A new Terrarana frog of genus Pristimantis from an unexplored cloud forest from the eastern Andes, Colombia

Abstract A new species of Pristimantis (Craugastoridae, subgenus Pristimantis) is described from a relict and unexplored cloud forest in the western slope from Cordillera Oriental of the Colombian Andes. The specific name was chosen by consensus expert scientists and local people. Pristimantis chamezensissp. nov. is easily distinguished from congeneric species by having a gray iris with black reticulations in life, subconical tubercles on the upper eyelid, the chin edged with irregular, dark-brown blotches, and conical heel tubercles. The phylogenetic analyses suggest that the origin and radiation of its clade may have occurred in the highlands. With the description of P. chamezensissp. nov., we identify 14 species distributed throughout the eastern slope of the Andes that are associated with the Orinoco Basin.


introduction
The amphibian fauna from Colombia is among the richest and most diverse in the world (Lynch 1999;Grant et al. 2008) and includes 843 species (Acosta-Galvis 2020). A significant number of these species is grouped in the so-called Terrarana; an unranked taxonomic grouping of at least four closely related families characterized by direct development, egg embryos, and terrestrial reproduction (Hedges et al. 2008;Heinicke et al. 2009Heinicke et al. , 2018. Terrarana richness in Colombia includes 268 species in 13 genera (Acosta-Galvis 2020), among which the frogs of the Pristimantis genus represent the greatest diversity with 83% of the described species.
Morphologically, frogs of the genus Pristimantis are easily recognizable among other features by terminal discs on expanded digits and T-shaped terminal phalanges, a dentigerous process of the vomers usually present, and toe IV much longer than toe III (Hedges et al. 2008;Duellman and Lehr 2009). However, this genus still has latent phylogenetic challenges (Navarrete et al. 2016), and recent proposals, based on molecular phylogenies (Hedges et al. 2008;Padial et al. 2014;Páez and Ron 2019;Reyes-Puig et al. 2020), reassigned or excluded members of the species groups from evolutionary arrangements, which were previously based solely on morphological evidence Duellman 1980, 1997).
The genus Pristimantis in Colombia is represented by 223 formally described species (Acosta-Galvis 2020). The geographic and ecological complexity of the Andes harbors the greatest richness and rate of endemism in contrast to the lowlands of the Amazon and Pacific basins . Current geological evidence of the north-Andean region indicates that the northern formations in Colombia (Occidental, Central, and Oriental mountains ranges) have promoted speciation processes in the genus Pristimantis and, therefore, have high diversity and endemism (Lynch and Duellman 1997;Lynch 1999; García-R et al. 2012;Mendoza et al. 2015;Meza-Joya and Torres 2016;Acevedo et al. 2020). Among these geographical units, the Cordillera Oriental contains 44 species, with 13 species inhabiting the Andean and sub-Andean forests on the eastern slopes (Table 1), as part of the Orinoco basin (Acosta-Galvis et al. 2010;; Ospina-Sarria and Angarita-Sierra 2020; Acevedo et al. 2020).
During field studies along an unexplored cloud forest (2140 m a.s.l.) in the Cordillera Oriental, we collected several specimens of Pristimantis that, due to their morphological characters, are not assignable to any described species in this region. Based on the analysis of its molecular data and morphology, we describe a new species recognized by its molecular and morphological distinctiveness.

Study area
We collected by actively searching from September 2 to November 29, 2010, using intensive visual encounter surveys (Crump and Scott 1994) during evenings in the cloud forests in the municipality of Chámeza (05°15'24.4"N, 072°53'51.6"W), Department of Casanare, Colombia (Fig. 1). This locality is part of an elevated area between 1700-2200 m a.s.l. in an unexplored northern portion of the Cordillera Oriental. This mountainous area consists mainly of pristine natural forests of the Andes orobiome ( Fig. 2) within the ecoregion of the Eastern Cordillera montane forests of Colombia (Dinerstein et al. 1995;Olson and Dinerstein 2002). We recorded geographical coordinates and elevations at collecting sites using a Garmin GPSMAP 60CSx (map datum WGS 84).

Data collection and laboratory procedures
Molecular distinctiveness and phylogenetic relationships of the new species were assessed by analyzing DNA sequences of mitochondrial DNA (mtDNA) which included a fragment of the 16S ribosomal RNA (16S) and a fragment of the cytochrome oxidase subunit 1 (COI) genes. We assembled a data set that included only the 16S gene fragment by aligning sequences from all known Pristimantis species from the eastern slopes of the Cordillera Oriental of Colombia together with the most similar sequences already published in Genbank (Table 2). For this, we conducted a search for sequences similar to the 16S gene fragment of the new species using the BLAST algorithm in GenBank. The most similar 127 BLAST hits to the sequences from the new species  200-600 Lynch 1975;1980, 1994Padial et al. 2014;Heyer and  were downloaded, aligned, and assessed using Bayesian (BA) and maximum likelihood (ML) analyses. After removing distant and redundant sequences, the final dataset contained 58 sequences of 827 base pairs (bp) of the 16S, including the new species and Pristimantis medemi (Lynch, 1994) obtained in this study (Table 1). We assembled a complete data set comprising sequences of the 16S, concatenated with sequences of the COI gene for a subset, including the new species and its following six most-related species, selected based on the results of the analyses: Pristimantis carranguerorum (Lynch, 1994), P. bowara Acevedo et al., 2020, P. lutitus (Lynch, 1984, P. medemi (Lynch, 1994), P. nicefori (Cochran & Goin, 1970), and P. savagei (Pyburn & Lynch, 1981). From two tissue samples of the new species and a tissue sample of Pristimantis medemi we extracted total genomic DNA using a standard Phenol-Chloroform method (Sambrook et al. 1989). We amplified the gene fragments using the primers pairs 16Sbr-H/16SC-16L (Palumbi et al. 1991;Darst and Cannatella 2004, respectively) and LCO1490/HCO2198 (Folmer et al. 1994) for the 16S and COI, respectively. We carried out PCRs in a total volume of 30 μl containing one unit Taq polymerase (Bioline; Randolph, MA), 1× of a buffer (Bioline), a final concentration of 1.5 mM MgCl2 (Bioline), 0.5 μM of each primer, 0.2 mM of each dNTP (Bioline), 0.2 μg of bovine serum albumin (BSA), and approximately 50 ng of total DNA. We purified the PCR products using the ammonium acetate protocol (Bensch et al. 2000), and we sequenced them on an ABI 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) using the BigDye Terminator v. 3.1 Cycle Sequencing Kit (Applied Biosystems). We stored the remaining DNA extractions at -80 °C in the tissue collection of the Instituto de Genética, Universidad Nacional de Colombia (for voucher numbers see Table 2). We performed the thermocycling conditions as indicated by the authors, who reported the primers for the obtained fragments. The GenBank accession numbers of the obtained sequences are MK776946-MK776948 and MK789293-MK789295. We edited and aligned the sequences using Chromas 1.51 (http://www. technelysium.com.au/chromas.html) and BioEdit v. 7.0.5.2 (Hall 1999). To exclude divergent regions and poorly aligned bases from the 16S dataset, we used the software Gblocks v. 0.91b (Castresana 2000;Talavera and Castresana 2007; available as a web server at http://molevol.cmima.csic.es/castresana/Gblocks server.html), which resulted in a final alignment of 528 base pairs (bp). The COI alignment consisted of 652 bp.

Phylogenetic and genetic divergence analyses
We analyzed the complete evidence dataset using the following partition scheme: (i) unpartitioned; (ii) partitioned by gene (i.e., each gene fragment treated as a distinct partition); and (iii) maximum partitioning (i.e., we treated each codon of the proteincoding gene COI and the ribosomal gene fragment as distinct partitions). We assessed the optimal partitioning scheme and best-fit evolutionary models using PartitionFinder v. 1.1.1 and the Bayesian Information Criterion (Lanfear et al. 2012), resulting in the selection of the maximum partitioning scheme. For the 16S dataset, the obtained model (SYM + G) was applied in a Bayesian analysis (BA) with MrBayes v. 3.2.1 (Ronquist et al. 2012). For the complete evidence dataset, we applied the 16S fragment model plus the following complementary COI fragment resulting models in a Bayesian analysis with MrBayes: COI 1 st codon -TrNef + G, COI 2 nd codon -HKY, COI 3 rd codon -HKY. We incorporated these models into a single tree search (mixed model partition approach; Nylander et al. 2004). For both analyses, we carried out two parallel runs using four Markov chains, each starting from a random tree. We ran the Markov chains for 10 million generations. The burn-in was set to sample only the plateau of the most likely trees that were used for generating a 50% majority rule consensus. We used the software TRACER v. 1.5.4 (Rambaut and Drummond 2007) to assess an acceptable level of the MCMC chain mixing and to estimate effective sample sizes for all parameters. Additionally, maximum likelihood (ML) analyses were run using RAxML 7.2.8 (Stamatakis 2006) and the GTR+G model. We performed five independent maximum likelihood searches with different starting conditions and the rapid bootstrap algorithm to explore the robustness of the branching patterns by comparing the best trees. Afterward, 1000 non-parametric thorough bootstrap values were computed and plotted against the best tree. The Genbank sequence of Eleutherodactylus johnstonei Barbour, 1914, EF493561, was used as outgroup. To assess the genetic divergence between the new and the other Pristimantis species, we calculated uncorrected p genetic distances for the 16S and the COI fragments using MEGA v. 7.0.21 (Kumar et al. 2016).

Morphology
We euthanized specimens using Clorethone, which were then fixed in 10% formalin, preserved in 70% ethanol, and deposited in the biological collections of the Instituto de Investigación de Recursos Biológicos Alexander von Humboldt, Villa de Leyva, Boyacá, Colombia (IAvH-Am). Other specimens examined are listed in Suppl. material 1. The criteria for the definition of descriptions and diagnostic characters followed Duellman and Lehr (2009), Lynch and Duellman (1997), and Navarrete et al. (2016).
To identify sex and sexual maturity, we made a small incision in the groin region for macroscopic observation of the gonads. Adult males have the granular testis, while females show enlarged, thickened, and convoluted oviducts. Morphometric measurements were made with digital calipers (nearest 0.01 mm) or a Nikon stereoscopic microscope SMZ-1B with high Intensity Illuminator NI-150 Nikon as follows: SVL (snout-vent length), HW (head width), HL (head length from the tip of the snout to the posterior border of the skull, posterior edge of prootic, noted through the skin), IOD (interorbital distance), ED (eye diameter), EN (eyes-nares distance), UEW (upper eyelid width), ETS (distance between the anterior edges of the eye to the tip of the snout), TD (horizontal tympanum diameter), RW (rostral width), InD (internarial distance), TL (tibial Length), FL (femur length), FtL (foot length), and HnL (hand length). Means are reported ± one standard error. We photographed habitats and specimens using Canon EOS 30D and EOS 5D Mark II digital cameras inside a Photo Safe-box using 5.500 kelvins LED lights.

Phylogenetic and genetic divergence analyses
The resulting phylogenetic tree including all 58 sequence of the 16S fragment is shown in the Suppl. material 2: Fig. S1. A reduced phylogenetic tree including the 16S fragment sequences of the new species and its closest 29 sequences is shown in Figure 3.
The following description is referring to the reduced tree. Based on the phylogenetic relationships, the new species could be assigned to the genus Pristimantis, subgenus Pristimantis. Both tree-building methods revealed Pristimantis chamezensis sp. nov. with maximum support within a supported monophyletic group comprising Pristimantis carranguerorum, P. bowara, P. lutitus, P. medemi, P. nicefori, and P. savagei (Fig. 3). Both analyses concurred in placing the new species as a sister taxon of P. nicefori with low support (ML: 40%; BA: 0.80). The other 23 Pristimantis species were revealed by both analyses within three separated, weakly supported clades, exhibiting low supported evolutionary relationships (Fig. 3). For the complete evidence dataset, both tree building methods revealed Pristimantis chamezensis sp. nov., as part of a monophyletic clade also comprising P. carranguerorum, P. bowara, P. lutitus, P. medemi, P. nicefori, and P. savagei with maximum support (Suppl. material 3: Fig. S2). Both analyses revealed that the new species is the sister taxon of a clade showing the following weakly supported phylogenetic relationships: (((P. lutitus + P. bowara) P. nicefori) P. carranguerorum). Finally, P. medemi and P. savagei were revealed as successive sister taxa of the that clade plus the new species, with low support (Suppl. material 3: Fig. S2). Genetic distances for the 16S gene between P. chamezensis sp. nov. and P. nicefori, P. carranguerorum, and P. savagei were 4.8%, 5.2%, and 5.9%, respectively. Distances between P. chamezensis sp. nov. and P. medemi, P. lutitus, and P. bowara were 6.2%, 6.2%, and 6.7%, respectively (Table 3). The sequence divergence range of the monophyletic group compared to the other analyzed taxa was 5.9-4.1% (Table 3). The uncorrected p distances for the COI gene revealed that sequence differentiation values between P. chamezensis sp. nov. and P. carranguerorum, P. nicefori, P. lutitus, P. savagei, and P. medemi were 6.2%, 6.4%, 6.7%, 6.7%, and 6.7%, in that order. For the same gene fragment, the distance between P. chamezensis sp. nov. and P. bowara was 7.8%.
Description of the holotype. An adult female (Figs 4, 5) with a snout-vent length (SVL) of 23.8 mm; the skin of cephalic region, dorsum, eyelids, lateral surfaces, and dorsal thighs shagreen with scattered larger tubercles; dorsolateral folds absent and discoidal folds visible; skin on venter areolate. Head length (HL), diagonally from the corner of mouth to tip of snout 10.4 mm; head width (HW) 9.8 mm, approximately equal to width of the body and 41.1% of the SVL. Snout broadly rounded in dorsal view (type F, sensu Duellman and Lehr 2009;Fig. 4) and rounded in lateral view (type A, sensu Duellman and Lehr 2009;Fig. 4); internarial distance (between center of naris) 2.9 mm; nostril moderately protuberant, directed dorsolaterally; canthus rostralis well defined; loreal region slightly concave; lips not prominent. Eye diameter (ED) from its posterior to anterior corner 3.3 mm; its length 73.3% of the ETS (distance between the anterior edge of the eye to the tip of snout); interorbital region wider than upper eyelid; the upper eyelid width (UEW) 55.2% of interorbital distance (IOD); upper eyelid bearing three smaller subconical tubercles (Figs 4,5); no cranial crests. Supratympanic fold low and short. Tympanic membrane and tympanic annulus present, small, and rounded (Figs 4, 5), its dorsoposterior border converges with supratympanic fold; its diameter 1.1 mm and equivalent to 33% of eye diameter (ED). Choanae subovoid, not concealed by the palatal shelf of the maxillary arch; dentigerous processes of vomers prominent, nine teeth positioned posterior to level of choanae and widely separated from each other. Tongue rounded, its posterior border notched for half of its extension is adherent to the floor of mouth; teeth present on the maxillary arch.
Hindlimbs slender; foot length (FtL) 12.1 mm, 50.8% of SVL. Toe webbing and toe fringes absent. Relative lengths of appressed toes IV>V>III>II>I. Discs of the toes expanded; width of adjacent phalange 53.7% of disc of toe IV; disc of toe III does not reach penultimate subarticular tubercle of toe IV; toe V beyond that of the level of penultimate subarticular tubercle of toe IV. Femur length (FL) 13.6 mm, tibia length (TL) 14.5 mm, its length is equivalent to 60.9% of SVL. Subarticular tubercles 1-1-2-3-2; supernumerary plantar tubercles numerous, suboval, and low; inner metatarsal tubercle oval; outer metatarsal tubercle rounded, prominent, and smaller than inner metatarsal tubercle. Diameter outer metatarsal tubercle 52.8% of inner metatarsal tubercle; outer tarsal fold absent; inner tarsal fold short. Numerous supernumerary plantar tubercles rounded and barely visible; subarticular tubercles large, round, and conical; toes without lateral fringes; no webbing. Cloacal sheath absent; subcloacal tubercles absent.
Color of holotype in preservative (Fig. 4). Dorsum and flanks dark brown; hands in dorsal view, with fingers I and II cream-colored, while fingers III and IV brown with cream-colored bars; dorsal surfaces of the thigh with diffuse dark-brown transversal bars; hidden surfaces of thighs pale brown; venter light brown with a dark-brown suffu- sion and mottled brown; ventral surfaces of hindlimbs and forelimbs dark brown with a cream-colored suffusion; edge of chin with irregular blotches of dark brown; hands, in ventral view, with palmar tubercle cream-colored and palmar region dark brown.
Color of holotype in life (Fig. 5). Dorsal surfaces of body and limbs pink-orange; flanks salmon and sides of the head pink-orange; venter reddish cream-colored on chest and belly, cream-colored on throat; axillary region, groin, and anterior thigh pale orange; ventral surfaces of thighs light brown; iris gray, medially reddish, with black reticulations.
Variation of type series (Fig. 5, Table 4). In this section, coloration refers to specimens in life and is based on field notes and digital photographs, unless otherwise noted. Dorsal coloration reddish brown with mottled, dark-brown chevrons, usually surrounded by a thin band of lighter color; canthal stripe black; dorsal surfaces of thigh with dark-brown transversal bars; axillary region, groin, and anterior thigh bright orange (e.g., IAvH-Am-10283, IAvH-Am-10276; Fig. 5) or uniformly dark brown (e.g., IAvH-Am-10267-68, IAvH-Am-10272; Fig. 5). An adult female (IAvH-Am-10277) has a gold paravertebral line (Fig. 5). Labial bars dark brown, and postorbital and supratympanic stripe dark (e.g., IAvH-Am-10268, IAvH-Am-10270, IAvH-Am-10272, IAvH-Am-10276-7; Fig. 5). In IAvH-Am-10270 and IAvH-Am-10276, flanks with oblique, irregular, dark-brown bars (Fig. 5); IAvH-Am-10267 with a W-shaped, lightbrown marking on scapula; some specimens with a dark-brown interorbital bar (e.g., IAvH-Am-10268, IAvH-Am-10273-4, IAvH-Am-10279-10280). Pristimantis chamezensis is metachromatic, being lighter in color at night. Teeth positioned posterior to level of choanae and widely separated from each other, which vary between eight to nine. The variation in the skin texture is noteworthy (Fig. 5), varying from smooth (e.g., IAvH-Am-10283) to shagreen with scattered tubercles (e.g., IAvH-Am-10267, IAvH-Am-10277). The SVL of adult males ranges from 19.6 to 23.7 mm (Table 4), and the SVL of adult females ranges from 19.0 to 24.9 mm ( Distribution and natural history. This species is only known from the type locality at an altitude between 2125-2160 m a.s.l. in an Andean and relictual cloud forest in the Casanare region on the eastern flank of the Cordillera Oriental of Colombia (Fig. 1). The locality is within the Cordillera Oriental montane forest ecoregion (sensu Dinerstein et al. 1995) in the Andean region (Middle Orobiome). The forest (Fig. 2) is unaffected by human activities and is typified by a canopt of medium-height (up to 20 m). The annual precipitation is between 4600 and 5600 mm with bimodal seasonality. Specimens were found active during the second annual rainy season (August to November) at a temperature of 14 °C resting on mosses and lower leaves of shrubs and ferns in the undergrowth. Pristimantis chamezensis is syntopic with an undescribed species of genus Pristimantis.
Etymology. The specific epithet is derived from the Municipality of Chámeza, a geopolitical area where the type locality is located. We decided on this name using a citizen science approach, where expert scientists and local people met and discussed a list of possible names and their corresponding meanings. There was consensus on P. chamezensis as the preferred name.
Conservation status. The direct evaluation of the landscape units (e.g., broad-leaved forest) at the type locality, as well as the map of land cover of Colombia (CORINE Land Cover, IDEAM 2010), allowed us to identify a rapid reduction and low connectivity of its habitat. Based on land cover maps of Chameza's forest, the potential extent of occurrence is 301,624 km 2 . Consequently, we propose to categorize P. chamezensis as Vulnerable using the criteria B2a (IUCN Red List Categories and Criteria 2019).

Colombian diversity of the genus Pristimantis in a biogeographical context
The genus Pristimantis, with 556 described species, comprises of a substantial number of identified taxa (Frost 2020). Colombia harbors 40% of this diversity. The Andean Cordilleras harbor 183 species (Acosta-Galvis 2020), evidencing the high rate of speciation and endemism of the genus in this ecoregion (Lynch 1999), while in the lowlands (Pacific, Middle Magdalena, and Amazon basins) there are just 52 species. The diversity of Pristimantis of the Andean-Cordillera and Sierra Nevada of Santa Marta reflects the geological history of these mountains (Lynch and Ruiz-Carranza 1985;. Consequently, the geological formations of the Cordillera occidental (25 Ma old, with the greater species richness), Cordillera Central, and the Central Massif exhibit a 30% similarity of species. While, the Cordillera Oriental (10 Ma old; Gregory-Wodzicki 2000) and Sierra Nevada de Santa Marta (2.6 Ma old; Idárraga et al. 2011) have allowed the evolution of an unparalleled diversity with a high degree of endemism   (Fig. 8).
Despite this rough correspondence between the geological history of the Colombian Andes and Pristimantis diversity, the inventory of species in each region is far from being completed. Socio-political factors affecting the various regions of Colombia have limited scientific access, leaving several crucial regions with pronounced gaps in our knowledge of amphibians. Among these regions, we highlight the northern lowland regions of the upper Amazon, including Putumayo, Caquetá, Guaviare, Guainía, and Vaupés departments, as well as neighboring areas such as the Darien region. Additionally, some other unsampled areas are the tropical rainforests in the Pacific basin and the Andean region, such as the Serranias of Perijá and San Lucas, southern Cordillera Oriental (including the Andean-Amazonian foothills) and mountainous areas associated with the Orinoco drainage (Fig. 8).
Over the past six years of scientific studies in unexplored mountainous areas within the Orinoco drainage, including cloud forests and foothills of the Cordillera Oriental, several species of Pristimantis have been described (e.g., Acosta-Galvis et al. 2010; Acosta-Galvis and Alfaro-Bejarano 2011; Pedroza-Banda et al. 2014;Acevedo et al. 2020; Ospina-Sarria and Angarita-Sierra 2020). However, there is still a long way to go to characterize the amphibian fauna of this region.

Phylogenetic relationships of Pristimantis chamezensis
In our research, the integration of morphological and genetic data allowed us to establish that P. chamezensis is distinct from the other 13 Pristimantis species from Andean and sub-Andean forests on the eastern flank of the Cordillera Oriental. Taking into account the agreement between all phylogenetic analyses revealing a supported monophyletic group comprised of P. chamezensis, P. carranguerorum, P. bowara, P. lutitus, P. medemi, P. nicefori, and P. savagei, as well as the altitudinal (450-4170 m a.s.l.) and longitudinal distribution of those species along the Andean and sub-Andean forest on the eastern flank of the Cordillera Oriental (almost all are syntopic except by P. lutitus and P. nicefori from the western flank), it is probable that the origin of the new species and the radiation of the monophyletic group may have occurred at higher altitudes within this region. It might be possible that these Pristimantis lineages show the same pattern of recent diversification due to climatic changes, as seen in both, a high altitude dendrobatid frog (Hyloxalus felixcoperari Acosta-Galvis & Vargas-Ramírez, 2018) and a group of Andean anoles (Anolis heterodermus species group; Vargas-Ramírez and Moreno-Arias 2014) from the middle part of the eastern Cordillera.
Nevertheless, the generalized low support of the phylogenies emphasizes the need to increase the molecular dataset to reveal with confidence the evolutionary relationships within Pristimantis. This is clear from the recent changes in the phylogenetic position of several species (e.g., Hedges et al. 2008;Padial et al. 2014;Reyes-Puig et al. 2020). In addition, it is still required to incorporate a large number of unassigned Colombian taxa into evolutionary based species groups. There are about 117 species not yet analyzed using phylogenetic methods.
Our phylogenetic analyses unequivocally revealed that P. chamezensis is part of the subgenus Pristimantis. However, we do not force its allocation into one of the several species group (Hedges et al. 2008;Padial et al. 2014;Acevedo et al. 2020). Although our results validate some arrangements (e.g., conspicillatus or danae species groups; Fig. 3), some other individual assignments are weakly supported, and do not correspond to arrangements within the already proposed groups. Among the examples that we can identify, is the nesting of P. chamezensis with P. nicefori, which was formerly assigned within the unistrigatus group by Hedges et al. (2008) and later transferred to unassigned species group by Padial et al. (2014). Likewise, the close relationship of the chamezensis+ P. nicefori clade with the P. lutitus + P. medemi + P. carranguerorum clade (Fig. 3) is inconsistent with previous species groups assignments; P. medemi and P. carranguerorum were assigned to the conspicillatus species group by Hedges et al. (2008) and, later, validated by Padial et al. (2014). Additionally P. lutitus (Fig. 3), which was formerly assigned to the unistrigatus species group but subsequently transferred to an unassigned species group by Padial et al. (2014) and later inferred as sister to P. anolirex by .

Acknowledgments
We extend our thanks to all inhabitants of the Municipality of Chámeza, who by referendum selected the name of the species described here. Many thanks go to our local guides, José Pérez and Antonio Montaña. Natalia Novoa and Luis Daniel Prada offered their support and helped with fieldwork during the monitoring of Chameza's forest between 2010 and 2011. We also thank Dr José Rigoberto Ruiz Castillo and Dr Campo Elias Cardozo Tafur