**4. Discussion**

Molecular genetics, and in particular barcoding analyses, proved to be a powerful tool to generate preliminary information about the taxonomic status of problematic taxa [20,75]. We present here the most comprehensive analysis of genetic samples of Squamata from Paraguay. The results obtained here will be useful to help identify questionable specimens and in some cases also to clarify some taxonomic issues of the Squamata fauna from the central region of South America. Thus, the data generated here will have a positive impact in a larger geographic context, beyond Paraguay's borders.

As said before, genetics alone will not yield a well-founded taxonomy. Nevertheless, molecular genetics open a path for defining operational taxonomic units (OTUs), identifying potential undescribed species and pointing to taxonomic problems, and thus have to be seen as a first informative step and a complementary evidence line in the framework of the modern integrative taxonomic approach [20,76].

Some taxonomic results of this project were already published. For instance, the samples of *Colobosaura* exhibit large genetic distances, and then *Colobosaura kraepelini* was revalidated [41]. The *Tropidurus* samples show monophyly in the species of the *torquatus* group (*T. catalanensis* and *T. etheridgei*), but indicate several uncertainties within the *spinulosus* group (formerly *T. guarani*, *T. lagunablanca*, *T. spinulosus*, *T. tarara*, and *T. teyumirim*), that resulted in the synonymization of *T. guarani* with *T. spinulosus*, and *T. tarara* and *T. teyumirim* with *T. lagunablanca* [43]. Regarding the Family Phyllodactylidae, there is strong evidence for the recognition of two different *Homonota* species in the Chaco [42,44] and a highly distinctive *Phyllopezus* clade, separated from populations from Cerrado and Chaco [45].

The samples of *Vanzosaura rubricauda* from Cerrado (field number "ALA") show a high branch distance compared with *Vanzosaura rubricauda* from Chaco (GK 3801) which is even larger than the distance from *V. multiscutata* (Figure S10). Integrating molecular and morphological data, a new species of *Vanzosaura* (*V. savanicola*) was previously described, and *Gymnodactylus multiscutatus* was transferred to the genus *Vanzosaura* [35]. Nevertheless, their genetic tree [35] included only a single sample from Paraguay and none from Argentina. In their map, obviously two divergent populations of *Vanzosaura rubriauda* are recognized: One west of the Paraguay River in the Dry Chaco, and another east of the Paraguay River in the Cerrado. Keeping a conservative approach, the authors maintained *V. rubricauda* as a single taxonomic unit, but with our additional samples it might be possible to generate new taxonomic hypotheses.

Furthermore, we recommend further studies on Amphisbaenidae, because one of the major and latest revisions of Amphisbaenidae in the Neotropics concluded that *A. mertensi* and *A. cunhai* (not recorded in Paraguay) are the most basal lineages of the genus [77]. Our analysis showed that the most basal sample (*Amphisbaena* sp. PCS 314) seems to be a different species as those within the remaining clade. Additional analyses, including more samples and a detailed morphological revision, are necessary to assess the specific status of that specimen.

The weakest part of this work was the analysis of snakes. These animals are usually the harder ones to sample. Compared to the actual diversity of Colubridae, our dataset had fewer samples of this family and therefore it was not possible to draw detailed taxonomic conclusions. However, the presence of the genus *Xenodon* in two different clusters suggests that more taxonomic work with this group of snakes is needed. Several taxonomic modifications occurred within the Colubridae in the last decade, where the genera *Lystrophis* and *Waglerophis* were synonymized with *Xenodon*, based on the analysis of gene sequences 12S and 16S for the genus *Lystrophis*, and Cytb and bdnf for one sample of *Waglerophis merremii* [78]. In our analysis, we found the samples of *X. pulcher* (previously *Lystrophis pulcher*) separated from *X. merremii* (previously *Waglerophis merremii*). It is desirable to perform phylogenies in this group using more nuclear data to ge<sup>t</sup> more robust relationships in the deep nodes.

In the clade of the genus *Thamnodynastes*, the two species used in our analysis (*T. chaquensis* and *T. hypoconia*) are nested in the same node. A more specific genetic study of the genus is highly recommendable. Furthermore, the revision of the phylogenetic status of *Phimophis* is advisable, since here it appears polyphyletic. In a former study, two samples of *Phimophis* were used: *P. guerini* (GQ457761) and *P. iglesiasi* (JQ598891) and due to polyphyly the authors described the genus *Rodriguesophis* to include the latter species [78]. This genus is characterized by the absence of the loreal scale. Both *P. guerini* and *P. vittatus* have a loreal scale, so they cannot be assigned to *Rodriguesophis*. Thus, a deeper integrative (morphological and molecular) analysis is needed to understand their relationships.

The genus *Micrurus* is scarcely represented in GenBank, and comparisons are not possible. The only sample from GenBank is *M. altirostris*, which is differentiated from Paraguayan samples by a rather long branch distance. There is a polytomy with three samples (PCS 310, 334, and 337), and the only identified specimen is *Micrurus pyrrhocryptus* (PCS 310) from Pantanal (northern part of Paraguay). The other samples are from Concepción at the other side of the river, and with a body color different from the pattern of *M. pyrrhocryptus*. It is important to highlight here that some of the specimens that we have of *Micrurus*, were decapitated (killed by rural farmers) and therefore, without genetic samples for comparison (in GenBank) and without cephalic data (which contains important diagnostic characters) [79], its specific taxonomic allocation becomes difficult.

Some of our Paraguayan samples of *Bothrops* of the *neuwiedi* complex, from distant parts of Paraguay, are clustered with a sample of *B. diporus* from GenBank (Samples PCS 302, PCS 318, PCS 331, PCS 504, Figure S7). A thorough analysis of this complex of *Bothrops* is needed to understand the true diversity in the group.

Regarding the Scincidae, a sample from GenBank of *Manciola guaporicola* (KX364960) from Mbaracayú Reserve (Paraguay) appears out of the clade of the remaining *M. guaporicola* from Paraguay and Brazil. This is also a topic that should be further investigated.

Regarding conservation, one of the major problems in Paraguay for several years was habitat loss due to extensive soybean crops in the eastern part of Paraguay [80]. Nevertheless, habitat fragmentation is currently also affecting the landscapes of the Occidental Region of Paraguay [81,82]. Thus, currently, the protected areas are the best strategy for conservation of biodiversity in Paraguay, although many conservation units face legal problems (e.g., lack of official measurements, managemen<sup>t</sup> plans, forest guards, infrastructure, etc.) and then the long-term maintenance of their biodiversity is not guaranteed [80]. There are some reptile species absent from protected areas in Paraguay; therefore, monitoring and conservation efforts should be intensified for these taxa [83]. One species recently revalidated is *Colobosaura kraepelini*. This lizard is known only from the holotype from the locality of Puerto Max (San Pedro Department), the neotype from Altos and an additional specimen from San Bernardino, both localities in Cordillera Department [41]. There are no protected areas in the Cordillera Department, but there are some in the northern portion of Central Department (border with Cordillera), located less than 10 km from the known localities of *C. kraepelini*. The presence of this species in a conservation unit should be confirmed, but is possible that it is protected by "Monumento Natural Cerro Chororí" and "Monumento Natural Cerro Kõi". It is important to note that the conservation unit closer to the distribution of *C. kraepelini* is the "Parque Nacional Lago Ypacaraí", although only the lagoon is protected and not the surroundings. The species *Homonota septentrionalis* was described from the driest part of Paraguay (northwestern Chaco) and is abundant in the "Parque Nacional Teniente Enciso" [44]. The four species of *Tropidurus* found in Paraguay [43] are well represented in several protected areas. The last herpetofaunal conservation assessment was published in 2009 [84], and thus a new conservation assessment of Paraguayan reptile fauna including the new taxa, is necessary to provide a sound basis for conservation planning for those species that require special attention.

Finally, given that the sequenced specimens are ye<sup>t</sup> a small portion of the actual diversity of Paraguay (Figure 5), it will be of the utmost importance to continue and expand these studies that will further improve our taxonomic knowledge.

**Figure 5.** Comparison between known diversity of Paraguayan Squamata (blue bars) vs. diversity of sampled taxa for this study (red bars).

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1424-2818/11/9/152/ s1, Table S1: Reagents used for molecular procedures; Table S2: Specimens used in the barcoding analysis; Figures S1–S16: Details of sub-trees sections from the general barcoding analysis.

**Author Contributions:** Conceptualization, P.C., E.B., and G.K.; data curation, P.C., E.B., and G.K.; methodology, P.C.; formal analysis, P.C.; supervision, G.K.; writing—original draft preparation, P.C.; writing—review and editing, G.K. and E.B.; funding acquisition, P.C. and E.B.

**Funding:** This research was developed in the framework of the Ph.D. thesis project "Lizards of Paraguay: an integrative approach to solve taxonomic problems in central South America" presented by P.C. in the Faculty of Biosciences of the Johann Wolfgang Goethe-Universität, which had a strong financial support from the Deutscher Akademischer Austauschdienst (DAAD). Additionally, P.C. received a subsidy from the Consejo Nacional de Ciencia y Tecnología (CONACYT) through the Programa Nacional de Incentivo a los Investigadores (PRONII), field equipment provided by Idea Wild, and a gran<sup>t</sup> to perform fieldwork activities in the Mbaracayú Reserve by funded by the GNB Bank through the Fundación Moisés Bertoni (FMB). Additionally, we received financial support from PRESIDENT ENERGY to pay the APC of this work.

**Acknowledgments:** This was a major project that had the help of many people. For this, we owe gratitude to Norman Scott, Luciano Avila, Mariana Morando, Tony Gamble, Sebastian Lotzkat, Martin Jansen, Arne Schultze, Christian Printzen, Raúl Maneyro, and Abel Batista, who provided advice and technical support at different stages of the project. This project would not have been achieved without the valuable help of many colleagues and friends who offered their help during field work. Hence, thank you so much Jorge Ayala, Frederick Bauer, Aníbal Bogado, Enrique Bragayrac, Diego Bueno, Emilio Buongermini, "Huguitus" Cabral, Paulo Campos Filho, Gloria "Lolex" Céspedes, Jorge A. Céspedez, Julia Coda, Sixto Fernández, Andrea Ferreira, Marcela Ferreira, Celeste "Tita" Gauto, Kevin Guest, Hugo del Castillo, Stefan Harrison (you rock man!), Monica Kozykariski, Victoria Kuntz, Arne Lestheruis, Pamela Marchi, Nicolás Martínez, José Méndez, Martha Motte, Cristian F. Pérez, Pastor Pérez, Rachel Pitts, Sergio Ríos, Lía Romero, Mirtha Ruiz Díaz, Humberto "Ka'umber" Sánchez, Rebecca Sheehan, Nelson Silva, María E. Tedesco, Dulcy Vázquez, Thomas and Sabine Vinke, Akira Yoshikawa, and Víctor Zaracho. Furthermore, we have to acknowledge the hospitality of persons (Rubén Ávila and Nilda Torres de Ávila, Massimo and Angela Coda, Rosario Gabaglio, Ana María Macedo, and Goli Stroessner) and organizations (Fundación Moisés Bertoni, Guyra Paraguay, and Para La Tierra) that provided accommodations during field wok. We also want to express our gratitude to some friends and colleagues that provided tissue samples, literature, or valuable information. They are Luciano J. Avila, Francisco Brusquetti, Arley Camargo, Santiago Carreira, André de Carvalho, Mariana Morando, and Paul Smith. Additionally, we are grateful to the staff (especially Heike Kappes) of the Grunelius-Möllgaard Laboratory (Senckenberg Forschungsinstitut und Naturmuseum Frankfurt - SMF), and to Linda Mogk (SMF) for lab support. P.C. also wants to thank to Tachi,

417

Dante and Rafa for patience and support. Finally, to the CONACYT for providing important tools for scientific research in the country.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
