New phylogeny of tanystropheids fails to include many tanystropheids and dozens of pertinent outgroup taxa

Summary for those in a hurry:
Three PhDs cherry-pick taxa, “Pull a Larry Martin” by cherry-picking a convergent neck trait, omit dozens of taxa that separate protorosaurs (pre-archosaurs) from tanystropheids (lepidosaurs), and omit several small tanystropheids that learned how to flap, then fly because the authors thought the details too confusing to deal with.

They took the blue pill to maintain their status and ‘make nice’ in academic circles

Spiekman SNF, Fraser NC and Scheyer TM 2021
propose a “new phylogenetic hypothesis of Tanystropheidae”. It’s not new. So, others call this headline-grabbing.

From the abstract:
“The historical clade “Protorosauria” represents an important group of
archosauromorph reptiles that had a wide geographic distribution between the Late
Permian and Late Triassic.”

Simply adding taxa, as in the large reptile tree (LRT, 1839+ taxa) separates members of the Protorosauria that do indeed nest within the new Archosauromorpha (e.g. Protorosaurus, Prolacerta, Ozmik) from convergent tanystropheids that nest within the new Lepidosauromorpha. This was presented in Peters 2007, uncited in the Spiekman et al. paper.

“Protorosaurs” are characterized by their long necks, which are epitomized in the genus Tanystropheus and in Dinocephalosaurus orientalis. “

Three PhDs are here caught “Pulling a Larry Martin”. Dr. Martin would have had a good chuckle at this short-sighted blunder. As we learned many times earlier, never-ever cherry-pick a trait like this. You don’t know if it is convergent (as it is in this case) with unrelated taxa without the benefit of a phylogenetic analysis.

“Recent phylogenetic analyses have indicated that “Protorosauria” is a polyphyletic clade, but the exact relationships of the various “protorosaur” taxa within the archosauromorph lineage is currently uncertain.”

That’s a clue that their taxon list is incomplete.

“Several taxa, although represented by relatively complete material, have previously not been assessed phylogenetically.”

If so… good! However, not only did these three authors add unrelated taxa, they omitted ingroup taxa. So ignore this Spiekman et al. paper and wait for a valid phylogenetic analysis that includes all ingroup taxa and no outgroup taxa.

“We present a new phylogenetic hypothesis that comprises a wide range of archosauromorphs, including the most exhaustive sample of “protorosaurs” to date and several “protorosaur” taxa from the eastern Tethys margin that have not been included in any previous analysis.”

This is false. The LRT represents the most exhaustive sample of protorosaurs and tanystropheids to date. The authors do not realize that tanystropheids are lepidosaurs when appropriate taxa are added. Both analyses, you should note, were published online.

“The polyphyly of “Protorosauria” is confirmed and therefore we suggest the usage of this term should be abandoned.”

By contrast, in the LRT the Protorosauria is a monophyletic clade (Fig. 1). Tanystropheidae is also monophyletic, but nests elsewhere (Fig. 2). These PhDs did not do enough work to see that their cherry-picked taxa had convergent traits. Let a wide-gamut family tree, like the LRT, choose your taxon list. Do not cherry-pick taxa that ‘look like they belong. Do not let textbooks tell you cretain taxa belong together. Find out for yourself. Don’t be lazy.

Figure 1. Two unrelated clades from the LRT. Compare to figure 2 from the Spiekman et al. paper.

From the abstract (continued):
“Tanystropheidae is recovered as a monophyletic group and the Chinese taxa Dinocephalosaurus orientalis and Pectodens zhenyuensis form a new archosauromorph clade, Dinocephalosauridae, which is closely related to Tanystropheidae.”

Figure 2. Cladogram from Spiekman et al. 2021(above). Colors added to show how many clades were mixed together during the cherry-picking process.

From the abstract (continued):
“The well-known crocopod and former “protorosaur” Prolacerta broomi is considerably less closely related to Archosauriformes than was previously considered.”

In the LRT, Prolacerta is a basal protorosaur. Just add taxa. Don’t cherry-pick to your own biases. And that concludes the abstract.

these authors omitted so many pertinent taxa that they nested tanystropheids within Archosauromorpha, rather than Lepidosauromorpha (Peters 2007). They also omitted the tanystropheids, Cosesaurus, Sharovipteryx, Longisquama and the clade Pterosauria (Fig. 3). Their motivation: keep the academic status quo.

a jumbled roadkill specimen of Tanystropheus was first named Tribelesodon because the long neck bones were originally considered wing phalanges and the pedal morphology and trident teeth were nearly identical to those of basal pterosaurs. That’s a clue!

At least Spiekman et al.
cited Peters 2000 and 2005. Other peer-reviewed academic publications on this subject by Peters were omitted (see below). What Spiekman et al. said about the small bipedal taxa featured in those pubs (Fig. 3) sheds light on this subject (see below).

Click to enlarge. Squamates, tritosaurs and fenestrasaurs in the phylogenetic lineage preceding the origin of the Pterosauria.
Figure 3. Squamates, tritosaurs and fenestrasaurs in the phylogenetic lineage preceding the origin of the Pterosauria.

From Spiekman et al. 2021:
“Cosesaurus aviceps is known from a single specimen, which represents an impression of a complete skeleton. As such, the outline of the specimen is well-preserved, but the detailed morphology of the taxon is very poorly known.”

This is false. I encourage readers to click this Cosesaurus page link to see the wealth of detail captured in this holotype fossil. Better to ‘show’ than ‘tell’.

“Due to the lack of morphological information, the phylogenetic affinities of Cosesaurus aviceps are unclear.”

This is an excuse not to visit the fossil or to cite online data.

“Previous analyses by Peters (2000), which has been widely criticized (e.g., Hone & Benton, 2007), it was concluded that Pterosauria are a derived lineage within “Prolacertiformes”.

Spiekman et al. apparently never read Hone & Benton 2007, 2009, in which Hone and Benton mis-scored data and upon learning that Cosesaurus was igoing to appear as a pterosaur ancestor in their supertree (as opposed to professor Benton’s personal favorite, the crocodilomorph with tiny fingers, Scleromochlus, Hone and Benton 2009 dropped all reference to Peters 2000 and gave credit to Bennett for the Peters 2000 hypothesis. Not sure why others haven’t jumped on this scandal (see below for link) other than to preserve the academic status quo and keep the origin of pterosaurs ‘an enigma’.

“This was largely based on several morphological characters observed in Cosesaurus aviceps, as well as the poorly known, gracile reptiles Sharovipteryx mirabilis and Longisquama insignis (Peters, 2000).”

This is false. Those two taxa are so well preserved that soft tissue is readily apparent, as everyone knows. Link to Sharovipteryx here. Link to Longisquama here. Link to a manuscript on these taxa here.

“Although Cosesaurus aviceps might represent a “protorosaur”, the lack of morphological information does not allow this taxon to be reliably incorporated in phylogenetic analyses, and recent phylogenetic investigations into archosauromorph or “protorosaurian” affinities did not consider this taxon.”

This is how academics avoid a subject. No one wants to do the work or be accused of helping an amateur who somehow got this ‘discovery’ published.

“Peters (2000) used the matrices of Evans (1988), Jalil (1997), and Bennett (1996) and reran each of them after adding a number of characters and rescoring some characters for certain taxa, for a total taxon sample that included 11 “protorosaurs”, other non-archosauriform archosauromorphs, the pterosaur Eudimorphodon, and two enigmatic and possibly gliding diapsids, Longisquama insignis (Sharov, 1970) and Sharovipteryx mirabilis (Cowen, 1981; Sharov, 1971).”

This is true. This hopefully shows that Peters 2000 followed the scientific method by personally examining the specimens, employing previously published cladograms and simply adding taxa. But keep reading. That’s not going to be good enough.

“Sharovipteryx is an enigmatic gliding reptile with a membrane stretched between the hindlimbs, which represents an entirely unique morphology among gliding reptiles.”

Gliding is a traditional assumption. Probably a mistake. What Spiekman et al. fail to note is Sharovipteryx was an obligate biped with a longer tibia than femur, indicating cursorial abilities. Again, don’t you think these three PhDs should have examined the specimens firsthand, instead of parroting traditional literature?

“Its phylogenetic position is highly uncertain due to its highly specialized, yet very poorly
known morphology.

This is false, as noted above. Even though it was tested three times by Peters 2000, Sharovipteryx still does not enter the Spiekman et al. cladogram. Trying to avoid a trip to Moscow, the authors considered ‘these grapes are probably sour.”

“Peters (2000) found “protorosaurs”, and Longisquama and Sharovipteryx, to be very closely associated with Eudimorphodon, from which a “protorosaurian” ancestry for pterosaurs was concluded. However, the exact topologies varied strongly between the different analyses, and this hypothesis of pterosaur ancestry has widely been rejected by other phylogenetic studies on pterosaurs and early archosaurs (e.g., Ezcurra, 2016; Ezcurra et al., 2020; Hone & Benton, 2007; Nesbitt, 2011; Padian, 1997).”

Wait a minute! Padian 1997 can’t reject Peters 2000! The other authors listed in the paragraph above also cited Hone and Benton 2007, 2009 (it was a two-part study) and that’s why they reject Peters 2000. Not because they looked at the specimens and added them to their analyses. They didn’t want to do the decade of work shown in the LRT and visit the specimens firsthand.

“The datasets of Benton & Allen (1997), Dilkes (1998), and Jalil (1997) were combined into one larger character list of 239 characters by Rieppel, Fraser & Nosotti (2003), which was used specifically to address “protorosaur” phylogeny, and in particular the question of “protorosaur” monophyly, which had now been put in doubt (Dilkes, 1998). This approach included seven “protorosaur” taxa (Protorosaurus, Drepanosaurus, Megalancosaurus, Prolacerta, Macrocnemus, Langobardisaurus, and Tanystropheus longobardicus), and four outgroup taxa (Petrolacosaurus, Youngina, Rhynchosaurus, and Trilophosaurus).”

Readers… now do you recognize when PhDs cherry-pick taxa to their own biases? And note: Rieppel, Fraser and Nosotti (2003) omitted adding the taxa published in Peters 2000? Why not just add them to test them? (Note: co-author Fraser is also a co-author of the present work under discussion, Spiekman et al. 2021.)

“Additional analyses were performed after subsequently including Euparkeria and Proterosuchus, and the lesser known “protorosaurs” Boreopricea and Jesairosaurus. Although the first analysis found a monophyletic “Protorosauria”, the other two resulted in paraphyly for the group. Although Rieppel, Fraser & Nosotti (2003) concluded that the monophyly of “Protorosauria” as previously regarded (e.g., Benton & Allen, 1997; Jalil, 1997) could not be maintained, they argued the need for an extensive phylogenetic investigation into “protorosaurs”.

That has been provided for all to see, test, confirm or refute in the LRT.

“Senter (2004) investigated the phylogenetic position of drepanosaurids in an analysis that comprised “protorosaurs” (Prolacerta, Macrocnemus, and Langobardisaurus), Longisquama, non-archosaurian Archosauriformes, birds, a non-avian dinosaur, and a number of early diapsids. This study found drepanosaurids to form a clade with Longisquama and Coelurosauravus, which was termed “Avicephala”, as the sister group to Neodiapsida, which in his analysis encompassed Youngina, the rhynchocephalian Gephyrosaurus, and several archosauromorphs. The included “protorosaurs” formed a monophyletic clade within Archosauromorpha.”

To be fair, Senter 2004 was publishing his PhD dissertation in 2004. It suffered from cursory examination of the specimens (or photos of the specimens) in which Senter could only pick out a few elements and many of those were misidentified.

“However, an analysis using the same character list by Renesto & Binelli (2006) could not reproduce the same topology.”

To be fair, you need the same taxon list to reproduce the same topology. Cladograms lump and separate taxa, not characters.

“Renesto et al. (2010) reaffirmed the position of drepanosaurids among “protorosaurs”, whereas Pritchard & Nesbitt (2017) recovered Drepanosauromorpha as a separate clade of non-saurian diapsids.”

These workers did not employ enough taxa to know that the diapsid skull morphology is convergent, with one clade appearing in the Archosauromorpha and another in the Lepidosauromorpha. Spiekman et al. are also in the dark with regard to diapsid monophyly.

“Müller (2004) included four different “protorosaur” taxa in his broad-scale analysis of diapsid relationships, which consisted of 184 characters compiled mainly from Rieppel, Mazin & Tchernov (1999) and DeBraga & Rieppel (1997). This study also inferred a polyphyletic “Protorosauria”, with Tanystropheus, Macrocnemus, and Prolacerta being successive sister taxa to rhynchosaurs and Trilophosaurus, whereas drepanosaurids were only quite distantly related to these taxa.”

Taxon exclusion is once again the problem here.

In summary,
don’t cherry-pick taxa, like these PhDs did. Let your wide-gamut cladogram choose your clades for you. Don’t trust anyone, even me. Test everything for yourself. I hope everyone was able to pick up a few clues as to how to avoid ‘believing’ certain cherry-picked citations. See how they accumulate? If you know Spiekman, Fraser or Scheyer, send them this link. They should know that what they produced was unprofessional and misleading.

Peters D 2000. A reexamination of four Prolacertiformes with implications for pterosaur
phylogenesis. Rivista Italiana di Paleontologia e Stratigrafia 106:293–336
DOI 10.13130/2039-4942/6148.
Peters D 2005. Suction feeding in a Triassic protorosaur? Science 308(5725):1112c–1113c
DOI 10.1126/science.308.5725.1112c.
Peters D 2007. The origin and radiation of the Pterosauria. In D. Hone ed. Flugsaurier. The Wellnhofer pterosaur meeting, 2007, Munich, Germany. p. 27.
Spiekman SNF, Fraser NC and Scheyer TM 2021. A new phylogenetic hypothesis of Tanystropheidae (Diapsida, Archosauromorpha) and other “protorosaurs”, and its implications for the early evolution of stem archosaurs. PeerJ 9:e11143 DOI 10.7717/peerj.11143

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.