Sphodrosaurus: here identified as a stem soft shell turtle

Known for decades as an enigma,
Sphodrosaurus pennsylvanicus nests here more primitive than Odontochelys and sheds light on the pareiasaur-to-soft shell turtle transition.

FIgure 1. Partial reconstruction of Sphodrosaurus based on tracings in figure 2.

FIgure 1. Partial reconstruction of Sphodrosaurus based on tracings in figure 2. This turtle is more basal than Odontochelys. Lots of loose parts here and no attempt was made to reassemble the manus or pes.

Colbert 1960
described Sphodrosaurus pennsylvanicus (Fig. 1) as, “A new Triassic procolophonid from Pennsylvania” based on North Museum No. 2321, a natural mold in ventral view of a partial skeleton (Fig. 2) resembling Hypsognathus and located less than a mile from the skull of this genus. In the large reptile tree (LRT, 1308 taxa; subset Fig. 3) Sphodrosaurus nests between the tiny pareiasaur Sclerosaurus (Fig. 4) and the basal soft-shell turtle (known only from skull material) Arganaceras.

Based on the appearance of a shape in the mudstone
beneath the ribs (in ventral view, thus dorsal in life), Sphodrosaurus appears to be (by observation and phylogenetic bracketing) the first taxon to have some sort of soft carapace without developing any sort of expanded ribs or any sort of plastron. Thus it informs on the likely appearance of the currently missing post-crania of Arganaceras. Some loose gastralia-like ossifications (in cyan) are apparent. These are plastron precursors (again, based on phylogenetic bracketing). These inform on a previously unknown genesis for the plastron in soft shell turtles. Sclerosaurus lacks them. Odontochelys has massive plastron elements.

Figure 2. Sphodrosaurus in situ with colors added to bones and possible soft carapace impression.

Figure 2. Sphodrosaurus in situ, ventral view, with colors added to bones and possible soft carapace impression overlooked originally. Colbert  1960 tracing also shown here.

Sclerosaurus nests with procolophonids, but that nesting is based on taxon exclusion. Sphodrosaurus is very similar to Sclerosaurus, but a little more derived toward the soft shell turtles.

Figure 3. Sphodrosaurus nests with other soft-shell turtles arising from pareiasaurs.

Figure 3. Sphodrosaurus nests with other soft-shell turtles arising from pareiasaurs without invoking the carapace.

Sphodrosaurus pennsylvanicus (Colbert 1960; North Museum No. 2321; Newark Supergroup, latest Carnian, Late Triassic). Distinct from Sclerosaurus, the femur is longer, the coracoid is smaller. The antebrachium is longer. As in Trionyx, pedal digit 5 is gracile. The specimen was found in mudstones. Note the wide, flat torso, the tall, slender scapula, sigmoidal femur and long-clawed toes… all turtle traits.

Colbert reported,
“The skull seems to have been unusually large in comparison to the size of the postcranial skeleton. The posterior portion of the skull is produced back into a “frill,” as is common in the advanced procolophonids, this frill covering about five cervical vertebrae. There are 25 presacral vertebrae, to which are articulated widely spreading holocephalous ribs. The scapula is rather slender, the ilium seemingly deep. The pubis and ischium are platelike bones, the former being proximally constricted and distally expanded. The hind limbs are large, the extended limb being approximately equal in length to the total length of the presacral series of vertebrae. In total length and in each of its component sections the linear dimensions in the hind limb are about double those in the fore limb. The metatarsals are rather slender, and long. The ungual phalanges of the pes are large, pointed claws.”

Figure 4. Sclerosaurus reconstructed.

Figure 4. Sclerosaurus reconstructed.

Colbert continues,
“Perhaps the most striking differences between this form and the established genera of procolophonids are in the great length and robust size of the hind limb in the Pennsylvania specimen, and the long, sharp claws of the pes. Such characters might lead one to doubt the true procolophonid relationships of Sphodrosaurus, but other characters, such as size, the obviously large skull, the extension of the back of the skull in a sort of frill over the cervical region, the evidently broad vertebral neural arches (as indicated by the separation of the heads of the ribs), and the holocephalous, flaring ribs, are all characters that point to procolophonid affinities for Sphodrosaurus.”

The following paper
was discovered after the reconstruction and phylogenetic analysis were made:

Sues, Baird and Olsen 1993 reexamined Sphodrosaurus
and determined that the specimen was not a procolophonid, but some sort of diapsid or neodiapsid. They note, Baird (1986) suggested rhynchosaurine affinities. They also note “This combination of characters has not been found in any other known diapsid.” 

The authors note
the preservation of the posterior mandibles, rather than a set of dorsal skull bones as Colbert reported. They failed to see the detached retroarticular process. The cervicals and anterior dorsals have a ventral ridge. So do soft-shell turtles, but the authors did not make that connection. What they identify as extremely long cervicals parallel to the spine and apparently coosified are interpreted here as clavicles. They remarked on the “great width of the trunk region,” as in pareiasaurs and turtles, but the authors did not make that connection. They note the scapula has a “slender  blade”, as do turtles, but the authors did not make that connection. They note the femur is sigmoidal, as in turtles, but the authors did not make that connection.

The authors conclude,
“The mode of preservation of the holotype and only known specimen of Sphodorsaurus pennsylvanicus leaves very few anatomical features for assessing its phylogenetic position.” This is true, but phylogenetic analysis over a wide gamut of potential candidates leaves no doubt in the LRT about where this specimen nests, based on the characters that are visible. There is no mention of pareiasaurs or turtles in the Sues, Baird, Olsen 1993 paper.

As in many enigma taxa studied here,
the solution to their nesting problem appears whenever the enigma taxon is permitted to be tested against a wide gamut of taxa. This minimizes initial bias and lets the software do what it was intended to do… keep human preconceptions from interfering in a cold-blooded scientific process.

Added later the same afternoon
Rice et al. 2016 report: “We show that plastron development begins at developmental stage 15 when osteochondrogenic mesenchyme forms condensates for each plastron bone at the lateral edges of the ventral mesenchyme.” In this way ontogenesis recapitulates the phylogenesis demonstrated by Sphodrosaurus.

Colbert EH 1960. A new Triassic procolophonid from Pennsylvania. American Museum Novitates 2022:1–19.
Rice R, Kallonen A, Cebra-Thomas J and Gilbert SF 2016. Development of the turtle plastron, the order-defining skeletal structure. PNAS 113 (19):5317–5322.
Sues H-D, Baird D and Olsen PE 1993. Redescription of Sphodrosaurus pennsylvanicus Colbert, 1960 (Reptilia) and a Reassessment of its Affinities. Annals of Carnegie Museum 62(3):245-253


North Museum of Nature and Science
Franklin and Marshall College
400 College Avenue
Lancaster, PA 17603


Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Connecting to %s

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