Chile has a high level of endemism of reptiles, which currently reaches 60% (^{Ruiz de Gamboa 2020}). Most of the endemic species correspond to lizards of the genus Liolaemus Wiegmann, 1834 (one of the most diverse genus of reptiles of the world; ^{Abdala et al. 2021}), but there are also representatives of other genera shared with other South American countries. Among them is the genus Pristidactylus Fitzinger, 1843, whichcurrently comprises 10 species distributed in south-central Chile (^{Garín et al. 2020}) and Argentina (^{Morando et al. 2015}).

Like other genera of the same family (Leiosauridae), lizards of the genus Pristidactylus are characterized by stout bodies and legs, robust heads, and short tails that usually do not autotomize (^{Cei 1986}). Traditionally, the species of this genus have been divided into two groups based on morphological and ecological (habitats) characteristics and geographic distribution: the Chilean group (four species) and the Argentine one (six species) (^{Etheridge & Williams 1985}; ^{Cei et al. 2001}, ²⁰⁰⁴).

There are several phylogenetic studies, using morphological and/or molecular characters, where Pristidactylus and other genera of leiosaurids have been included (reviewed in ^{Morando et al. 2015}). These studies have included a variable number of species of the genus and have recovered it as monophyletic (^{Frost et al. 2001}; ^{Abdala et al. 2009}, in some of their morphological analysis) or polyphyletic (^{Abdala et al. 2009}, in their supertree combining morphological and molecular data; ^{Pyron et al. 2013}; ^{Morando et al. 2015}).

The most complete phylogenetic studies of the genus to date, in terms of species and number of genes (mitochondrial and nuclear), are ^{Morando et al. (2015}) and ^{Femenias et al. (2020}). One of the main results of those studies is that P. torquatus (Philippi, 1861) does not group with the rest of the species of the genus. However, both studies included almost exclusively species of Pristidactylus from Argentina and only included P. torquatus among the Chilean representatives. Among the four species from Chile, P. torquatus is the one with the widest geographic distribution; the other three species, P. alvaroi (Donoso Barros, 1974), P. valeriae (Donoso Barros, 1966) and P. volcanensis Lamborot and Díaz 1987, have much more restricted distributions in central Chile (^{Núñez & Urra 2016}; ^{Garín et al. 2020}) (Fig. 1). In addition, the quality of the habitats has decreased or deteriorated, which is why these three species are considered Endangered by Chilean legislation (^{Garín et al. 2020}).

FIGURE 1 Adult individuals of the Pristidactylus species from central Chile and sampling sites. A. Pristidactylus alvaroi. B. Pristidactylus volcanensis. C. Pristidactylus valeriae. D. Location of sampling sites (see details in Table 1). / Individuos adultos de las especies de Pristidactylus de Chile central y sitios de muestreo. A. Pristidactylus alvaroi. B. Pristidactylus volcanensis. C. Pristidactylus valeriae. D. Ubicación de los sitios de muestreo (ver detalles en la Tabla 1). 

The results of ^{Morando et al. (2015}) and ^{Femenias et al. (2020}) ratify that the genus Pristidactylus is polyphyletic and suggest that the Chilean and Argentine species correspond to lineages that have evolved separately, on opposite sides of the Andes. Previous phylogenetic studies that have included part or all of the Chilean species (using only morphological characters for the three from central Chile) have not recovered the reciprocal monophyly of these two lineages (^{Frost et al. 2001}; ^{Abdala et al. 2009}). Therefore, the objective of this study is to assess the phylogenetic position of the four Chilean species of Pristidactylus with sequences of the mitochondrial gene cytochrome b.

We sampled two individuals from one locality of each of the four Chilean species: P. alvaroi, P. torquatus, P. valeriae and P. volcanensis (Fig. 1; Table 1). In the case of these last two species, the individuals come from their respective type localities. Oral mucosa samples were obtained with swabs Copan 516CS01 to extract DNA. The sampled individuals were photographed and released at the same capture site. DNA was extracted with the kit ReliaPrep^TM gDNA Tissue Miniprep System (Promega, Madison, WI), following the manufacturer’s instructions. A fragment of the mitochondrial gene cytochrome b (cytb) was obtained. The primers and PCR protocols to obtain that fragment are found in ^{Morando et al. (2015}) and references therein. Sequences were edited with Bioedit v7.1.3 (^{Hall 1999}) and then aligned with Muscle (^{Edgar 2004}). The eight sequences of the four species were deposited in GenBank under the numbers ON787825- ON787832.

We inferred the phylogenetic relationships of the four Chilean species of Pristidactylus in relation to the Argentine species of the genus and the genera most closely related to Pristidactylus (Diplolaemus Bell, 1843 and Leiosaurus Duméril & Bibron, 1837) (^{Morando et al. 2015}; ^{Femenias et al. 2020}). Therefore, we included representatives of all species of genera Diplolaemus and Leiosaurus and of five of the six species of Pristidactylus from Argentina. Also, we included one specimen of P. torquatus from Cordillera de Nahuelbuta (^{Morando et al. 2015}). We performed a Bayesian Inference (BI) analysis with the program MrBayes v3.2.7a (^{Ronquist et al. 2012}). A reversible-jump method to explore the space of all General Time Reversible sub-models, plus gamma and proportion of invariable sites parameters was applied independently to each codon position (three partitions). The analysis consisted of four independent chains run for 20 million generations, sampled every 1000 generations. The first 25% of generations was conservatively discarded as burn-in after observing the stationarity of ln-likelihoods of trees in Tracer v1.7.1 (^{Rambaut et al. 2018}). Convergence and mixing of chains were assessed by examining values of average standard deviation of split frequencies (ASDSF), and expected sampling sizes (ESS) and Potential Scale Reduction Factor (PSRF) for all parameters. Trees were rooted with a representative of the genus Urostrophus Duméril & Bibron, 1837 following the most recent phylogenetic studies of these leiosaurid genera (^{Morando et al. 2015}; ^{Femenias et al. 2020}).

We obtained an alignment of 809 nucleotide sites for the cytb. We recovered four main clades, two corresponding to the genera Diplolaemus and Leiosaurus, and other two made up of the Chilean and Argentine species of Pristidactylus (Fig. 2). According to our results, the Chilean clade of Pristidactylus is the sister group of Leiosaurus, but this relationship is supported by a low posterior probability (pp = 0.84), whereas the Argentine species of Pristidactylus comprise the sister group of the genus Diplolaemus, also with a low support (pp = 0.70). The four Chilean species of Pristidactylus comprise a highly supported clade (pp = 0.99), where P. torquatus is the first to diverge, followed by P. alvaroi, which corresponds to the sister taxon of P. valeriae + P. volcanensis (this last relationship with low support, pp = 0.80).

FIGURE 2 Bayesian consensus tree (50% majority-rule) of the fragment of the cytochrome b, showing the relationships of Pristidactylus, Diplolaemus and Leiosaurus. Specimens of the two Pristidactylus lineages are shown with different colors: Chilean (red) and Argentine (blue). The values next to the nodes correspond to posterior probabilities and the scale bar below the tree represents the expected substitutions per site along the branches. / Árbol de consenso bayesiano (regla de la mayoría del 50%) del fragmento del citocromo b, mostrando las relaciones de Pristidactylus, Diplolaemus y Leiosaurus. Los dos linajes de Pristidactylus se muestran con diferentes colores: chileno (rojo) y argentino (azul). Los valores junto a los nodos corresponden a las probabilidades posteriores y la barra de escala debajo del árbol representa las sustituciones esperadas por sitio a lo largo de las ramas. 

In the last two decades, several studies have been conducted to elucidate the phylogenetic relationships and evolution of the leiosaurid genera Pristidactylus, Diplolaemus and Leiosaurus (reviewed in ^{Morando et al. 2015}). The most recent and complete phylogenetic studies of these genera (^{Morando et al. 2015}; ^{Femenias et al. 2020}) have been consolidating the position that Diplolaemus and Leiosaurus are monophyletic, whereas Pristidactylus is polyphyletic. Here, we obtained similar results, recovering the same four main groups of leiosaurids, but there are some differences from those studies.

We obtained low support values (pp < 0.95) for the clades corresponding to the genera Diplolaemus and Leiosaurus and a relationship between Diplolaemus and the Argentine species of Pristidactylus (also with low support) (Fig. 2). Instead, ^{Morando et al. (2015}) (in their species tree) obtained high levels of support for the genera Diplolaemus and Leiosaurus and the clade formed by the Argentine species of Pristidactylus. However, like ^{Morando et al. (2015}), we recovered a relationship between P. torquatus and Leiosaurus, although with low support. On the other hand, ^{Femenias et al. (2020}) (in their species tree) also obtained high support for Diplolaemus, Leiosaurus and the Argentine species of Pristidactylus, but recovered to P. torquatus as sister of these three clades. These differences can be attributed to the fact that we used only one mitochondrial marker, but it should be noted that despite the fact that ^{Morando et al. (2015}) and ^{Femenias et al. (2020}) used several genes (two mitochondrial and eight nuclear), in general they did not obtain high support values for the relationships among genera.

Both ^{Morando et al. (2015}) and ^{Femenias et al. (2020}) recovered to P. torquatus separated from the Argentine species of the genus, although in different positions. Here, we obtained the same result, but including all species of Pristidactylus from Chile, which grouped with high support; moreover, a division is observed between P. torquatus and the three species from central Chile, also with a high support (Fig. 2). The relationship of P. torquatus as sister to the other species from Chile had already been obtained in previous studies, for example, by ^{Frost et al. (2001}) (in their analysis of morphological and molecular data combined, where the Argentine species P. fasciatus is closely related to the Chilean species) and ^{Abdala et al. (2009}) (in their morphological analysis, where P. alvaroi was not included). Among the three species from central Chile, we recovered P. valeriae and P. volcanensis as sister species (Fig. 2), a result also obtained by ^{Abdala et al. (2009}) in their morphological analysis.

In summary, despite using only one mitochondrial gene, we confirmed previous studies that recovered Pristidactylus as a polyphyletic genus (^{Morando et al. 2015}; ^{Femenias et al. 2020}). Our analysis also shows that the four endemic species of Chile make up a well-supported lineage that has evolved separately from their counterparts on the other side of the Andes. This is consistent with the morphological and ecological differences that have traditionally been used to distinguish the two groups of species within the genus. However, a larger number of samples from different populations and the inclusion of nuclear markers are required to better understand the evolutionary history and establish the taxonomic status of the Pristidactylus lineage that is restricted to the west of the Andes in Chilean territory.