Skip to main content

A reassessment of the enigmatic diapsid Paliguana whitei and the early history of Lepidosauromorpha

Cite this dataset

Ford, David Paul et al. (2021). A reassessment of the enigmatic diapsid Paliguana whitei and the early history of Lepidosauromorpha [Dataset]. Dryad.


Lepidosaurs include lizards, snakes, amphisbaenians and the tuatara, comprising a highly speciose evolutionary radiation with widely varying anatomical traits. Their stem-lineage originated by the late middle Permian 259 million years ago, but its early fossil record is poorly documented, obscuring the origins of key anatomical and functional traits of the group. Paliguana whitei, from the Early Triassic of South Africa, is an enigmatic fossil species with potential to provide information on this. However, its anatomy and phylogenetic affinities remain highly uncertain, and have been debated since its discovery more than 100 years ago. We present microtomographic 3D imaging of the cranial anatomy of Paliguana whitei that clarifies these uncertainties, providing strong evidence for lepidosauromorph affinities based on the structure of the temporal region and the implantation of marginal dentition. Phylogenetic analysis including these new data recovers Paliguana as the earliest known stem-lepidosaur, within a long-lived group of early-diverging lepidosauromorphs that persisted to at least the Middle Jurassic. Our results provide insights into cranial evolution on the lepidosaur stem-lineage, confirming that characteristics of pleurodont dental implantation evolved early on the lepidosaur stem-lineage. In contrast, key functional traits related to hearing (quadrate conch) and feeding (streptostyly) evolved later in the lepidosaur crown-group.


We included new data on the anatomy of Paliguana to the phylogenetic matrix of Griffiths et al. 2021, which was extensively modified from Simões et al. 2018. In addition to re-scoring the anatomy of Paliguana, we also revised the characters concerning the implantation of marginal dentition. Multistate characters 212 and 213 of the original character–taxon matrix were removed and replaced by six new binary characters (ca 376–381), which recognise independent variation in the morphology of tooth implantation among reptiles. We also added a recently reported species to the data-set (Vellbergia bartholomaei) (refer to electronic supplementary material, data S5 for details on these changes, together with a revised character list).

The modified character-taxon matrix was analysed using a non-time calibrated Bayesian Mkv model in MrBayes v.3.2.5 (Ronquist et al. 2012). The ‘standard’ datatype was used for morphological data, and the coding was set at ‘variable’ (that is, only variable characters being sampled) with rates set to a gamma distribution. The Mkv model analysis was performed with two runs of four chains each for 50 million generations, with tree sampling every 500th generation and a 25% burn-in. The average effective sample size for the two parameters (tree length and shape of gamma distribution) were in excess of 50,000, and the average potential scale reduction factor was 1.000, with both chains demonstrating good convergence (average standard deviation of split frequencies 0.0028). Our results were expressed in a 50% majority rule tree (halfcompat command in MrBayes) and in a maximum clade credibility tree (MCC). The MCC tree is a single tree in the posterior sample with the maximum sum of posterior clade probabilities across all the constituent bifurcations (Heled and Bouckarert 2013). We used TreeAnnotator (v1.10.4 Rambaut and Drummond 2002-2018) to recover the MCC from a combination of the posterior tree files of the two runs from MrBayes output (.t files), less the 25% burn-off, producing a total of 150,000 trees. The script for the Bayesian analysis is provided in Part 6 of Supplementary Data. Optimisation of the 50% majority rule tree and the MCC tree were performed for both ACCTRAN and DELTRAN in PAUP v.4.0a, build 169 (Swofford, D.L. 2002)(refer to electronic supplementary material, data S2 and S3 for extended details).


National Research Foundation of South Africa (NRF), Award: 98800

National Research Foundation of South Africa (NRF), Award: 118794

National Research Foundation of South Africa (NRF), Award: 98800