DaSilva et al. 2018
bring us a new perspective on snake evolution that employs molecules, physical traits, embryos, fossils, CT scans… a huge amount of data and labor… perhaps all for nought because they excluded so many pre-snake taxa (Fig. 2). And their results do not produce a gradual accumulation of derived traits (Fig. 1), even when they omit the mosasaur skulls listed at their base of snakes. Here I added that missing mosasaur skull.
The DaSilva et al. abstract:
“The ecological origin of snakes remains amongst the most controversial topics in evolution, with three competing hypotheses: fossorial; marine; or terrestrial. Here we use a geometric morphometric approach integrating ecological, phylogenetic, paleontological, and developmental data for building models of skull shape and size evolution and developmental rate changes in squamates. Our large-scale data reveal that whereas the most recent common ancestor of crown snakes had a small skull with a shape undeniably adapted for fossoriality, all snakes plus their sister group derive from a surface-terrestrial form with non-fossorial behavior, thus redirecting the debate toward an underexplored evolutionary scenario. Our comprehensive heterochrony analyses further indicate that snakes later evolved novel craniofacial specializations through global acceleration of skull development. These results highlight the importance of the interplay between natural selection and developmental processes in snake origin and diversification, leading first to invasion of a new habitat and then to subsequent ecological radiations.” Fossorial = burrowing.
The DaSilva et al. Supplementary Data reports:
“To include a large dataset of squamate specimens, including extant, fossil, and embryonic taxa (see details below as well as Fig. 1 (main text) and Supplementary Fig. 1), we used a composite phylogenetic hypothesis based on the most recent molecular as well as combined molecular and morphological studies on squamate evolution.”
As readers know by now,
molecular data fails at large phylogenetic distances. It produces false positives. Even so, their large number of physical traits (691 morphological characters and 46 genes) should have given them a good cladogram… unless they omitted huge swaths of taxa.
Which is what they did (Fig. 2).
Even though they used fossil and embryological data,
their results do not produce a gradual accumulation of traits (Fig. 1). Nor do they employ appropriate outgroup taxa, either for squamates or for snakes (Fig. 2).
Without these key transitional taxa,
the authors have no idea what the basalmost squamates and snakes should look like. Here’s what the large reptile tree (LRT, 1152 taxa) recovered (Fig. 1):
If nothing else,
I hope readers gain a critical and skeptical eye toward published material. Sometimes it’s not what they say, but what they omit that spoils their results.
The LRT is a good base
to begin more focused studies in tetrapod evolution. It covers virtually all the possible candidates so workers can have high confidence that their more focused studies include relevant taxa and exclude irrelevant taxa.
DaSilva FO et al. (7 co-authors) 2018. The ecological origins of snakes as revealed by skull evolution. Nature.com/Nature Communications (2018)9:376 1–11. DOI: 10.1038/s41467-017-02788-3 pdf