Wachtlerosaurus: a thalattosaur, not an archosaur

Perner 2018
introduces Wachtlerosaurus ladinicus (Fig. 1, 2), a tiny disarticulated reptile from the Middle Triassic Dolomites of northern Italy. Perner considers the specimen an archosaur. In the large reptile tree (LRT, 1264 taxa) the specimen nests with thalattosaurs and Thalattosaurus in particular. The purported antorbital fenestra is the naris.

Figure 1. From Perner 2018, the original reconstruction of Wachtlerosaurus. Scale bar added.

Figure 1. From Perner 2018, the original reconstruction of Wachtlerosaurus. Scale bar added. Note the elongate ribs considered parts of the pelvis here. Pes of Euparkeria added by Perner. Manus appears to be from a coelophysoid theropod and flipped left to right.

The Dolomites are about 250 million years old
and are formed from coral reefs in the Tethys Sea, a perfect niche for a marine reptile like a thalattosaur.

Figure 1. Wachtlerosaurus in situ and reconstructed in lateral view.

Figure 2. Wachtlerosaurus in situ and reconstructed in lateral view.

A reconstruction of the skull helps
(Fig. 1) put the pieces of the broken skull back together again.

A few other new interpretations on the paper.

  1. Perner 2018 identifies two long, parallel dorsal ribs surrounding a jumble of ?vertebrae as parts of an oversized pelvis (Fig. 1).
  2. Perner employs the humerus in place of a scapula (Fig. 1).
  3. Probably this scattered bone specimen is incomplete, not nearly complete, as described.
Figure 3. Subset of the LRT nesting Wachtlerosaurus with Thalattosaurus.

Figure 3. Subset of the LRT nesting Wachtlerosaurus with Thalattosaurus.

References
Perner T 2018. A new interesting archosaur from the Ladinian (Middle Triassic) of the Dolomites (Northern Italy) Preliminary report. Pp 1–8 in Some new and exciting Triassic Archosauria from the Dolomites (Northern Italy). Perner T and Wachtler M eds. Dolomythos-Museum, Oregon Institute of Geological Research.

Advertisements

What would Vancleavea be, if thalattosaurs were not known?

This is lesson 5 in taxon exclusion…
to see where select clades would nest in the absence of their proximal taxa.

Figure 1. Vancleavea surrounded by purported sister taxa as figured by Nesbitt and Wikipedia. None of these taxa share more traits with Vancleavea than does Helveticosaurus, a taxon ignored since it was proposed here.

Figure 1. Vancleavea surrounded by purported sister taxa as figured by Nesbitt and Wikipedia. None of these taxa share more traits with Vancleavea than does Helveticosaurus (Fig. 2), a taxon ignored before and since it was proposed here.

Traditionally,
(Nesbitt et al. 2009) aquatic Vancleavea (Figs. 1, 2) was nested with Doswellia and other archosauriforms (Fig. 1), but only in the absence of a long list of taxa including thalattosaurs.

By contrast,
as we learned seven years earlier, in the large reptile tree (LRT, 1242 taxa) Vancleavea nests with Helveticosaurus, as a derived thalattosaur. Vancleavea is so different from archosauriforms, that I wondered where it would nest in the absence of thalattosaurs? Would it nest with archosauriforms?

Figure 2. Vancleavea with its sister, Helveticosaurus.

Figure 2. Vancleavea with its sister, Helveticosaurus. This connection might have been missed previously because the skull of Helveticosaurus is so badly smashed. These two share more traits in the LRT with each other than either does with any other taxon. Taxon exclusion prevented earlier workers from seeing this.

Having a wide gamut taxon list
permits us to ‘play’ with hypothetical scenarios like this.

If we restrict potential candidates to the Archosauromorpha
(in order to keep calculation time down) then Vancleavea nests with a wide variety of basal sauropterygians. The absence of thalattosaurs (except Vancleavea) creates this loss of resolution. However, the absence of Vancleavea, too, returns complete resolution to basal sauropterygians, as in the complete analysis.

If we restrict potential candidates to Arcosauromorpha sans Enaliosauria (ichthyosaurs + plesiosaurs)
then Vancleava nests with the first outgroup taxon to the Enaliosauria, Anarosaurus.

Figure 2. Anarosaurus nests at the base of the Enaliosauria (= Sauropterygia + Ichthyosauria)

Figure 3. Anarosaurus nests at the base of the Enaliosauria (= Sauropterygia + Ichthyosauria)

If we restrict potential candidates to the Lepidosauromorpha
(sans Archosauromorpha) then Vancleavea nests with Lanthanosuchus (Fig. 3).

The skull of Lanthanosuchus in several views and colorized.

Figure 4. The skull of Lanthanosuchus in several views and colorized.

In none  of these tests
does Vancleavea nest with any Archosauriformes (Fig. 1). Once again, a biased, cherry-picked and restricted taxon list missed several opportunities/possibilities that the LRT includes due to its wide gamut of tested taxa. In 2018 many workers are still including Vancleavea in archosauriform studies. Why? That ‘by default‘ nesting due to cherry-picked taxa needs to stop immediately. There are many better candidates out there for sister taxa. Sprinkle a few thalattosaurs in there and see where Vancleavea nests when given the opportunity.

Taxon exclusion
has been the number one problem in traditional paleontology. That’s why this blogpost exists. That’s why the LRT includes such a wide gamut of taxa. The result is a minimizing of taxon exclusion and the problems that attend it.

References
Nesbitt SJ, Stocker MR, Small BJ and Downs A 2009. The osteology and relationships of Vancleavea campi (Reptilia: Archosauriformes). Zoological Journal of the Linnean Society 157 (4): 814–864. doi:10.1111/j.1096-3642.2009.00530.x.
Parker WG and Barton B 2008. New information on the Upper Triassic archosauriform Vancleavea campi based on new material from the Chinle Formation of Arizona. Palaeontologia Electronica 11 (3): 20p.

wiki/Vancleavea

 

 

Juvenile articulated Eusaurosphargis discovered

Scheyer et al. 2017
bring us a new largely articulated juvenile Eusaurosphargis specimen (PIMUZ A/III 4380; Figs. 1–3) very similar to the adult disarticulated specimen described by Nosotti and Rieppel 2003 (BES SC 390; Middle Triassic, ~240 mya, ~20 cm snout to vent length). Scheyer et al. had trouble nesting Eusaurosphargis correctly as a derived thalattosaur largely due to taxon exclusion (see below).

Figure 1. Adult and juvenile Eusaurosphargis specimens to scale.

Figure 1. Adult and juvenile Eusaurosphargis specimens to scale. The adult was disarticulated.

The old ‘adult’ specimen
was considered more closely related to Helveticosaurus (Fig. 4) than to placodonts. Here both Eusaurosphargis and Helveticosaurus nest within the Thalattosauriformes close to armored Vancleavea. Here Eusaurosphargis does not nest close to turtle-like Sinosaurophargis. The adult skeleton is completely disarticulated. That makes reconstruction particularly difficult. Thus the order of the traced vertebrae in dorsal view (Fig. 1) is largely guesswork. Likewise, the skull included some guesswork helped by phylogenetic bracketing.

Figure 2. The in situ juvenile specimen of Eusaurosphargis, the original tracing and DGS tracing of dorsal ribs (blue) and sternal ribs (green).

Figure 2. The in situ juvenile specimen of Eusaurosphargis, the original tracing and DGS tracing of dorsal vertebrae and elongated transverse processes (blue) and dorsal ribs (green). The specimen was exposed from below, but preserved right side up, hence the slight disarticulation of dorsal elements and the skull in marine sediments. CT scans indicate the buried sacral ribs were longer than traced here. 

The new ‘juvenile’ specimen
has a disarticulated skull, but most of the elements appear to be present, though some were originally unidentified and the squamosal, now a jugal, was misidentified.

Figure 3. Eusaurosphargis juvenile skull, pectoral and pelvic girdles reconstructed.

Figure 3. Eusaurosphargis juvenile skull, pectoral and pelvic girdles reconstructed. GIF animation second frame shows two views of the in situ skull. The juvenile includes articulated extremities. Boxed elements are the purported squamosals, here identified as jugals. Scheyert et al. did not attempt a skull reconstruction.

Scheyer et al. report
the armor and other elements “support an essentially terrestrial lifestyle for Eusaurosphargis and and that within the marine reptile ‘superclade’ E. dalsassoi potentially is the sister taxon of Sauropterygia.” Neither are supported by the large reptile tree (LRT 1027 taxa), which resurrected the clade, Enaliosauria for Scheyer’s ‘superclade.’

The elongated dorsal transverse processes
and osteoderms are convergent with those in placodonts and sinosaurosphargids.

The jugals are much larger than the squamosals
as in Helveticosaurus and Vancleavea. For reasons unknown, Scheyer et al. erroneously compared these elements with those of the more distantly related Askeptosaurus, which ALSO has tiny squamosals, like most, if not all, thalattosaurs.

Phylogenetic analysis
The Scheyer et al. inclusion set excludes so many pertinent taxa that it nests turtles with archosaurs and lepidosaurs. It also nests Eusaurosphargis close to placodonts. Correctly it nests Eusaurosphargis close to Helveticosaurus and Thalattosauriformes. Vancleavea was not included. It is clear that Scheyer et al. have no idea how the major taxa are actually arranged as documented in the LRT for the last seven years. They also employed suprageneric taxa. There’s no reason for such unprofessional  guessing to continue in professional studies.

Figure 4. Helveticosaurus had cheek teeth that look like baleen strainers and long fangs anteriorly. It was also much larger than Eusaurosphargis but was coeval. Vancleavea is shown to scale and to the same length.

Figure 4. Helveticosaurus had cheek teeth that look like baleen strainers and long fangs anteriorly. It was also much larger than Eusaurosphargis but was coeval. Vancleavea is shown to scale and to the same length.

‘Homologies’ reported by Scheyer et al.:
“PIMUZ A/III 4380 shares with Palatodonta bleekeri (and placodonts such as Paraplacodus broilii and Placodus gigas) the deep skull shape and wide snout with large external nares, as well as the double tooth row in the upper jaw (on the maxillae and palatines) and a single row in the lower jaw.” These traits are likewise found by homology in Helveticosaurus and Vancleavea where known. Scheyer et al. feel the freedom to make these comparisons to placodonts because their incorrect (based on massive taxon exclusion) cladogram nests Eusaurosphargis close to placodonts. This is the authority of the LRT and its large gamut, specimen-based taxon list at work. When Scheyer et al. have a comparable taxon list, then we can discuss differences in scoring, if they arise.

Terrestrial?
Scheyer et al. report, “Given the large number of pachypleurosaurs of similar size range, among a plethora of thousands of other fossils, we corroborate the previous idea that E. dalsassoi had a terrestrial habitat preference.” This makes no sense. In the LRT pachypleurosaurs arise from marine taxa and give rise to marine taxa. Thus, based on phylogenetic bracketing. pachypleurosaurs are marine (or at least aquatic), too,

Scheyer et al. report, “the short and proximally dorso-ventrally wide tail would be similarly inefficient in providing propulsion.” You don’t have to get around fast in order to be aquatic. The flattened turtle-like appearance of several saurosphargids and placodonts have similar short-comings. And look at Helveticosaurus, the acknowledged sister (Fig. 4).

Scheyer et al. report,The stylopodial elements (humerus and femur) are tubular, moderately thin-walled bones with large marrow cavities” typical of terrestrial, not marine diapsids. They do not report similar tests on the universally accepted sister taxon, Helveticosaurus (and Vancleavea), but the proximal limb elements look similar from the outside (Fig. 4).

The All-Aquatic Superclade of Chen et al. 2014
does not include mosasaurs, but does include a few representatives of most other marine clades (but not nearly the number of taxa as in the LRT). While this is confirmation of the results first reported here in 2011, the topology of the Chen et al. cladogram has serious problems all based on taxon exclusion. Such problems are minimized in the LRT based on its large gamut where there is no need to ‘delete problematic characters’ in order to achieve a result that makes sense. 

A larger gamut phylogenetic analysis
nests Helveticosaurus, Vancleavea and Eusaurosphargis within the Thalattosauriformes, despite over 1020 opportunities to nest elsewhere. Their disparate morphologies hint at further transitional and unusual morphologies to come.

Congeneric? Yes. Conspecific? No.
The smaller Eusaurosphargis nests with the larger one in the LRT. So they could be congeneric. However, comparing the reconstructions of the two shows several differences in the skull bones that preclude the two from being conspecific. Such juvenile/adult conspecific relationships in fossils found years apart and miles apart are, by their very nature, very rare, but they do occur.

References
Chen X-H, Motani R, Cheng L, Jiang D-Y and Rieppel 2014. The Enigmatic Marine Reptile Nanchangosaurus from the Lower Triassic of Hubei, China and the Phylogenetic Affinities of Hupehsuchia. PlosOne online.
Nosotti S and Rieppel O 2003. Eusaurosphargis dalsassoi n.gen. n.sp., a new, unusual diapsid reptile from the Middle Triassic of Besano (Lombardy, N Italy). Memories of the Italian Society of Natural Science and the Museum of Natural History in Milan, XXXI (II).
Scheyer T et al. (5 other authors) 2017. A new, exceptionally preserved juvenile specimen of Eusaurosphargis dalsassoi (Diapsida) and implications for Mesozoic marine diapsid phylogeny. Nature.com/scientific reports online.

wiki/Eusaurosphargis

SVP 24 – Last abstract and a note by Holtz on the DML

Li et al. 2015
present a new thalattosaur specimen, complete and preserved in ventral view.

Figure 1. New Xinpusaurus described by Li et al. 2015.

Figure 1. New Xinpusaurus described by Li et al. 2015.

From the abstract
During the excavation in the Wusha District, Xingyi City, Guizhou Province from 2010 to 2013, a new thalattosaur specimen was excavated from fossil bed 53 of the Upper Assemblage of the Xingyi Fauna, where abundant marine reptiles, fishes and invertebrates are found. This new specimen represents a new species of the thalattosauroid Xinpusaurus. The skeleton is nearly complete and articulated, with only some caudal vertebrae missing. Its total length is about 2.1 m. The skull is preserved in ventral view, with a length of more than 40 cm. This specimen is ascribed to the genus Xinpusaurus based on the dorsally curved anterior end of maxilla, and a proximal end of the humerus that is wider than the distal end. It differs from the type species Xinpusaurus suni by several features. The posterior process of its jugal is undeveloped, while other thalattosaurs all possess triradiate jugals with an elongate posterior process. The coracoid is oval, different than the arch-shaped coracoid of X. suni. The radius is relatively short, only about half the length of the humerus, whereas in all specimens of X. suni, the length ratio of the radius to humerus is more than 0.63. The constriction of the femur at mid-shaft is unremarkable, which is quite unique in Thalattosauria. In addition, the posterior part of the iliac blade is rectangular without a pointed tip, an element that has not yet been described for X. suni. 

I think the authors did a good job here.

On a side note…
Dr. Thomas Holtz (Dept. of Geology, University of Maryland)

reported this bit of wisdom on the Dinosaur Mailing List regarding a previously published specimen that was later redescribed as a chimaera (mix of several specimens).

“It happens sometimes. Your search image is off, and you misconstrue one element for another similar-shaped one. For instance, a bone identified originally as the squamosal of Zuniceratops turned out to be the ischium of Nothronychus. Science works by reciprocal illumination: we correct our misjudgements and build on that new information. As you can see in the new PeerJ paper, those turtle parts DO look fairly furcula-like at first glance.”

Mistakes of omission and commission DO happen to all paleontologists.
whether prior to or during publication. Many of those are reviewed and repaired here. On the flip side, I’m still finding mistakes and creating updates in my own work. It’s all part of the process… and we should all keep that in mind.

What should NOT be part of the process
is the practice of holding up past mistakes as a reason to reject a new manuscript. Try to start every day fresh. Try to sprinkle in a little praise before and after every criticism. And, kids, always beware of black-washers and data deniers. They don’t practice what Dr. Holtz preaches.

References
Li et al. 2015. A new species of Xinpusaurus (Reptilia: Thalattosauria) from the Middle Triassic of Southwestern China. Journal of Vertebrate Paletontology Abstracts

SVP23 – A new thalattosaur from Oregon

Metz et al. 2015
reports several new associated and likely conspecific bits and pieces of a new thalattosaur from the Middle Triassic of  Oregon.

From the abstract:
“Thalattosauria is a clade of secondarily aquatic Triassic reptiles generally found as
isolated skeletons from localities in Europe, China, and North America. In North America, four genera of thalattosaurs have been described from two formations: the Hosselkus Limestone of California (Carnian) and the Sulphur Mountain Formation of British Columbia (Lower to Middle Triassic). In 2011, a new thalattosaur locality was discovered from the Late Triassic (Carnian) Brisbois Member of the Vester Formation in central Oregon. This formation records deposition in a nearshore environment in the forearc region of the Izee Terrane. The material consists of one large block of highly concentrated, three-dimensionally preserved, disarticulated skeletons of five or more individuals. Preparation to date has revealed partial crania, including complete braincases, as well as numerous axial and appendicular elements from individuals of varying size.

Preliminary examination of the most complete skull, University of Oregon Museum of Natural and Cultural History F64236, reveals a unique combination of characters including a strong degree of rostral ventral deflection, frontals that are excluded from the orbital margin, nasals separated by a long anterior projection of the frontal, a small degree of contribution by the jugal to the ventral margin of the orbit, the absence of an upper temporal fenestra and diastema, and a peg-like homodont dentition. Based on the morphological similarity of several multiple-sized elements, we interpret that this material represents the first ontogenetic series of any known thalattosaur. The new Oregon material represents the largest thalattosaur species yet found in North America, and is both the oldest vertebrate remains and the first occurrence of a thalattosaur from Oregon. Finally, the quality of preservation, particularly of poorly-known cranial elements, contributes important new morphological data about thalattosaurs for resolving phylogenetic relationships within the clade as a whole.”

Photos of Eric Metz
and his SVP poster have appeared online here and here. From the images and description F64236 (Fig. 1) is a sister to Clarazia (Peyer 1936, Rieppel 1987, Figs. 2, 3) with a more deeply downturned premaxilla.

Figure 1. Oregon thalattosaur skull parts.

Figure 1. Oregon thalattosaur skull parts. Middle Triassic 238-228 mya. Gray elements are restored.

According to Metz
“The specimen had a downturned snout, which it likely used to break apart reefs made of mollusks and sponges. Adults would have measured 3 meters in length.” Like similar thalattosaurs, rather large vomer teeth were present.

Figure 2. The Thalattosauria and outgroups (Wumengosaurus and Stereosternum) to scale.

Figure 2. The Thalattosauria and outgroups (Wumengosaurus and Stereosternum) to scale. Note the heretical and verified inclusion of Vancleavea here. The new specimen would be as large as the largest of these.

Clarazia was a much smaller thalattosaur and more primitive with a straighter snout.

Figure 4. Clarazia, a thalattosaur sister to the new Oregon specimens.

Figure 3. Clarazia, a thalattosaur sister to the new Oregon specimens.

 

References
Metz E, Druckenmiler PS, and Carr G 2015. A new thalattosaur from the Vester Formation (Carnian) of central Oregon. Journal of Vertebrate Paleontology abstracts 2015.
Peyer B 1936. Die Triasfauna der Tessiner Kalkalpen. X.  Clarazia schinzi nov. gen. nov. spec. Abhandlungen der Schweizerischen Pala¨ontologischen Gesellschaft, 57, 1–61.
Rieppel O 1987. Clarazia and Hescheleria; a reinvestigation of two problematic reptiles from the Middle Triassic of Monte San Giorgio (Switzerland). Palaeontographica, A, 195, 101–129.

wiki/Thalattosaur

Yahoo news

Back to Vancleavea

Several years ago we looked at Vancleavea campi (Figs. 1,2 ), a Triassic aquatic reptile described by Nesbitt, et al. 2008 as an archosauriform nesting with Erythrosuchus, Euparkeria, Turfanosuchus and Doswellia (according to Wikipedia, based on published work listed above). Unfortunately, Vancleavea shares few traits with these archosauriforms. It has no antorbital fenestra, no upper temporal fenestra and no mandibular fenestra.

Figure 1. Vancleavea surrounded by purported sister taxa as figured by Nesbitt and Wikipedia. None of these taxa share more traits with Vancleavea than does Helveticosaurus, a taxon ignored since it was proposed here.

Figure 1. Vancleavea surrounded by purported sister taxa as figured by Nesbitt and Wikipedia. None of these taxa share more traits with Vancleavea than does Helveticosaurus, a taxon ignored since it was proposed here.

Not yet tested in academic publications,
the thalattosaur, Helveticosaurus, shares more traits with Vancleavea than 569 other tested taxa in the large reptile tree.

Figure 2. Vancleavea with its sister, Helveticosaurus.

Figure 2. Vancleavea with its sister, Helveticosaurus.

This counter argument
was made more than 4 years ago. To date no one else has supported or refuted the argument. Nevertheless, in the last four years Vancleavea has appeared in several cladograms without Helveticosaurus. Unfortunately this demonstrates that paleontologists are really not interested in its correct nesting node, but would rather just add new taxa to existing flawed analyses and cladograms. Testing prior work is not their strong suite. Discovery is.

Vancleavea campi  (Nesbitt et al. 2009) Late Triassic,~210 mya, ~1.2 meters in length, was originally considered a very weird archosauriform close to DoswelliaTurfanosuchusChanaresuchus and Erythrosuchus, but that’s because the authors did not compare it to Helveticosaurus with which Vanclevea shares more traits. It turns out that Vanclevea was a not-so-weird thalattosaur and a prime example of what happens when the gamut of the inclusion set is decided prior to the analysis. Vancleavea was the last in its lineage. Unlike other thalattosauriforms, Vancleavea was armored with a variety of ossified scales covering the body.

There must be dozens
of Vancleavea-like thalattosaurs yet to be discovered, judging by the variation present between it and Helveticosaurus. Even so, after adding hundreds of taxa to the large reptile tree, these two still nest together.

Not the only time a taxon’s correct nesting ignored.
These taxa are also traditionally incorrectly nested based on the results of the large reptile tree.

  1. Turtles
  2. Pterosaurs
  3. Fenestrasauria and Tritosauria
  4. Snakes
  5. Caseasauria
  6. Mesosauria
  7. Poposauria
  8. Rhynchosauria
  9. Synapsida
  10. Chilesaurus and Daemonosaurus
  11. Gephyrostegus
  12. Procolophon
  13. Cartorhynchus
  14. Youngina and Youngoides
  15. Xianglong
  16. Tetraceratops
  17. Eudibamus
  18. Doswellia
  19. Revuletosaurus
  20. Scleromochlus
  21. Pseudhesperosuchus
  22. Marasuchus
  23. Lagerpeton
  24. Teraterpeton
  25. Ticinosuchus
  26. and whatever else I’ve forgotten to list here

References
Nesbitt SJ, Stocker MR, Small BJ and Downs A 2009. The osteology and relationships of Vancleavea campi (Reptilia: Archosauriformes). Zoological Journal of the Linnean Society 157 (4): 814–864. doi:10.1111/j.1096-3642.2009.00530.x.
Parker WG and Barton B 2008. New information on the Upper Triassic archosauriform Vancleavea campi based on new material from the Chinle Formation of Arizona. Palaeontologia Electronica 11 (3): 20p.

wiki/Vancleavea

Update on the Rossman 2002 “Brazilosaurus”

Figure 1. The nesting of Rossman's "Brazilosaurus" (the PIMUZ AIII 0192 specimen) at the base of the Thalattosauria is confirmed with the addition of post-cranial and new cranial data.

Figure 1. Click to enlarge. The nesting of Rossman’s “Brazilosaurus” (the PIMUZ AIII 0192 specimen) at the base of the Thalattosauria is confirmed with the addition of post-cranial and new cranial data.

Today: Some new data and a new reconstruction of the skull and postcrania of the intriguing and potentially important postcrania of PIMUZ A/III 0192. This mesosaur-ish reptile was attributed by Rossman 2002 to Brazilosaurus sanpauloensis. 

Earlier the only data I had on this specimen was a line drawing of a skull from Rossman 2002. This data resulted in a nesting outside of Brazilosaurus + Stererosternum + Mesosaurus, at the base of the Thalattosauria. Notably, with the additional data, the nesting did not change.

Crushed skulls are often the best Because all the parts are crushed into a single plane and you can rebuild that “house of cards” or “crushed eggshell” in many views. Some parts are visible through the orbit. The occiput often flips to the side.

Mesosaurs are key
Workers have attempted to nest mesosaurs based on its lack of temporal fenestration — and they (Modesto 2006) end up with pareiasaurs and millerettids and procolophonoids. But with it’s hyper-long teeth, Mesosaurus is clearly a derived form. What we’re looking for is a basal taxon that looks like Mesosaurus with small plesiomorphic teeth. Then perhaps we’ll see more evidence for the diapsid skull morphology. And that’s exactly what we find.

Figure 2. Click to enlarge. The Rossman "Brazilosaurus" PIMUZ AIII 0192. This is a basal thalattosaur and a derived mesosaur. Due to severe crushing elements from the other side of the skull made it appear that the skull lacked fenestration following the generally accepted but mistaken hypothesis that all mesosaurs lacked temporal fenestra.

Figure 2. Click to enlarge. The Rossman “Brazilosaurus” PIMUZ AIII 0192. This turns out to be THE  basal thalattosaur as well as a basal mesosaur. Due to severe crushing elements from the other side of the skull made it appear that the skull lacked fenestration following the generally accepted but mistaken hypothesis that all mesosaurs lacked temporal fenestra. Note the tiny forelimbs and deep tail chevrons. Low dorsal spines and gracile ribs are also noteworthy.

Here (Fig. 1) the Rossman (2002) “Brazilosaurus” nests at the base of the Thalattosauria, whether using the Rossman data as is, or pulling slightly different traits out (Fig. 2).

Rossman (2002) presented photos of several interesting mesosaurs. Among them was the SMF-R-4710 specimen attributed to Stereosternum (Fig. 3). This one does nest with Stereosternum, but with more open temporal fenestrae, its nests at the base of the mesosaurs, close to the base of the ichthyosaurs and thalattosaurs.

Figure 3. The SMF-R-4710 specimen attributed to Stereosternum, but with larger temporal fenestrae, smaller teeth and other distinct traits.

Figure 3. Click to enlarge. The SMF-R-4710 specimen attributed to Stereosternum, but with larger temporal fenestrae, smaller teeth and other distinct traits. The high unfused dorsal processes are also seen on Hupehsuchus and Utatsusaurus.

If you think this Stereosternum specimen is starting to look a lot like Wumengosaurus, you’d be right. And Wumengosaurus now nests at the base of Hupesuchus + Ichthyosaurs, so we’re getting closer and closer to the common origin of both. One keeps its teeth, the other does not.

Figure 3. Stereosternum SMF-R-4710 reconstructed from traced image (in color below). Here temporal fenestration is clearly diapsid.

Figure 3. Stereosternum SMF-R-4710 reconstructed from traced image (in color below). Here temporal fenestration is clearly diapsid.

Lest you doubt
Running the large reptile tree with all temporal fenestration traits deleted recovers a single tree unchanged from the full character list tree. If anyone wants to come up with better data or a better tracing, please use specimen numbers.

There are so many specimens of mesosaurs
and so many variations and only three names for them (Mesosaurus, Brazilosaurus, Stereosternum). Someone needs to put it all together and catalog some of the important specimens (e.g. the holotypes). We’ll need new generic names for the specimens above.

If someone has a good photo of the Brazilosaurus holotype (Shikama and Ozaki 1966), please let me know.

References
Cope ED 1886. A contribution to the vertebrate paleontology of Brazil. Stereosternum tumidum, gen. et sp. nov. Proceedings of the American Philosophical Society 23(121):1-21.
Modesto S 2006.
 The cranial skeleton of the Early Permian aquatic reptile Mesosaurus tenuidens: implications for relationships and palaeobiology. Zoological Journal of the Linnean Society, 2006, 146, 345–368.
Shikama T and Ozaki T 1966. On a Reptilian Skeleton from the Palaeozoic Formation of San Paulo, Brazil” Transactions and Proceedings of the Palaeontological Society of Japan, New Series 64: 351–358.