Abstract
We generated a phylogeny for Caucasian rock lizards (Darevskia), and included six other families of true lizards (Lacertini), based on complete mitochondrial genome analysis. Next-generation sequencing (NGS) of genomic DNA was used to obtain 16 new mitogenomes of Darevskia. These, along with 35 sequences downloaded from GenBank: genera Darevskia, Zootoca, Podarcis, Phoenicolacerta, Takydromus, Lacerta, and Eremias-were used in the analysis. All four analytical methods (Bayesian Inference, BI; Maximum Likelihood, ML; Maximum Parsimony, MP; and Neighbor-Joining, NJ) showed almost congruent intra-generic topologies for Darevskia and other lizard genera. However, ML and NJ methods on one side, and BI and MP methods on the other harvested conflicting phylogenies. The ML/NJ topology supports earlier published separation of Darevskia into three mitochondrial clades (Murphy, Fu, Macculloch, Darevsky, and Kupinova, 2000), but BI and MP topologies support that the basal branching occurred between D. parvula from the western Lesser Caucasus and the rest of Darevskia. All topologies altered the phylogenetic position of some individual species, including D. daghestanica, D. derjugini, and D. chlorogaster. Reanalysis after excluding four saturated genes from the data set, and excluding genus Eremias gives fully convergent topologies. The most basal branching for true lizards was between Far Eastern Takydromus and the Western Eurasian genera (BI). Comparing phylogenetic performance of individual genes relative to whole mitogenome data, concatenated 16S RNA (the least saturated gene in our analyses) and Cytochrome b genes generate a robust phylogeny that is fully congruent with that based on the complete mitogenome.