*3.1. Phylogenetic Relationships*

Our phylogenetic ML (Figure 2) tree inferred using fragments of four mitochondrial and one nuclear gene is congruen<sup>t</sup> with other studies [6,11,12,51,52], except that it recovered the monophyly of *Proctoporus* and that the recently described genus *Wilsonosaura* is nested within the *Proctoporus* lineage (Bootstrap: 76). However, trees using only mitochondrial (Figure S1) and nuclear (Figure S2) markers are not congruent. Our ML analyses using the full dataset recovered the polyphyly of genera *Cercosaura* and *Pholidobolus*. *Pholidobolus anomalus* and *Cercosaura* sp. (MUBI 14515) were nested within *Cercosaura*, whereas *Cercosaura manicata boliviana* was nested with *Potamites* and *Selvasaura*. Likewise, our mixed ML analyses recovered the monophyly of the genera: *Anadia* (98), *Andinosaura* (99), *Cercosaura* (including *C. anomala*, 85), *Dendrosauridion* (99), *Echinosaura* (100), *Gelanesaurus* (100), *Neusticurus* (100), *Macropholidus* (100), *Pholidobolus* (99), *Placosoma* (100), *Potamites* (100), *Proctoporus* (76), *Selvasaura* (100), and *Riama* (100). *Euspondylus* and *Rheosaurus* (both with a single known species) were recovered as independent lineages. Our phylogeny inferred using the full dataset is congruen<sup>t</sup> with other studies, and may better reflect the evolutionary history of the lineages [53].

**Figure 2.** Maximum likelihood tree (log likelihood = −44,674.306, ultrafast boostrap = 1000) showing the phylogenetic relationships of *Cercosaura*, *Pholidobolus anomalus,* and other gymnophthalmid lizards. The numbers next to the nodes are bootstrap values. The analysis was constructed from a concatenated dataset of 2167 bp of four mitochondrial genes (12S, 16S, Cytb, ND4) and a nuclear gene (c-mos). In blue are the samples obtained in this study.

The uncorrected p-distance of the 16S gene between *Pholidobolus anomalus* and other taxa were: *Cercosaura manicata* (3.7–5.6%),*Cercosaura* sp. (6%), and*C. ocellata* (9.5%).*Cercosaura* sp. has a distance of 3.2% with respect to its sister species (*Cercosaura doanae*). Genetic distances among *Cercosaura manicata boliviana* and any other species of *Cercosaura* exceeded 4.5% (Table 3). The uncorrected p-distances between *Cercosaura* sp. and *C. doanae* in 12S (5.2%), ND4 (9.3%), c-mos (0.7%) genes were always greater than the intraspecific distance of *C. manicata* in 12S (2.2%), ND4 (9.0%), and c-mos (0.0%). These uncorrected

p-distance between *Cercosaura* sp. and *C. doanae* are outside the intraspecific range, and within the interspecific range for species of gymnophthalmid lizards [11].

Based on our molecular analyses (Figure 2, Tables 3 and 4), we conclude (1) that *P. anomalus* needs to be redescribed and allocated to the genus *Cercosaura*, including designating a neotype; (2) *Cercosaura* sp. (MUBI 14515) is a new species; and (3) *Cercosaura manicata boliviana* is a valid species, the sister lineage of the semiaquatic lizards of the genus *Potamites* [20], but considered here as *incertae sedis*.
