Middle Triassic Plagiosternum enters the LRT

Shoch et al 2025 wrote:
“Plagiosaurids form a small but highly disparate clade of Triassic temnospondyls that are characterized by extremely flattened and wide skulls, large orbits and a knobby to pustular ornamentation. The largest European taxon is Plagiosternum granulosum from the Middle Triassic of Germany. Originally known from fragmentary bonebed material only, recent finds add well-preserved specimens that reveal the structure of the skull, mandible and pectoral girdle in great detail. Plagiosternum granulosum was the most salt-tolerant temnospondyl in the Ladinian palaeoenvironments, where it dwelled brackish lagoons and lakes and formed a community with nothosaurids and other euhaline taxa.”

When added to
the large reptile tree (LRT, 2340 taxa) Plagiosterum nested, as expected, between two other plagiosaurs: Gerrothorax and Plagiosuchus.

Click links to see images.

The authors
“raised the question of what caused the convergent increase of orbit size in plagiosaurids.”

“The loss of circumorbital elements, which occurred convergently in up to three plagiosaurid clades (Witzmann & Schoch, 2024), might well have been triggered by different factors in divergent taxa.”

The LRT tests fewer plagiosaurids, so only one loss of circumorbital elements is indicated as the orbit increases in size, creating a confluent upper temporal fenestra by way of erosion of the postorbital elements.

“Plagiosternines, in contrast, did not have enlarged adductor muscles, as evidenced by the normal size of the adductor fenestra in the mandible.”

“Steyer (2014) suggested that salinity tolerance may have been facilitated by salt or lacrimal glands in euhaline temnospondyls, and the enormous size of the orbits in P. granulosum might well have housed such an organ.”

In the LRT
plagiosaurs are among the most primitive tetrapods, not derived temnospondyls. Add taxa to test this. Let us know if you confirm, refute or modify this LRT hypothesis.

References
Shoch RR et al (5 co-authors) 2025. Morphology and ontogeny of the plagiosaurid temnospondyl Plagiosternum granulosum from the Middle Triassic of Germany. PalZ https://doi.org/10.1007/s12542-025-00748-7

wiki/Plagiosternum
wiki/Plagiosauridae

Scyllacerta enters the LRT as a tiny early member of the Diapsida (within Archosauromorpha)

Scyllacerta creanae
(Jenkins et al 2026, SAM-PK-K7710a, Late Permian, South Africa) was originally classified as the oldest known member of the stem-reptile family Younginidae and the oldest neodiapsid. Adding taxa in the LRT modifies these preliminary findings. The authors did not realize a diapsid skull architecture evolved more than once by convergence and that the clade Reptilia extended back to the last common ancestor, Silvanerpeton, from the Early Cretaceous (Viséan) in the large reptile tree (LRT, 2340 taxa).

Thus the claim by the authors that ‘the origin of the tympanic fossa in reptiles is revealed by’ Scyllacerta creanae cannot be true.

Akkedops bremneri  
(Mooney, Scott and Reisz, SAM-PK-K6205, Late Permian, South Africa) described as a new early reptile from South Africa. The authors included the SAM-PK-K77710a specimen in this genus. They claimed that the cranial anatomy of all three specimens was indistinguishable. The taxa are similar enough to nest together in the LRT, but the straight rostral margin of Scyllacerta is distinct from the convex rostral margin of Akkedops bremneri. Likewise these authors did not realize a diapsid skull architecture evolved more than once by convergence and that the clade Reptilia extended back to Silvanerpeton.

Click the links above to see images. WordPress issues have not yet gone away.

References
Jenkins XA et al (10 co-authors) 2026. The origin of the tympanic fossa in reptiles revealed by a late Permian neodiapsid. Palaeontology. 69 (1) e70041: 1–15.
Mooney ED, Scott D and Reisz RR 2025. A new stem saurian reptile from the late Permian of South Africa and insights into saurian evolution. Swiss Journal of Palaeontology. 144 (1): 10.

wiki/Scyllacerta
wiki/Akkedops

https://pterosaurheresies.wordpress.com/2025/03/13/akkedops-new-name-for-the-assembly-of-juvenile-youngina-2-new-skulls/

 

 

Evolutionary transition to birds on YouTube video

https://www.youtube.com/watch?v=GV49b9RFCRw

Transformations… the complex evolutionary transition to birds by Dr Corwin Sullivan.

@7:40 Sullivan’s cladogram divides Archosauria into Pseudosuchia and Ornithodira. Adding taxa recovers a clade Archosauria containing only Crocodylomorpha + Dinosauria and does not recover Pseudosuchia, nor Ornithodira. The latter traditionally includes Pterosauria, but those were shown to be lepidosaurs related to Huehuecuetzpalli and small bipedal tanystropheids like Sharovipteryx, Longisquama and Cosesaurus almost 20 years ago (2000, 2007). In this new order Ornithodira is a junior synonym for Reptilia and so is Amniota.

@8:18 Sullivan reports the origin of birds from theropods has been widely accepted for three decades = since 1990 (on this 2016 video). This points out a major fault with paleontology: failure to embrace the obvious. By 1996 Ostrom’s work with Archaeopteryx and Deinonychus was, by then, twenty years old and Huxley’s advocacy for the transitional status of feathered Archaeopteryx was over a century old.

@9:00 Sullivan fails to acknowledge the variety of Solnhofen birds, labeling them all Archaeopteryx, which has become a brand name, unfortunately, more valuable to lump than to separate.

@9:57 Sullivan shows several feathered theropods, labeling them all ‘first-cousins’ and ‘close’ to birds. This is incorrect. Many theropods unrelated to birds also had feathers. Sullivan more or less acknowledges this fact later.

@14:46 Sullivan discusses scansoriopterygids as closely related to birds. After analysis these are birds, some of them secondarily flightless, which Sullivan later reports.

@15:23 Sullivan brings up microraptors, which are convergent with birds, not closely related after analysis. Cladogram here: http://reptileevolution.com/reptile-tree.htm

@21:00 Yi qi does not have a strange wing. Rather that was a misinterpretation of a torsion fracture detailed here: https://pterosaurheresies.wordpress.com/2021/07/04/when-a-simple-torsion-fracture-turned-an-early-cretaceous-bird-into-bizarre-bat-wing-dinosaur/

Which brings up another widespread weakness in paleontology: a failure to test odd interpretations and resulting ‘unique’ concepts like the styliform element. So Sullivan is promoting a myth here. He wonders why this theropod had membraneous wings when no others do. That should have led him and others to figure out why this element was misinterpretated.

@28:38 Sullivan moves on to the rib cage and other elements associated with flight – completely ignoring the key to flapping: the elongate, locked-down coracoid found in birds and pterosaurs. Bats substitute an elongate, locked-down clavicle since they lack a coracoid. Non-avian theropods have a short, round, sliding coracoid. Flapping preceeded flying and gliding in birds. This insight has been overlooked and ignored by avian paleontologists, other than Ken Dial (2003).

 

 

 

Four camera-type eyes in the earliest vertebrates from the Cambrian Period

Lei et al 2026 re-interpteted former nasal capsules
in Cambrian Haikouichthys as twin, close-set pineal eyes with lens and retina medial to larger traditional lateral eyes.

The identification of retinal tissue, along with the eyeball-ish shape, shifted the traditional paradigm from nose to eye. Good job, authors!

“Vertebrate vision is mainly accommodated by a pair of lateral image-forming camera-type eyes and is supplemented in non-mammalian vertebrates by a dorsal pineal complex (pineal and parapineal organs) functioning as photoreceptive and/or endocrine organs.

“The pineal complex shares a common genetic and embryological basis with the lateral eyes, both derived from evaginations during the development of diencephalon. Despite being widely heralded as the ‘third eye’ in crown vertebrates, the nature of the pineal complex and its presumed visual capability in early vertebrates² remain unknown.

“Here we describe two pigmented features situated between the lateral eyes in two species of myllokunmingids, the earliest known fossil vertebrates (approximately 518  million years ago), and interpret these as pineal/parapineal organs.

“In both myllokunmingid species, the pineal complex contains abundant melanin-containing melanosomes identical to those in the retinal pigment epithelium in the lateral eyes, together with a distinctive, regularly ovoid structure interpreted as a lens.

“Our results indicate that the lateral eyes and pineal complex in myllokunmingids probably functioned as camera-type eyes capable of image formation. Thus, we propose that the four camera-type eyes represent an ancestral vertebrate character, corroborating hypotheses about the deep homology between the eyes and pineal complex.”

After an earlier phylogenetic analysis
in the large reptile tree (LRT, 2340 taxa) Cambrian Haikouichthys nested basal to Silurian Kalanaspis + Hemicyclaspis (with close-set tiny dorsal eyes) AND extant sturgeons (with large lateral eyes). All of these taxa have/had a flat ventral surface and a ventral oral cavity, rather than jaws, posterior to a ventral olfactory organ. They also had similar armor plates and a heterocercal tail.

Thus the new interpretation of Haikouicthys with four eyes
provides a novel solution to the phylogenetic split between Kalanaspis + Hemicyclaspis, (which lose or cover up the large lateral eyes) AND extant sturgeons, (which lose or cover up the tiny dorsal eyes).

A former hypothesis imagined a migration of eyeballs between the Silurian taxa and extant sturgeons.

Early Silurian
Thelodus is also related. It had large lateral eyes. If it had tiny dorsal eyes those are obscured in the fossil data presently available.

These taxa are all members of the clade
Acipenseriformes = sturgeons and their kin. Apparently this four-eyed trait is restricted to this clade, not vertebrates or chordates in general. See the LRT for details.

References
Let X-T et al (5co-authors) 2026. Four camera-type eyes in the earliest vertebrates from the Cambrian Period. Nature https://www.nature.com/articles/s41586-025-09966-0
Luo H, Hu and Shou 1997. New occurrence of the early Cambrian Chengjiang fauna from Haikou, Kunming, Yunnan province. Acta. Geol. Sin. 71, 97-104.
Shu D-G et al. (8 co-authors) 1999. Lower Cambrian vertebrates from China. Nature 402:42-46.

wiki/Haikouichthys

Acipenseriformes

https://pterosaurheresies.wordpress.com/2023/12/24/early-cambrian-haikouichthys-in-dorsal-view-and-the-naris-migration-problem/

Giant wombat Diprotodon ancestors in the LRT: Tillodon and Trogosus

This match took too long to surface.
– but click here: http://reptileevolution.com/diprotodon.htm to see how Tillodon and the giant wombat Diprotodon compare.

[Sorry. Still no imagery possible after the WordPress ‘upgrade’]
As you might imagine, that kinda takes the fun out of posting.

This makes stylinodons and taeniodonts marsupials
close to koalas. Wikipedia considers both of these eutherian taxa, not metatherian = close to koalas. This may be due to lingering faith in genomic analyses (that omit fossils), rather than trait analyses (that includes fossils).

https://en.wikipedia.org/wiki/Stylinodontidae

Taeniodonta (“banded teeth”)
is an extinct order of eutherian mammals, that lived in North America and Europe from the late Cretaceous (Maastrichtian) to the middle Eocene.

https://en.wikipedia.org/wiki/Taeniodonta

http://reptileevolution.com/reptile-tree.htm

Coombs 1983 looked at
‘Mammalian clawed herbivores’. She wrote, “Tillodonts are a Paleocene-Eocene clawed group known from Eurasia and North America. Most of the described forms are from North America, though Asiatic forms are becoming increasingly known (see, for example, Zhou et al., 1977), and the group may have arisen in Asia (Gingerich and Gunnell, 1979; Zhou and Wang, 1979). Much of the discussion surrounding tillodonts has regarded their affinities. A number of hypotheses of relationship have recently been proposed (Gazin, 1953; McKenna, 1975; Zhou et al., 1977; Szalay, 1977; Gingerich and Gunnell, 1979; Zhou and Wang, 1979). Several have suggested pantodont affinities, but none has yet received general acceptance.”

“Tillodont morphology gives no real suggestion of bipedal browsing (fig. 4) or scansorial habits, though the latter cannot be discounted in view of the well-developed hindclaws, large tail, and other characters. Force adaptations of the forelimb, which could be associated with digging or tearing, are present but less developed in general than in Titanoides or taeniodonts (see fig. 3). Existing evidence leads to a tentative suggestion of
digging or especially tearing, combined with pulling by the anterior teeth, as a method of obtaining fairly coarse plant food.”

Thank you for your patience
during this latest time-off from posting.

References
Coombs MC 1983. Large mammalian clawed herbivores: a comparative study. Transactions of the American Philosophical Society 73(7). 95pp.

A fourth Protorothyris specimen described as a µCT scan

Jenkins, Behlke and Sues 2025 applied
“µCT imaging to a previously undescribed specimen of Protorothyris archeri. Our segmentations revealed previously unknown anatomical details of this species, including enlarged dentary teeth, denticles on the parabasisphenoid, and the morphology of the internal surfaces of the dermatocranial elements. We included this specimen within a phylogenetic matrix of 177 operational taxonomic units and 628 morphological characters designed to examine the relationships of early-diverging amniotes, particularly stem reptiles. Bayesian and parsimony analyses reveal that there is still topological discordance among some clades (e.g. caseasaurs, varanopids), although these analyses confirm that ‘protorothyridids’ do not form a clade, consistent with other recent phylogenetic analyses.”

This new dorsoventrally crushed skull specimen,
UNSM PAL382180, was added to the large reptile tree (LRT, 2340 taxa) as the fourth Protorothyris tested. There is no topological discordance among clades in the LRT.

IMHO each of the four specimens
should be considered a separate species based on bone suture patterns and skull proportions. Only one, CM 8617, has been assigned to its own species, P. morani.

Please click on the above links to see images. My OS is now incompatible with WP and I am unable to upgrade probably due to that. I am also unable to upload images at present. Apologies.

The authors reported,
“We conducted phylogenetic analyses in both parsimony and Bayesian frameworks, and Seymouria baylorensis was chosen as the outgroup in both instances. We coded character, clock, and tree models after Sim~oes et al. (2022), which hypothesize a root age for the origin of amniotes to be 315–330 Ma (=mya) which is consistent with the known fossil record (Carroll & Baird 1972; Tuinen & Hadley 2004) and estimates from genomic datasets.”

The LRT narrows the origin of amniotes
= reptiles to Silvanerpeton as a last common ancestor in the Viséan 330–346 mya. A long list of basal reptiles, including Silvanerpeton, do not appear in the cladogram of Jenkins, Behlke and Sues (their figure 8). That list can be found in the LRT. Rather an invalidated arrangement is presented that nests dissimilar caseasaurs with synapsids, mesosaurs with millerettids and rib-gliders, like Weigeltisaurus, with aquatic taxa ike Claudiosaurus and Tangasaurus. These mistakes arise from borrowed cladograms borrowing from borrowed cladograms. More taxa are needed. See the LRT for a list.

To their credit,
the authors added 27 new taxa to their borrowed cladogram, but those were unable to provide the first dichotomy within the Reptilia, the one splitting Lepidosauromorpha from Archosauromorpha (link below), nor were they able to recover two separate origins for the diapsid skull morphology (link below).

https://pterosaurheresies.wordpress.com/2016/06/14/when-synapsids-and-diapsids-split/

https://pterosaurheresies.wordpress.com/2011/07/31/the-big-kahuna-the-reptilia-is-diphyletic/

References
Clark J and Carroll RL 1973. Romeriid Reptiles from the Lower Permian. Bulletin of the Museum of Comparative Zoology 144 (5): 353-408.
Jenkins KM, Behlke ADB and Sues H-D 2025. New anatomical details concerning the cranial structure of the early Permian stem reptile Protorothyris archeri revealed by µCT, with implications for the evolution of olfaction in reptiles. Palaeontology 68(6): e70038.
Price LI 1937. Two new Cotylosaurs from the Permian of Texas: Proceedings of the New England Zoological Club, v. 16, p. 97-102.

wiki/Protorothyris

The mammal subset of the LRT is at a stopping point

The synapsid/mammal subset
of the large reptile tree (LRT, 2340 taxa) has arrived at a stopping point = it’s done. At least, for now, it’s as finished as it can be.

That means the interrelationships recovered by the LRT seem to make sense.

That means sister taxa throughout appear to be the result of microevolution.

That means they look like each other, and look like they could have evolved one from another, without any odd disjointed pairings.

Unfortunately,
this subset of the LRT is also more fragile than any other similar subset of the LRT with 600+ taxa. Practically that means deletion of one mammal taxon can add many instances of ‘loss of resolution’ at several nodes in a wide spectrum up and down the taxon list while ironically reducing the number of MPTs overall. That is disheartening.

(Don’t ask me how or why that happens. It’s a puzzlement.)

So, that’s what we’re dealing with: a fragile conclusion.
Mammals are different than other chordate subsets. Mammals converge here and there. Some substitute a first premolar for a reduced and lost canine. Males differ from females. Juveniles have fewer teeth than adults. Placental reproduction appeared by convergence 5x at present.

It took 600+ 12-hour days of housekeeping
to fine tune 600+ mammals to the present fragile state of interrelationships. Every day I learned something new. Mistakes were made along the way, as the number of corrections rose to six then seven figures employing data that ranged from µCT scans in several views to century-old engravings without a dorsal or palatal view.

Thank you for your patience as I slogged through this quagmire.

So, it’s been a journey of frustration and elation
= discovery in a stepwise fashion. The LRT started in late 2010, so here at the start of 2026 the LRT has been a 16-year-long thesis project without an advisor and without hope or goal of achieving a degree. No $$. No fame. Just an interest in finding out.

This journey was made possible
by a quick access digital library of color-coded skulls and skeletons. The work could not have been started or finished without the many workers who posted data in library journals and on the web over the decades.

I was driven by
a vacuum (or several vacuums) that needed to be filled in this topic. While others focused on finding taxa, digging out fossils and studying a few skeletons in their lifetime, it was my task to take 2500 skeletons and arrange them in order and in vivo poses. Several times I discovered interrelationships that others overlooked. It was my task to sweep up the litter after others had published on various taxa.

I discovered no new taxa. My job was to join them together.

Other workers must have also experienced similar frustrations.
Perhaps that’s why academic workers stopped testing traits and passed this task over to genomic testing.

Unfortunately too often genomics did not recover similar interrelationships.
(Again, it’s a puzzlement.)

Unfortunately genomic interrelationships became widely accepted despite the obvious dicontinuities and its avoidance of fossil taxa.

The interrelationships recovered by the LRT
represent a hypothesis = stab at understanding how extinct and extant taxa are interrelated. The first stab was by Linneaus. Others have followed. All hypotheses require others to confirm, refute or modify their findings. The LRT is a contribution. It can be ignored or it can serve as a guide, a list of taxa that probably should be tested in future, smaller, more focused studies. I can’t imagine anyone else spending 16 years trying to figure out if badgers are small bears or bears are large badgers, etc. etc. etc.

All data = matrices for the LRT
are available online at FigShare.com. Look for ‘The ReptileEvolution.com project” in your keyword search.

Web pages now need to be repaired and updated at ReptileEvolution.com.

Ed note: I see six comments in the queue now. I will get to them over the next few days. Thank you for your comments, your patience and your interest in paleontology and systematics.