Testudines (Panpleurodira – Pancryptodira)
Unfortunately Marjanovic relies on tradition when he splits turtles into pleurodiran (side-neck) and cryptodiran (hidden-neck) clades. He reports, “With one short series of exceptions (Gaffney et al., 2006, 2007; Gaffney and Jenkins, 2010), all treatments of Mesozoic turtle phylogeny from the 21st century have consistently found Proterochersis and all other turtles older than Late Jurassic to lie outside the crown group. The oldest known securely dated crown-group turtle is thus the mid-late Oxfordian (158 Ma) stem-panpleurodire Caribemys. The observed absence of cryptodires is likely real; combining this with more rootward Middle and Early Jurassic stem turtles from other continents, I suggest a hard maximum age of 175 Ma based on the beginning of the Middle Jurassic (174.1 ± 1.0 Ma ago: ICS).”
Neither sea turtles nor soft-shell turtles hide their head within their carapace, nor could their ancestors do so. In the large reptile tree (LRT, 1630+ taxa; Fig. 1) the basal dichotomy between soft shell and hard shell turtles extends back to small horned pareiasaurs from the Latest Permian (255 mya). Thus the crown group of all living turtles also includes all extinct turtles. Cryptodires and pleurodires appeared later, both within the hardshell clade, timed as noted above.
Lepidosauria (Rhynchocephalia + Squamata)
Marjanovic reports, “The minimum age of this calibration, given as 238 Ma, has to be slightly revised to 244 Ma (hard) based on Megachirella, the oldest known stem-squamate, which is older than the oldest known rhynchocephalian (238–240 Ma). An Early Triassic or perhaps Late Permian maximum age seems reasonable, but, given the rarity of stem-lepidosauromorphs and of Permian diapsids in general, I rather propose to use the ecologically similar small amniotes of Richards Spur (289 ± 0.68 Ma, see Node 107) to support a soft maximum age of 290 Ma.”
In the LRT the last common ancestors of rhynchocephalians + squamates (Fig. 2) include the basal rhynchocephalian (not stem-squamate) Megachirella (earliest Middle Triassic, 244 mya) and the earlier Palaegama (Late Permian). A proximal outgroup taxon is Tridentinosaurus (Earliest Permian, 295mya) approximating Marjanovic’s proposal.
Toxicofera (Pythonomorpha + Anguimorpha including Iguanomorpha)
Marjanovic reports, “I agree with Irisarri et al. (2017) in not assigning a maximum age other than that for Node 125 (Lepidosauria, see above).”
In the LRT Toxicofera is a junior synonym for Squamata (Fig. 2). The basalmost squamate taxon in the LRT is Euposaurus (Late Jurassic, Kimmeridgian, 155 mya). An Early Permian outgroup taxon, MNC-TA1045 (Spindler 2017) in a traditionally unrecognized clade, Protosquamata, which includes extinct taxa only. Lacertulus (Late Permian, not mentioned by Marjanovic) is a basal taxon.
Iguania (Chamaeleonformes + Iguanoidea)
Marjanovic reports, “I cannot assign a maximum age other than that for Node 125.” (See above).
In the LRT Euposaurus (overlooked by Marjanovic, and see above, Fig. 2), is the basalmost member of the Iguania and Squamata.
Marjanovic D 2019. Recalibrating the transcriptomic timetree of jawed vertebrates.
bioRxiv 2019.12.19.882829 (preprint)