PAUP* used the first terminal taxon in the list as the outgroup. 
Online Resources 4 and 5 reveal that's *Limnoscelis*. Yay, paraphyletic 
diadectomorphs... though fortunately that doesn't matter for the topology of 
the rest of the ingroup.

PAUP* makes unrooted trees (unless there are asymmetric characters in the 
matrix, which almost never happens) and then displays them as rooted on the 
outgroup. If no outgroup was defined a priori, the first OTU in the list is 
used. If an outgroup consisting of several OTUs was defined, but the ingroup is 
not monophyletic with respect to it in the unrooted trees, then PAUP* gives you 
a warning message saying it can't do what you want, and again displays the 
trees as rooted on the first OTU in the list.

"The most stable tree could be found when including *Paleothyris* and 
*Protorothyris* as the only non-synapsid OTU[s]."

That's not a reason to actually do that.

"Therefore, the original definition of âMycterosaurinaeâ without the 
eponymic Mycterosaurus is replaced by the new combination Mesenosaurinae."
"(unranked) Neovaranopsia comb. nov.
Subfamily Mesenosaurinae comb. nov.
(unranked) Afrothyra comb. nov."

If a species is referred to a genus it wasn't referred to before, its binomen 
becomes a new combination. A single word isn't a combination; these three names 
are simply new names.

The paper contains no indication of having been registered in ZooBank. Thus, 
the new potentially ICZN-governed names (Mesenosaurinae subfam. nov., 
Microvaranops parentis gen. et sp. nov.) will only become valid once the print 
version comes out.

And Mesenosaurinae most likely won't be a valid subfamily name even then, 
because it isn't explicitly given a type genus and isn't given a diagnosis. I 
can't look up the relevant ICZN Articles right now, though, because nhm.ac.uk 
has frozen.

Always read the supplementary information! Online Resource 3:

"Numerous errors occur in all cited datasets, including impossible codings of 
unknown skeletal elements (for example, hyoids of *Archaeovenator*, Reisz et 
al. 2010)."

Oh. Indeed, the hyoid apparatus of *Archaeovenator* is completely unknown 
according to its description (Reisz & Dilkes 2003). That same paper is the one 
that introduced the hyoid character into the matrix (the phylogeny of the 
matrices is admirably traced in Online Resources 1 and 2), and sure enough it's 
scored as unknown for *Archaeovenator*. I can't remember if I miscorrected the 
score, but if not, I still have to take responsibility for overlooking it.

"Multistate characters are avoided, after previous workers sometimes included 
several possible evolutionary steps within one character."

I don't understand what's supposed to be wrong with that...?

Online Resources 6â8, which can be opened in Photoshop CS3 but not in 
Illustrator CS3, are simply the figures already included in Online Resource 1.

Anderson JS, Reisz RR. 2004. *Pyozia mesenensis*, a new, small varanopid 
(Synapsida, Eupelycosauria) from Russia: âpelycosaurâ diversity in the 
Middle Permian. JVP 24:173â179.
Reisz RR, Dilkes DW. 2003. *Archaeovenator hamiltonensis*, a new varanopid 
(Synapsida: Eupelycosauria) from the Upper Carboniferous of Kansas. Can. J. 
Earth Sci. 40:667â678.
Reisz RR, Laurin M, MarjanoviÄ D. 2010. *Apsisaurus witteri* from the Lower 
Permian of Texas: yet another small varanopid synapsid, not a diapsid. JVP 
30:1628â1631.
Sansom RS, Gabbott SE, Purnell MA. 2010. Non-random decay of chordate 
characters causes bias in fossil interpretation. Nature 463:797â800.