From the Ebel et al 2020 abstract:
“We reconstructed the acquisition of a fossorial lifestyle in 2813 lepidosaurs and assessed the skull roof compactness from microCT cross-sections in a representative subset (n = 99). We tested this and five macroscopic morphological traits for their convergent evolution. We found that fossoriality evolved independently in 54 lepidosaur lineages. Furthermore, a highly compact skull roof, small skull diameter, elongate cranium, and low length ratio of frontal and parietal were repeatedly acquired in concert with a fossorial lifestyle.”
Unfortunately the Ebel team relied on a genomic cladogram. By contrast the large reptile tree (LRT, 1772+ taxa, subset Fig. 2) found only 6 lepidosaur fossorial clades. Perhaps this is so because the LRT is a phenomic (trait-based) cladogram in which fossorial and legless Dibamus is a highly derived skink, alongside Bipes and Amphisbaena, not a basalmost squamate nesting alone. In the LRT, basalmost squamates have legs, fingers and toes, traits many derived squamates retain then lose.
Ebel et al 2020 abstract continues:
“We report a novel case of convergence that concerns lepidosaur diversity as a whole. Our findings further indicate an early evolution of fossorial modifications in the amphisbaenian ‘worm-lizards’ and support a fossorial origin for snakes.”
By contrast the LRT employs a wider gamut of taxa, including fossils, and finds a basal dichotomy at the genesis of snakes. One branch remained terrestrial while the other branch became fossorial. The morphologically weirdest burrowing snakes are the most derived. That’s how evolution works: from simple and plesiomorphic to bizarre and derived.
The latest LRT addition to snakes, Atractaspis, is a burrowing venomous snake arising from terrestrial snakes. That taxon would have been overlooked prior to addition to the LRT, so there may be other squamate burrowers not yet tested by the LRT.
The LRT finds only six clades
of burrowing squamates (Fig. 2) and an entirely different tree topology that includes fossils and protosquamates, and, need I say it, tritosaurs.
Ebel R, Müller J, Ramm T, Hipsley C and Amson E 2020. First evidence of convergent lifestyle signal in reptile skull roof micro anatomy. BMC Biology 18…185. https://doi.org/10.1186/s12915-020-00908-y