The Reptilian Stem

John Merck

Amniote phylogeny, after Ford and Benson, 2019

Synapsida is only half of Amniota. The other half - "reptiles" - contains living turtles, squamates, Sphenodon, crocodylians, and birds, and their fossil relatives. Alas, we have no consensus on:

What we do know:

The Last Word on "reptile" phylogeny is the that of Ford and Benson, 2019, which GEOL431 provisionally follows. Where appropriate, we will look at the history of phylogenetic hypotheses of certain important taxa. We do impose one caveat:

Reptilia or Sauropsida? Both terms have been used for the sister-taxon of Synapsida and neither has a "clean title" to the clade because of issues that haunt us from the early, revolutionary days of cladistics when:

Ford and Benson favor Reptilia, however this name has at least four phylogenetic definitions in the literature (Modesto and Anderson, 2004.) Respecting priority, GEOL431 goes with the definition of Gauthier et al., 1988 which is:

Given Gauthier et al.'s convictions that:

...this yielded a nice amniote dichotomy between Synapsida and Reptilia. Alas, since then there has been no consensus on the position of turtles, with molecular phylogenies frequently placing them within Diapsida. As hypotheses of phylogeny change, so does the makeup of Reptilia. Worse, in some cases it is redundant with other common group names like "Sauria."

Sauropsida is defined by Laurin and Reisz, 1995 as:

This anchors the definition on Mesosauridae - a basal group never appearing on the synapsid side. But alas, Ford and Benson place Mesosauridae, along with other members of Parareptilia, far up the tree, within Diapsid, making Laurin and Reisz' definition useless to us as well.

So we punt!

For now and until a proper definition appears, the sister taxon of Synapsida will be referred to as Sauropsida, a name that is free of cultural association with traditional definitions of "reptiles."

Palatal views of pareiasaurs Deltavjatia and Scutosaurus from the U. C. Berkeley History of Life.


Sauropsid synapomorphies include:

Among living members we also note:

The Early Permian captorhinid Eocaptorhinus laticeps from Paleocritti

Basal sauropsids

Includes Hylonomus lyelli, a Joggins tree-stump victim and the earliest well-known amniote.

Diversity: Diapsida comprises the majority of Eureptilia, but we should note some stem-diapsids.

Diapsida: (Late Carboniferous to Quaternary.) Diapsids are among the first amniotes of the Late Carboniferous, however during the Paleozoic they were a minor component of the terrestrial fauna. That changed during the Mesozoic, when they achieved ecological dominance. Modern diapsids include : Their fossil members have included the largest/scariest land animals ever, however at their root, they were adapted to life as small, fast-moving predators of small prey items.

Petrolacosaurus kansensis from Reisz, 1981
Synapomorphies are visible in Petrolacosaurus (Right - Late Carboniferous) the earliest and among the most primitive diapsids:

Petrolacosaurus kansensis from Reisz, 1981
There are two general patterns here:

Diapsid diversity: Diapsids, as generally conceived, are an apomorphy-based group - the sauropsids that ancestrally possess the diapsid pattern of temporal fenestration, They were among the first groups to be studied with the methods of cladistics (See Gauthier, 1984 and Benton 1985). Between them on the saurian stem is an array of fossil stem saurians.

Petrolacosaurus kansensis

Araeoscelidia: (Late Carboniferous - Early Permian). Small slender animals characterized by:

Araeoscelidians are specialized either as arboreal or aquatic animals. Remarkable more for their plesiomorphies, including retention of:

At least one member, Araeoscelis secondarily closes the infratemporal fenestra, prefiguring the widespread modifications to the diapsid pattern seen in Diapsida.

Orovenator mayorum from Ford and Benson, 2018. (Scale = 1 cm.)

Mesenosaurus efremovi, a proper varanopid (Scale = 2 cm.)

Varanopidae: (Late Carboniferous - Middle Permian). Dr. Holtz introduced these in his discussion of basal synapsids, where they have traditionally been thought to reside. Ford and Benson's, 2018 redescription of the poorly-known diapsid Orovenator mayorum (right) revealed many synapomorphies of it and the varanopid "synapsids," causing its inclusion in existing matrices to pull varanopids into Diapsida despite the fact that no varanopid has an open supratemporal fenestra. Results:

Case study in character polarization: When moved into Diapsida, many of the characters that had made Varanopids seem derived with respect to other synapsids suddenly seem plesiomorphic. While others including: Become apomorphic.

Milleretta from
University of California Museum of Paleontology


(Early Permian - Late Triassic) This group has historically included reptilian-grade organisms closer to Eureptilia that did not have living members. The arrival of phylogenetic systematics allowed it to be phylogenetically defined as monophyletic, however its precise membership has been fuzzy because of uncertainties about the position of turtles.


Milleretta from
University of California Museum of Paleontology


(Early Permian - Late Triassic) This group has historically included reptilian-grade organisms closer to Eureptilia that did not have living members. The arrival of phylogenetic systematics allowed it to be phylogenetically defined as monophyletic, however its precise membership has been fuzzy because of uncertainties about the position of turtles.


Parareptilian diversity. Great diversity during the Late Paleozoic and earliest Mesozoic.

The basal parareptilian Mesosaurus from
Mesosauridae: (Early Permian) Mesosaurus, Braziliosaurus, and Stereosternum. Inhabiting the incipient basin of the South Atlantic, their fossils are from southern South America and Africa. Mesosaur distribution was among the data of Alfred Wegener supporting continental motion. The first amniotes with clear adaptations to aquatic life. Known from many well-preserved specimens, yet phylogenetically enigmatic. Characteristics:

Mesosaur issues:

Millerina rubidgei from Paleofile
Millerettidae: (Permian) Small (20 - 40 cm.) probably ecological generalists with simple peg-like teeth. Synapomorphies:

Lanthanosuchus from U. C. Berkeley Museum of Paleontology
Lanthanosuchoidea: (Early - Late Permian) Small (20 - 40 cm.) probably ecological generalists with simple peg-like teeth. Contains: Synapomorphies:

Bolosaurus from Wikimedia Commons
Bolosauridae: (Early - Middle Permian) Small (20 - 40 cm.) probably ecological generalists to herbivores with complex teeth. Contains the earliest known facultative biped - Eudibamus cursorius. Although bolosaurids had been known through the 20th century, it was Berman et. al, 2000 who demonstrated their parareptilian nature phylogenetically. Synapomorphies:

Hypsognathus from Stuart Sumida's BIOL622 - California State University San Bernardino
Procolophonoidea: (Late Permian - Late Triassic) Small (20 - 40 cm.) probably ecological generalists to herbivores with bulbous blunt teeth. Typically stocky with short tails, Many members sport cranial ornamentation and armor of dermal scutes. Contains: Synapomorphies: In addition, procolophonoids tend to expand the orbit posteriorly into the temporal region, however in contrast to many more basal parareptiles, there is no temporal fenestration. In most, the cheeks are broadly flaring.

Bashkyroleter mesensis (a) and Macroleter poezicus (b) from Wikipedia.
Nycteroleteridae: (Middle Permian - Late Permian) Small (20 - 40 cm.) Mostly known from skulls, however Emeroleter shows a large skull on a relatively gracile bodies and limbs. These include the parareptiles with the most extensive embayments of the posterior cheek margin and small slender stapes - strongly indicative of impedance-matching ears. Their phylogeny is poorly resolved but Tsuji et al. 2012 find Macroleter to be the most basal. Synapomorphies:

Scutosaurus karpinskii from
Pareiasauria: (Late Permian). Medium - big, squat, ugly, bumpy. These were major herbivores in the Late Permian world. They competed ecologically with therapsid dicynodonts, and were hunted by therapsid gorgonopsians.

General trends:

Thus, everything points to these creatures having been large, slow moving armored herbivores with extensive digestive systems.


Neodiapsida: (Early Permian - Quaternary) First coined in the 1980s to encompass diapsids closer to Sauria than to Araeoscelidia, its definition has been contentious. The last word belongs to Ford and Benson, 2019, who apply it to the sister taxon of Parareptilia. In Neodiapsida, the diapsid morphotype and fenestration pattern stabilizes around something like the traditional concept of a "diapsid."


Lanthanolania ivakhnenkoi (Middle Permian) from Modesto and Reisz, 2003
Lanthanolania reflects these trend by: Note: The posterior skull elements in the reconstruction are entirely speculative!

Megalancosaurus preonensis (Late Triassic) from Vertebrate Paleontology at Insubria University
Drepanosauromorpha: (Late Triassic) Odd arborealists - convergent on certain saurians, but lacking key saurian synapomorphies like evidence for an impedance-matching ear: The oversized ungual phalanges of some drepanosaurids (e.g. Drepanosaurus) has invited speculation about their function. Jenkins et al., 2020 determined that larger unguals closely resemble those of "hook-and-pull" diggers like the living tamandua and silky anteater - creatures that combine arboreality with the ability to dig into ant and termite nests.

Coelurosauravus elivensis (Late Permian - Early Triassic)
Wiegeltisauridae: the gliders:

Claudiosaurus germaini (Late Permian) from Paleofile
Claudiosaurus: The lower temporal bar is definitely incomplete in Claudiosaurus (right). In life, its place would be occupied by an unossified ligament. Various diapsids would eventually reossify that ligament in various ways, but from this point forward, we deal with diapsids who are descended from creatures whose infratemporal fenestrae have been transformed to broad embayments in the lower margin of their cheeks.

Claudiosaurus was a limb-propelled swimmer with a relatively long neck. When first described it was called a "plesiosaur ancestor." That's probably wrong, but it demonstrates that even in the Permian, diapsids displayed a tendency opportunistically to evolve aquatic forms. Claudiosaurus was fresh-water aquatic, however, not marine.

Youngina capensis (Late Permian) from Wikipedia
"Younginiformes:" (Late Permian) Our first relatively complete stem-neodiapsid belong to this probably paraphyletic group. Members of this grade contain the last diapsids to retain a fully-enclosed infratemporal fenestra, such as Youngina (right). Youngina was terrestral. Others, such as Hovasaurus were fresh-water aquatic.

Note: It's not clear that all members of this grade had complete lower temporal cheek bars.

Potential synapomorphy of Younginiformes:

Potential synapomorphies of "younginiformes" and Sauria:

Phylogeny: The more derived stem saurians display considerable convergence on other forms, leading to a diverse range of phylogeny hypotheses. Some highlights:


(Late Permian - Rec.) The most recent common ancestor of living lizards, Sphenodon, crocodylians, and birds.

Here, for once, in a pleasing cladogram is the arrangement of Gauthier, 1984, in which two crown-groups were recognized in Sauria:

Anchored on these, he defined total-groups that contained them: This setup was embraced so firmly that even though there were technical difficulties with his definition of Lepidosauromorpha, (ask me sometime) most systematists continue to use the term, and so shall we.


Stay tuned.

Additional reading: