New Champsosaurus paper perpetuates old myths

Whenever taxon exclusion mistakes are made and reviewed here,
I try to write to the lead author of the paper. Below is a recent email directed to Professor Dudgeon et al. 2020 on their recent review of the well-preserved skull of Champsosaurus (Figs. 1, 3), which they re-examined using computed tomography analysis.

Figure 1. Champsosaurus from Dugeon et al. Here the nasal is the ascending process of the premaxilla. The prefrontal is the nasal fused to the prefrontal. The postorbital is the postfrontal and vice versa.

Figure 1. Champsosaurus from Dugeon et al. Here the nasal is the ascending process of the premaxilla. The prefrontal is the nasal fused to the prefrontal. The postorbital (pro) is the postfrontal (pof) and vice versa.

Dear Dr. Dudgeon:

It’s always good to see new studies on old skulls.

Based on phylogenetic bracketing the bone traditionally identified as the ‘nasal’ is the ascending process of the premaxilla. That makes the purported ‘prefrontal’ a fused nasal + prefrontal. The postorbital and postfrontal are mislabeled with the other bone identity based on Tchoria (Fig. 2), a taxon not mentioned in your text. See attached.

Choristoderes are not ‘neodiapsid reptiles.’ Phylogenetically they are archosauriformes arising from Proterosuchus, Elachistosuchus and Tchoria. Phylogenetic miniaturization in that lineage lost the antorbital fenestra. See links below.

https://pterosaurheresies.wordpress.com/2013/08/13/champsosaurus-and-its-snorkel-nose/
http://reptileevolution.com/reptile-tree.htm
http://reptileevolution.com/champsosaurus.htm
http://reptileevolution.com/youngina-bpi2871.htm
http://reptileevolution.com/hyphalosaurus.htm
http://reptileevolution.com/lazarussuchus.htm

Best regards,

Figure 1. Tchoria and phylogenetic bracketing help identify bones in the skull of Champsosaurus (Fig. 2).

Figure 2. Tchoria and phylogenetic bracketing help identify bones in the skull of Champsosaurus (Fig. 2).

So, the Dudgeon et al. paper
is yet another great example of a situation in which phylogenetic analysis and bracketing (= comparing related taxa) sheds more light on a specimen than high-resolution micro-computed tomography scanning and/or adding characters (= looking more deeply into one taxon to the exclusion of others).

Figure 2. Champsosaurus skull with premaxilla in yellow.

Figure 3. Champsosaurus skull with premaxilla in yellow, nasal + prefrontal in pink. Bone identities determined by phylogenetic bracketing with Tchoria. See figure 2.

The greatest benefit 
available from the large reptile tree (LRT, 1631 taxa) is this sort of phylogenetic bracketing based on the validated nesting of sisters that have never been tested together in prior studies. You can look more deeply into one skull, as Dudgeon et al. did. Or you can examine many skulls, as ReptileEvolution.com and the LRT enable workers to do (Figs. 2, 4). In this case, using computed tomography on one skull did not put an end to traditional myths regarding the identity of bones in Champsosaurus.

Note to readers who like to harp on these issues:
More characters were not needed to resolve these problems. More taxa were needed.

Firsthand access + computed tomography did not help Dudgeon et al. Rather, a century-old drawing (Brown 1905, Fig. 3), access to several sister taxa for comparison (Figs. 2, 4) and Adobe Photoshop were the tools needed to resolve this issue.

It helps to know what you are dealing with.
Only a wide-gamut phylogenetic analysis that minimizes taxon exclusion can tell you where a specimen nests in the cladogram. Too often workers like Dudgeon et al. rely on vague citations, rather than running tests themselves or citing ongoing and self-repairing studies like the LRT. Publishing a mistake is to be avoided no matter how trivial.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Click to enlarge. Note the tiny remnant of the antorbital fenestra. The squamosal has been broken into several parts.

Figure 4. Dorsal, lateral and palatal views of Late Triassic BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Note the tiny remnant of the antorbital fenestra and the long ascending process of the premaxilla.  The squamosal has been broken into several parts. This is a tiny phylogenetically miniaturized sister to the ancestor of Champsosaurus.

Champsosaurus annectens (Cope 1876, Brown 1905) ~1.5 m in length, Late Cretaceous to Eocene. Champsosaurus was derived from a sister to Tchoiria, and was a sister to other choristoderes, such as Cteniogenys and Lazarussuchus. This clade must have originated in the Late Permian or Early Triassic, but fossils are chiefly from late survivors, hence the wide variety in their morphology.


References
Brown B 1905. The osteology of Champsosaurus Cope. Memoirs of the AMNH 9 (1):1-26. http://digitallibrary.amnh.org/dspace/handle/2246/63
Cope ED 1876.
On some extinct reptiles and Batrachia from the Judith River and Fox Hills beds of Montana: Proceedings of the Academy of Natural Sciences, Philadelphia. 28, p. 340-359.
Dudgeon TW, Maddin HC, Evans DC & Mallon JC 2020. 
Computed tomography analysis of the cranium of Champsosaurus lindoei and implications for choristoderan neomorphic ossification. Journal of Anatomy (advance online publication)
doi: https://doi.org/10.1111/joa.13134
https://onlinelibrary.wiley.com/doi/10.1111/joa.13134

http://reptileevolution.com/champsosaurus.htm

Coeruleodraco: Traditional choristodere mistakes resurface

Occasionally within the Archosauriformes
the antorbital fenestra disappears. That is the case with the clade Choristodera, which Wikipedia describes as “an extinct order of semiaquatic diapsid reptiles. Cladists have placed them between basal diapsids and basal archosauromorphs, but the phylogenetic position of Choristodera is still uncertain.” 

That is so unnecessarily vague.
Just run the analysis. In the large reptile tree chorisotderes are derived from phylogenetically miniaturized proterosuchians like the BPI 2871 specimen and its sister Elachistosuchus.

Figure 1. Coeruleodraco skull as originally interpreted (below) and interpreted here (colors). This is a traditional error. Also note the remnants of an antorbital fenestra in this phylogenetically miniaturized taxon. The maxilla continues posterior to the orbit as in other choristoderes.

Figure 1. Coeruleodraco skull as originally interpreted (below) and interpreted here (colors). This is a traditional error. Also note the remnants of an antorbital fenestra in this phylogenetically miniaturized taxon. The maxilla continues posterior to the orbit as in other choristoderes. Firsthand access does not guarantee better interpretations.

 

Matumoto, Dong, Wang and Evans 2018
bring us a new genus of short-snouted, small choristodere, Coeruleodraco jurassicus (Fig. 1; Late Jurassic). The authors use a ‘by default’ very distant outgroup for their choristodere cladogram: the basal diapsids, Petrolacosaurus and Araeoscelis, because “Outgroup choice is problematic for Choristodera, because the position of the group within Diapsida remains uncertain.” The LRT solved that problem years ago and posted it online. Unfortunatley, the authors did not test the listed outgroup taxa. That’s all they had to do.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Click to enlarge. Note the tiny remnant of the antorbital fenestra. The squamosal has been broken into several parts.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975.Note the tiny remnant of the antorbital fenestra in this phylogenetically miniaturized proterosuchid, basal to Choristodera.

A traditional mistake associated with choristoderes
is the mislabeling of the nasals as the prefrontals (Fig. 1). Both Coeruleodraco and outgroup taxa, like the BPI 2870 specimen demonstrate the ascending process of the premaxilla extends beyond the naris. That it becomes detached from the toothy lateral processes in Champsosaurus (Fig. 3) does not turn the premaxilla into a nasal. We looked at that earlier here and once again, it is due to the exclusion of taxa that clarify the issue. Choristodere workers are not looking at these outgroup taxa for guidance or analysis.

Figure 2. Champsosaurus skull with premaxilla in yellow.

Figure 3. Champsosaurus skull with premaxilla in yellow.

The authors also messed up the finger identification.
The original interpretation of the Coeruleodraco manus (Fig. 4) misidentified the lateral and medial digits along with the olecranon and ulna (violet), which extends behind the humerus as in all other tetrapods. DGS revealed the middle phalanges of manual digit 4 behind the others. The apparently short digit 4 becomes the longest digit when reconstructed (Fig. 4). This matches the manus of other choristoderes.

Figure 3. Manus of Coeruleodraco as originally identified and repaired and reconstructed in color.

Figure 4. Manus of Coeruleodraco as originally identified and repaired and reconstructed in color. Note frame that includes middle phalanges of digit 4. Digit 5 also has a semi-buried element.

Yes, I see things in fossils that others don’t see.
These are just a few of the many examples. In science it’s okay to point out where others have missed things, and the only way to convey that data over the Internet is by tracing and publishing photos (Fig. 4). Others are free to confirm or refute.

Firsthand access does not guarantee better interpretations.
It is important to understand what sister taxa are present and what traits they present. Without a good cladogram answers will not arrive. If there is any question, as in Champsosaurus and Coeruleodraco it’s okay to look at sister taxa for guidance.

Choristoderes are archosauriformes
in which the antorbital fenestra is reduced to absent. Others, like the Wikipedia authors, Matsumoto, Dong, Wang and Evans, who look only at a list of traits present or absent in a taxon are “Pulling a Larry Martin.” That’s a common problem that leads to taxon exclusion. The LRT is a science experiment that you can confirm or refute yourself. It’s time to put the choristodere enigma to rest.

References
Matsumoto R, Dong L, Wang Y and Evans SE 2019. The first record of a nearly complete choristodere (Reptilia: Diapsida) from the Upper Jurassic of Hebei Province, People’s Republic of China, Journal of Systematic Palaeontology
DOI:10.1080/14772019.2018.1494220

Thanks to co-author, Dr. S. Evans,
for sending a PDF link to the paper. I sent her a pdf of the LRT noting that it provided outgroups for choristoderes back to Devonian tetrapods, but no reply accompanied the pdf link.

wiki/Coeruleodraco

BPI 2871 has a new sister: Elachistosuchus huenei

Earlier we looked at a tiny basal choristodere, BPI 2871, which was derived from a line of much larger proterosuchids, according to the large reptile tree.

Recently a new PlosOne online paper (Sobral et al. 2015) reintroduces Elachistosuchus huenei (Janensch 1949, Late Triassicm, Norian, Germany; MB.R. 4520 (Museum für Naturkunde Berlin, Berlin, Germany)) with CT scans.

Figure 1. Elachistosuchus (Janensch 1949, Sobral et al. 2015) is a sister to BPI 2871, a basal choristodere.

Figure 1. Elachistosuchus (Janensch 1949, Sobral et al. 2015) is a sister to BPI 2871, a basal choristodere, has been misidentified for over fifty years.The left upper temporal fenestra has been largely closed by crushing here. Like BPI 2871, the nares were located on top of the skull, close to the snout tip. Note the vestige of the antorbital fenestra.

And they don’t know what it is. 
According to Sobral et al, Elachistosuchus could be “an archosauromorph, a lepidosauromorph or a more basal, non-saurian diapsid.” That confusion arises from using outdated matrices with too few generic taxa and too many suprageneric taxa.

Sobral et al. used the matrix from Chen et al. 2014, which nested Elachistosuchus in a polygamy with Choristodera, Prolacerta + Tanystropheus + Macrocnemus, and Trilophosaurus + Rhynchosauria + Archosauriformes. As readers know the large reptile tree found many of these taxa on opposite sides of the reptile cladogram.

Sobral et al. also used the matrix from Ezcurra et al. 2014, which nested Elachistosuchus with the gliding Permian lepidosauriform, Coelurosauravus.

Hmmmm…

Sobral et al. report: 
“These different placements highlight the need of a thorough revision of critical taxa and new character sets used for inferring neodiapsid relationships.” 

Exactly. 
That’s why large reptile tree and reptileevolution.com are here. It’s good to have hundreds of specimen-based taxa for new taxa to nest with. More choice. More accuracy. Complete resolution.

To their credit,
a Sobral et al. analysis nested Elachistosuchus with choristoderes.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Click to enlarge. Note the tiny remnant of the antorbital fenestra. The squamosal has been broken into several parts.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Click to enlarge. Note the tiny remnant of the antorbital fenestra. The squamosal has been broken into several parts. This is a sister to Elachistosuchus.

Among earlier workers
Janensch (1949) considered Elachistosuchus a pseudosuchian archosaur with an antorbital fenestra. Walker (1966 ) considered  Elachistosuchus a rhynchocephalian lepidosaur.

The large reptile tree (now 575 taxa)
finds Elachistosuchus nests firmly as a sister to the BPI 2871 specimen (Fig. 3) that Gow mistakenly attributed to Youngina, but it nests far from Youngina at the base of the large and small choristoderes. And these two taxa are both derived from much larger proterosuchids in yet another case of phylogenetic miniaturization at the genesis of a new clade, in this case the Choristodera.

Elachistosuchus has a larger orbit and a maxilla with a straight, not convex, ventral margin of the maxilla than the BPI 2871 specimen. The former extends the geographic range of the latter, from southern Africa to Germany.

Both probably look like juvenile proterosuchids (whenever they are discovered, we can compare them). Phylogenetic miniaturization often takes juvenile traits and sizes and makes them adult traits and sizes to start new clades.

Janensch thought Elachistosuchus had an antorbital fenestra. As in BPI 2871, that is the vestige of the antorbital fenestra found in ancestors and lost in descendants.

Contra the title of the Sobral et al. paper
Elachistosuchus huenei has nothing to do with the origin of ‘Sauria.’

Sauria definition: “.Any of various vertebrates of the group Sauria, which includes most of the diapsids, such as the dinosaurs, lizards, snakes, crocodilians, and birds. Sauria was formerly a suborder consisting ofthe lizards” Rather, Elachistosuchus is a basal choristodere and a derived proterosuchid according to the large reptile tree. Based on the current definition of ‘Sauria’ ‘Sauria’ is synonymous with ‘Amniota’ which is a junior synonym for ‘Reptilia’ because the last common ancestor of lizards and dinosaurs is the basalmost reptile/amniote, Gephyrostegus bohemicus.

The reason why Sobral et al. were confused
with regard to their blurred nesting of Elachistosuchus is due to taxon exclusion. BPI 2871 is a rarely studied taxon and was not included in their analyses. Moreover, traditional paleontologists are not sure what choristoderes are. They don’t recognize them as being derived proterosuchids. And to make matters worse, traditional paleontologists prefer to think of Proterosuchus specimens as members of an ontogenetic series, when they should consider them as a phylogenetic series.

Figure 4. This is where Elachistosuchus nests, next to BPI 2871, at the base of the Choristodera.

Figure 3. This is where Elachistosuchus nests, next to BPI 2871, at the base of the Choristodera. Click to see the complete reptile cladogram.

The large reptile tree (Fig. 3) has proven itself time and again to solve paleontological problems in the reptile family tree. It is unfortunate that it has been rejected for publication so many times. If published, it can be use.

A MacClade file is available on request.

References
Chen X, Motani R, Cheng L, Jiang D, Rieppel O. 2014. The enigmatic marine reptile Nanchangosaurus from the Lower Triassic of Hubei, China and the phylogenetic affinities of Hupehsuchia. PLoS ONE. 2014; 9: e102361. doi: 10.1371/journal.pone.0102361 PMID: 25014493
Ezcurra MD, Scheyer TM, Butler RJ 2014. The origin and early evolution of Sauria: reassessing the Permian saurian fossil record and the timing of the crocodile-lizard divergence. PLoS ONE. 2014; 9: e89165. doi: 10.1371/journal.pone.0089165 PMID: 24586565
Gow CE 1975. The morphology and relationships of Youngina capensis Broom and Prolacerta broomi Parrington. Palaeontologia Africana, 18:89-131.
Janensch W 1949. Ein neues Reptil aus dem Keuper von Halberstadt. N Jb Mineral Geol Palaeont B. 1949:225–242.
Sobral G, Sues H-D & Müller J 2015. Anatomy of the Enigmatic Reptile Elachistosuchus huenei Janensch, 1949 (Reptilia: Diapsida) from the Upper Triassic of Germany and Its Relevance for the Origin of Sauria. PLoS ONE 10(9): e0135114. doi:10.1371/journal.pone.0135114
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0135114
Walker AD 1966. Elachistosuchus, a Triassic rhynchocephalian from Germany. Nature. 1966; 211: 583–585.

wiki/Elachistosuchus

Did Champsosaurus and Tchoiria have an antorbital dimple?

The large reptile tree nests choristoderes, both large and small, with archosauriformes, derived from a long series of proterosuchids ending with the small, former younginid, BPI 2871 (Fig. 1). This tiny transitional taxon at the base of the Choristodera documents yet another case of phylogenetic miniaturization and independent loss of the antorbital fenestra.

All proterosuchids had an antorbital fenestra,
but tiny BPI 2871 had a dimple, sealed in back (Figs. 1, 7), not a fenestra.

BPI2871-3views588

Figure 1. Dorsal and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. A vestige antorbital fenestra may be present here. Compare to figure 2.

I ran across this image
of the skull of Champsosaurus (Fig. 2), a large, long-snouted choristodere. It also appears to retain a small dimple (pink arrow). The dimple is otherwise undocumented (Fig. 4) and not always duplicated, as shown here (Fig. 4).

Figure 2. This specimen of Champsosaurus appears to retain a tiny antorbital fenestra. Is this replicated in other specimens?

Figure 2. This specimen of Champsosaurus appears to retain a tiny antorbital fenestra. Is this replicated in other specimens? Or is it a shadow? If it is a shadow, this is exactly where the AOF would be if present. The nares (nostrils) are above the jaw tips.

Getting back to ‘did Champsosaurus have an antorbital dimple’?
Apparently only sometimes. In this clade the dimple is a vestige at best, and not always present. Most other choristoderes do not have an antorbital dimple. But some do.

Figure 3. Cast of Champsosaurus from Triebold Palentology. Here again we see that dimple in front of the eye.

Figure 3. Cast of Champsosaurus from Triebold Palentology. Here again we see that dimple in front of the eye, below the lacrimal prefrontal and above the maxilla. Click to enlarge.

Figure 4. How Barnum Brown illustrated the preorbital region of Champsosaurus, without a hint of a dimple or antorbital fenestra.

Figure 4. How Barnum Brown illustrated the preorbital region of Champsosaurus, without a hint of a dimple or antorbital fenestra.

Remember
living crocs also lack an antorbital fenestra. By convergence with choristoderes, crocs lost what their ancestors had. So there is precedence for such an evolutionary change. Brown 1905 did not illustrate a preorbital dimple in his treatise on Champsosaurus (Fig. 4).

Tchoiria is a basal choristodere
and this specimen (Fig. 5) appears to have a preorbital dimple along the same lines as the Triebold Champsosaurus cast (Fig. 3). But another Tchoiria (T. kauseni; “choy-er-ee-ya”??, Fig. 6, Ksepka Gao and Norell 2005) does not have such a dimple.

Figure 5. The skull of Tchoiria appears to also have a preorbital dimple.

Figure 5. The skull of Tchoiria appears to also have a preorbital dimple. This one appears to have a longer rostrum than in figure 6 and indeed, the two species are known to have different tooth counts.

 

What Tchoiria tells us about champsosaurs
Unlike other tetrapods, champsosaurs created an upper temporal arch from the postfrontal contacting the squamosal. In Tchoiria you can see that transition taking place (overlooked in the original paper, Fig. 6). Tchoiria also shows the separation of the prefrontal and nasal in their traditional places. By doing so, the very long ascending process of the premaxilla (not a narrow fused set of nasals) is revealed.

Figure 6. Tchoiria klauseni as originally interpreted and as interpreted using DGS. Note this specimen shows the the transition from the postorbital to the postfrontal contacting the squamosal. This also shows the extent of the premaxillary ascending process (yellow), nasals separate from prefrontals.

Figure 6. Tchoiria klauseni as originally interpreted and as interpreted using DGS. Note this specimen shows the the transition from the postorbital to the postfrontal contacting the squamosal. This also shows the extent of the premaxillary ascending process (yellow) and the nasals (pink_ separate from prefrontals (brown) Those would fuse in Champsosaurus (Fig. 4).

A closeup of the dimple on BPI 2871 might be instructive. 
The dimple here looks like the origin of the antorbital fenestra in Youngoides romeri (FMNH UC 1528 seen here), but it arrives at a much more derived node on the large reptile tree cladogram.

Rostral area of BPI 2871, formerly considered a younginid and here nesting at the base of the Choristodera.

Figure 7. Rostral area of BPI 2871, formerly considered a younginid and here nesting at the base of the Choristodera. Colors reappear every 5 seconds. Note the infilling of the antorbital fenestra. This is only known after cladistic analysis. The two sides of the same skull each provide clues as to the in vivo morphology. Splinters are often difficult to identify. This interpretation may change as time goes by. In the lower image, the fragile back of the dimple has been shifted dorsally. 

References
Brown B 1905. The osteology of Champsosaurus Cope. Memoirs of the AMNH 9 (1):1-26. http://digitallibrary.amnh.org/dspace/handle/2246/63
Ksepka, DT, Gao K-Q and Norell MA 2005. A New Choristodere from the Cretaceous of Mongolia. American Museum Novitates 3468. 22pp.

BPI 2871 – Now it is the oldest known choristodere

Updated July 5, 2015 with a lateral and occipital views of BPI 2871 and few text changes.

Earlier we looked at some of the earliest known Choristoderes recognized by traditional paleontologists. According to Wikipedia, Choristoderes are difficult to nest.Cladistshave placed them between basal diapsids and basal archosauromorphs, but the phylogenetic position of Choristodera is still uncertain. It has also been proposed that they represent basal lepidosauromorphs. Most recently, workers have placed Choristodera within Archosauromorpha.”

Figure 1. The sister to Doswellia, the BPI2871 specimen of Youngina.

Figure 1.  BPI 2871 specimen attributed to Youngina slightly modified from Gow 1975.

Gow 1975
considered the skull-only fossil, BPI 2871 (Bernard Price Institute, Fig. 1) a specimen of Youngina, despite its Late Triassic appearance (all other Youngina are Late Permian). For years his drawing (Fig. 1) was the only data I had for this specimen. Recently, and after several years of waiting, a requested image (Fig. 2) from the BPI was emailed and it clarified my understanding of this vaguely-croc-like basal archosauriform — perhaps without an antorbital fenestra — or perhaps this is one of the last taxa in the choristodere lineage to have a vestige antorbital fenestra, as it appears.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Click to enlarge. Note the tiny remnant of the antorbital fenestra. The squamosal has been broken into several parts.

Figure 2. Dorsal, lateral and palatal views of BPI 2871 with bones colorized above. Below, reconstructed images of BPI 2871 tracings. It is more complete than illustrated by Gow 1975. Click to enlarge. Note the tiny remnant of the antorbital fenestra. The squamosal has been broken into several parts.

The new image
of BPI 2871 (Fig. 2) indicates that the skull is more complete than illustrated by Gow 1975. With this new data added to the matrix of the large reptile tree (Fig. 3) the nesting of BPI 2871 shifts it closer to the base of the Choristodera, evidently making it the oldest known member in the Late Triassic (210 mya).

Figure 3. Subset of the large reptile tree focusing on the pararchosauriformes and the Choristodera.

Figure 3. Subset of the large reptile tree focusing on the pararchosauriformes and the Choristodera.

BPI 2871 is distinctly different
in chronology and morphology from Y. capensis, a fact overlooked or ignored by Gow who considered it another specimen of Youngina capensis. Like Bennett (1995, 1996, 2014) Gow was a lumper.

Tiny
BPI 2871 apparently lost the antorbital fenestra (requested lateral views are on the way, I am told). An absent antorbital fenestra is the reason why choristoderes have been difficult to nest in the reptile family tree. In the large reptile tree I simply popped in the traits and let the software recover the nesting. And it all makes sense. Note the resemblance of tiny BPI 2871 to the much larger Chanaresuchus, a related taxon. Like ancestral proterosuchids, BPI 2871 retains an overhanging snout and a shorter mandible. This specimen also retains postparietals.

Phylogenetic miniaturization
We keep meeting phylogenetic miniaturization at the base of novel clades and the same holds true for BPI 2871 and the Choristodera. Predecessor and successor taxa are both much larger.

Doswellia
and a rather large specimen of Proterosuchus (SAM PK-K 10603, early to Middle Triassic) also nest at the base of the Choristodera. Both retain an antorbital fenestra, rather feeble in the case of Doswellia (late Triassic).

Phylogeny
With all the new data BPI 2871 shifted one node over toward the choristoderes. This proves that inaccurate data, in the form of a simple drawing (Fig. 1), can still carry a large amount of data, that in this case, rather accurately places this taxon on the large reptile tree. More data refines that nesting.

References
Gow CE 1975. The morphology and relationships of Youngina capensis Broom and Prolacerta broomi Parrington. Palaeontologia Africana, 18:89-131.

wiki/Youngina
wiki/Choristodera

 

 

Philydrosaurus: another basal choristodere

Figure 1. Philydrosaurus in several views. This specimen nests as the most basal small, short-snouted choristodere. Juveniles surround it. DGS indicates this specimen had lateral temporal fenestrae, but greater resolution may modify this hypothesis.

Figure 1. Philydrosaurus in several views. This specimen nests as the most basal small, short-snouted choristodere. Juveniles surround it. DGS indicates this specimen had lateral temporal fenestrae, but greater resolution may modify this hypothesis.

Philydrosaurus proseilus (Gao and Fox 2005, Early Cretaceous, scale bar = 2 cm) is a basal choristodere with distinct ridges on the skull over the orbits. The lateral temporal fenestra is reported as closed, but the low-resolution image provided appears to show lateral temporal fenestra. It was scored without them. The cervicals do not decrease toward the skull. Several juveniles were found associated with the presumed mother. The right skull (above) does not seem to accurately reflect the fossil. Higher resolution images have been requested.

Figure 2. Philydrosaurus compared to the BPI 2871 specimen wrongly assigned to Youngina, itself a descendant of Proterosuchus.

Figure 2. Philydrosaurus compared to the BPI 2871 specimen wrongly assigned to Youngina, itself a descendant of Proterosuchus.

Update:
The blogpost on Nundasuchus has been updated with a new reconstruction (Fig. 3) and nesting with Qianosuchus and Ticinosuchus, also reflected at reptileevolution.com.

Figure 1. from Nesbitt et al. 2014. Plus foot reconstructed here and closeups of the mandible and tooth.

Figure 3. from Nesbitt et al. 2014. Plus foot reconstructed here and closeups of the mandible and tooth.

References
Gao K-Q and Fox RC 2005. A new choristodere (Reptilia: Diapsida) from the Lower Cretaceous of western Liaoning Province, China, and phylogenetic relationships of Monjurosuchidae. Zoological Journal of the Linnean Society 145 (3): 427–444.

The many faces of Proterosuchus: not a growth series

Recent papers by Ezcurra and Butler (2015) and Welman (1998) purported to show a growth series in Proterosuchus (Fig. 1; Broom 1903) using a number of small to large skulls. Anyone can see why these authors thought these two skulls could be samples from a growth series. But is that what these skulls truly represent? After all, the only way to become a giant proterosuchid is to evolve over hundreds of generations from an original pair of small proterosuchids (Figs. 2, 3). 

Figure 1. Two proterosuchid skulls which Ezcurra and Butler report represent a juvenile and adult. That hypothesis is not supported by phylogenetic analysis.

Figure 1. Two proterosuchid skulls which Ezcurra and Butler report represent a juvenile and adult. That hypothesis is not supported by phylogenetic analysis. Compare these images to RC96 and RC59 in figure 3.

Unfortunately,
Ezcurra and Butler did not perform the required phylogenetic analysis of the several specimens of Proterosuchus. Instead, like Bennett (1995) did with Rhamphorhynchus, Ezcurra and Butler assumed a single species of Proterosuchus was present with differences attributed to ontogeny. The smaller Proterosuchus skulls, Ezcurra and Butler reported, were juveniles. They reported the shape of the skulls changed during ontogeny, becoming taller and relatively narrower with ages. They did not consider the possibility of ‘Cope’s Rule’ or the process of evolution in the creation of a large Proterosuchus.

As with Rhamphorhynchus, the large reptile tree found the morphological differences in Proterosuchus/Chasmatosaurus were all due to phylogeny (evolution), not ontogeny. Basal Proterosuchus specimens more closely resembled outgroup Youngina specimens. Derived Proterosuchus specimens more closely resembled more derived genera (Fig. 3, like Diandongosuchus and Doswellia

Figure 2. The origin and evolution of Proterosuchus based on skulls. On the left to scale. On the right to the same length.

Figure 2. The origin and evolution of Proterosuchus based on skulls. On the left to scale. On the right to the same length. What you’re seeing here is the evolutionary steps taken to produce the large RC96 skull. You have to start somewhere, and here you can start with the AMNH 5561 specimen of Youngina.

Bottom line:
Despite their size differences, all of the Proterosuchus skulls in figures 1-3 are adults, or at least they can be scored as adults.

Figure 3. The many faces of Proterosuchus to scale and in phylogenetic order, among with their closest known relatives. Note the phylogenetic miniaturization, reduction of the drooping premaxilla and loss of the antorbital fenestra after the TM 201 specimen of Chasmatosaurus. Click to enlarge.

Figure 3. The many faces of Proterosuchus to scale and in phylogenetic order, among with their closest known relatives. Note the phylogenetic enlargement, phylogenetic miniaturization, reduction of the drooping premaxilla and loss of the antorbital fenestra after the TM 201 specimen of Chasmatosaurus. Click to enlarge. Yes there are two choristoderes related to the tiny BPI 2871 specimen wrongly attributed to Youngina. Moving the BPI 2871 closer to Youngina and the Choristodera adds 30 steps, so it appears that the antorbital fenestra disappeared in this lineage.

I’m not sure if we know
what a juvenile Proterosuchus specimen looks like. I don’t think we’ve found one yet. My guess is it will look like and phylogenetically nest with a specific adult, only smaller, as in pterosaurs and other reptile taxa. To that point, the smallest putative Proterosuchus specimen shown here (Fig. 3), RC 59, (formerly considered Elaphrosaurus), phylogenetically nests as a derived proterosuchian. At it nests with an even smaller little archosauriform, BPI 2871, formerly referred to Youngina. So the littlest Proterosuchus is not a juvenile, as imagined by Ezcurra and Butler. Rather RC 59 is going through phylogenetic miniaturization and it is not done shrinking.

By the way we can be pretty confident about these nestings because the taxon list has grown to 540+ taxa.

You might find this fascinating..
Everyone who sees a Proterosuchus skull must think, “That odd premaxilla drooping snout…that has to belong to a terminal taxon. What could possibly evolve from that?” Well, apparently the answer is: “a long list of taxa,” which you already know and we’ll meet in future blog posts. Some hints as to the identify of those taxa are in figure 3 and at reptileevolution.com. That premaxilla doesn’t stay so long and droopy in descendant taxa, perhaps due to neotony.

And one more thing…
The skulls of Proterosuchus and Chasmatosaurus are not shaped differently because they were crushed during burial in different directions. No, what you see is what you get. And it all appears to fit together in the tree of life — without any a priori assumptions as to relative ontogenetic age.

The many faces of Proterosuchus and Chasmatosaurus
have been perplexing, but phylogenetic analysis puts everything in order. Let’s get this problem behind us. Please encourage paleontologists to run all their taxa through analysis before assuming any are juveniles.

References
Broom R. 1903. On a new reptile (Proterosuchus fergusi) from the Karroo beds of Tarkastad, South Africa. Annals of the South African Museum 4: 159–164.
Ezcurra MD and Butler RJ 2015. Post-hatchling cranial ontogeny in the Early Triassic diapsid reptile Proterosuchus fergusi. Journal of Anatomy. Article first published online: 24 APR 2015. DOI: 10.1111/joa.12300
Welman J 1998. The taxonomy of the South African proterosuchids (Reptilia, Archosauromorpha). Journal of Vertebrate Paleontology 18 (2): 340–347.

 

The most primitive Choristoderes?

Choristoderes
are a difficult clade to figure out both inside and out.

Wikipedia reports, “Champsosaur skulls are actually very similar to lizard skulls, though heavily modified. This has led some researchers to consider champsosaurids as lepidosauromorphs. However, champsosaurs lack the complex quadrate of lepidosaurians. With features of both diapsid groups, the phylogenetic position of Choristodera is highly confused.”

Matsumoto et al. (2007) reports, “This tree confirms the monophyly of Neochoristodera (Evans and Hecht1993) including Champsosaurus, Ikechosaurus, Simoedosaurus, and Tchoiria. The relationships of the non−neochoristoderan taxa have been more controversial. In our analysis, the Jurassic Cteniogenys retains a basal position, with the Chinese Philydrosaurus (Gao and Fox 2005) one node above it.”

Note the loss of resolution at the base of their tree (Fig. 1).

Figure 1. The family tree of the Choristodera according to Matsumoto et al. 2007. Early eras added in color. Note that ancestral taxa were large and long-snouted with four temporal openings, not two.

Figure 1. Click to enlarge. The family tree of the Choristodera according to Matsumoto et al. 2007. Early eras added in color. Note that some ancestral taxa, like Diandongosuchus, were large and long-snouted. Others were small, like the BPI specimen of Youngina, but equally long snouted. So the long-snouted forms may be more primitive. Note the loss of resolution at the clade base.

The question in the Matsumoto et al. 2007 tree is what happened in pre-Jurassic times? In 2007 Youngina was indeed the best outgroup taxon, but several varieties are known and some of these have a longer rostrum. Unfortunately these were overlooked as Matsumoto et al. focused on known choristoderes. In 2012 the large younginidDiandongosuchusbecame known. While it nests at the base of the parasuchia, it also nests at the bases of the Choristodera and Chanaresuchia. Yes, it has an antorbital fenestra, and so do some Youngina specimens. It is possible that the Choristodera secondarily lost their aof. Or they never had one.

In any case, the large reptile tree recovered a similar basal split between the large and small choristoderes, but with complete resolution among the five small ones and three large ones. More precision in the character scoring of the various Youngina specimens should add clues to this mystery. Don’t discount those subtle variations!

References
Matsumoto R, Evans SE and Manabe M 2007. The choristoderan Monjurosuchus form the Early Cretaceous of Japan. Acta Palaeontologica Polonica 52(2):329-350.

 

The droopy snout of Doswellia

UPDATED
on April 2, 2014 with new data on the pmx/mx suture and pmx ascending process.

Earlier, we looked at that odd archosauriform, Doswellia (Late Triassic, Weems 1980, Dilkes and Sues 2009, Heckert et al. 2012). It was originally considered a sister to Proterochampsidae, which is very close. The large reptile tree nested it as a basal, but not very plesiomorphic, choristodere, derived from a sister to Youngina BPI 2871 (Fig.1 ) and otherwise close to ChampsosaurusSimoedosaurus and Diandongosuchus.

Figure 1. The sister to Doswellia, the BPI2871 specimen of Youngina.

Figure 1. The sister to Doswellia, the BPI2871 specimen of Youngina.

Earlier I accepted the straight mandible reconstruction of Weems (1980) and Heckert et al. (figure 6, 2012).

However, the mandible illustrated alone in Dilkes and Sues (2009) is ventrally concave. Putting that into a reconstruction (Fig. 2) also matches the referred broken rostrum of Heckert et al. (2012).

The Droopy Snout
Convergent with Proterosuchus (Fig. 3), Doswellia had a droopy snout that started to droop at mix maxilla, instead of a the pmx/mx suture.

 

Figure 2. Click to enlarge. Doswellia reconstructed from bone pieces. The postfrontal and postorbital are fused but colored separately. A tiny lateral temporal fenestra remains. The larger specimen may have had a stronger curve in the rostrum.

Figure 2. Click to enlarge. Doswellia reconstructed from bone pieces. The postfrontal and postorbital are fused but colored separately. A tiny lateral temporal fenestra remains. The larger specimen may have had a stronger curve in the rostrum.

Doswellia is reported to have lost lateral temporal fenestrae, but tiny vestiges are still apparent (Fig. 1). Small antorbital fenestra are present, but they appear to be vestiges, on their way to disappearing, too. Choristoderes do not have antorbital fenestra.

FIgure 3. Proterosuchus (above) and Doswellia (below). Note the similarity of the drooping premaxilla, the shape of the pes and the overall low slung body.

FIgure 3. Proterosuchus (above) and Doswellia (below). Note the similarity of the drooping premaxilla, the shape of the pes and the overall low slung body.

The nares
are at the tip of the snout, but dorsolateral with a short premaxilla. We also see something like this in champsosaurus. The posterior portion of the ascending process of the premaxilla appears to be present, stuck to the side of the maxilla due to taphonomic forces.

This new reconstruction comes from transferring existing drawings using DGS (digital graphic segregation) to create a reconstruction. Color also helps.

And this is blog post #900.

References
Dilkes D and Sues H-D 2009. Redescription and phylogenetic relationships of Doswellia kaltenbachi (Diapsida: Archosauriformes) from the Upper Triassic of Virginia. Journal of Vertebrate Paleontology 29(1):58-79
Heckert AB, Lucas SG and Spielmann JA 2012. A new species of the enigmatic archosauromorph Doswellia from the Upper Triassic Bluewater Creek Formation, New Mexico, USA. Palaeontology (Blackwell Publishing Ltd) 55(6): 1333-1348.
Sues H-D, Desojo JB and Ezcurra MD 2013. Doswelliidae: a clade of unusual armoured archosauriforms form the Middle and Late Triassic. Geological Society, London
Weems RE 1980. An unusual newly discovered archosaur from the Upper Triassic of Virginia, U.S.A. Transactions of the American Philosophical Society, New Series 70(7):1-53

wiki/Doswellia

Champsosaurus and its snorkel nose

Choristoderes are a varied clade of mostly aquatic, often croc-like reptiles descending from certain long-snouted younginids, like the BPI-2871 specimen (Fig. 2). Doswellia (Fig. 2) is closely related to choristoderes, splitting off at the base to form its own clade. Diandongosuchus likewise split off early, giving rise to later parasuchids (phytosaurs), proterochampsids and chanaresuchids.

Many from this varied clade of pararchosauriformes had nostrils shifted back from the snout tip, reaching an acme with parasuchians. Champsosaurus (Figs. 1, 2) was different. The naris was located at the very tip, probably to act as a snorkel, in order to breathe without surfacing. Given the unusual morphology of the snout tip, it’s very possible that the reversion to the tip was a secondary adaptation.

Figure 1. Neochoristoderes including Champsosaurus, Simoedosaurus and Ikechosaurus.

Figure 1. Neochoristoderes including Champsosaurus, Simoedosaurus and Ikechosaurus. Premaxilla in yellow. Nasals in pink. Lacrimals in orange. The prefrontals were fused to the nasals. Note: the largest genus here is the most primitive with lateral temporal fenestra oriented laterally and the nares still dorsal on the rostrum.

The identification of the rostral bones in Champsosaurus is controversial. Here we’ll look at some heretical labels for traditional paradigms.

Figure 2. Various choristoderes and their kin with a focus on the bones surrounding the naris and comprising the snout.

Figure 2. Various choristoderes and their kin with a focus on the bones surrounding the naris and comprising the snout. The nostrils migrate posteriorly by convergence in Lazarussuchus and Diandongosuchus.

Traditionally
In Champsosaurus (Fig. 1) the dorsal medial bone is traditionally considered the nasal and the paired bones following it are considered the prefrontals. However if you look at all the closest kin to Champsosaurus it becomes clear that the paired bones remain traditional nasals. The prefrontals are simply missing, likely due to fusion with the nasals. That means the tooth-bearing portions of the premaxilla wrapped completely around the rostrum and nares until they came into contact with the ascending process of the premaxilla, which extends beyond the naris in many related taxa.

Did Champsosaurus once have an antorbital fenestra?
Related taxa, including Diandogosuchus and Doswellia had an antorbital fenestra and there are signs of a nascent or vestigial antorbital fenestra in certain Champsosaurus (Fig. 3). If it’s there, it’s tiny, but worth searching for.

Figure 3. The rostrum of a champsosaur color coded to identify the premaxilla (yellow), nasals (pink), prefrontals (purple) and lacrimals (orange). The vestigial or nascent antorbital fenestrae are in black, along with the snout-tipped nares (at far right).

Figure 3. The rostrum of a champsosaur color coded to identify the premaxilla (yellow), nasals (pink), prefrontals (purple) and lacrimals (orange). The vestigial or nascent antorbital fenestrae are in black, along with the snout-tipped nares (at far right). If anyone has better data, please send it along.

Choristoderes are underrepresented in the fossil record.
So are doswellids and basal parasuchians. What this means, with present data, is basal taxa appear to be large forms, which goes against the grain of evolutionary patterns. These named taxa are all quite derived at their first appearance. I’m guessing we’re likely to find smaller basal and transitional forms, more like the BPI-2781 specimen (Fig. 2) as they become known.

It’s also interesting that the largest choristodere, Simoedosaurus (Fig. 1) has the more primitive skull, with a dorsal set of nares and more laterally-oriented lateral temporal fenestra. The smaller Champsosaurus has the more derived snout tip and more dorsally open lateral temporal fenestrae.

Popping paradigms is what we do here.
If you have data that supports other positions, please send them forward.

Updated were made to today to the post on Varanosaurus and archosauromorph diapsid origins.