Revision of the geographic distribution of three species of the montanus group of Liolaemus Wiegmann, 1834 (Reptilia: Squamata: Liolaemidae)

Liolaemus is one of the most diverse genera of iguanian lizards, with 239 species presently recognized (Uetz 2013). The genus is divided into two subgenera: Liolaemus (sensu stricto) and Eulaemus (Laurent 1985). The montanus group belongs to the nigriceps series of the montanus section of the subgenus Eulaemus (Fontanella et al. 2012). Lobo et al. (2010) listed 59 species for this group. However, they omitted to include L. lopezi IbarraVidal, 2005. Later, four new species were described for the group: L. cazianiae Lobo, Slodki & Valdecantos, 2010; L. halonastes Lobo, Slodki & Valdecantos, 2010; L. vulcanus Quinteros & Abdala, 2011; and L. porosus Abdala, Paz & Semhan, 2013. Therefore, the montanus group currently comprises 64 species. Here, I review the records of three species of the montanus group. For this purpose, I examined museum specimens (Appendix I) and reviewed the literature on each species. Examination of scales was made using lenses of different magnifications, and measurements were taken with a digital vernier caliper (±0.02 mm precision). Dorsal scales were counted between the occiput and the level of the anterior border of the hind limbs. Acronyms mentioned in this publication are: BMNH (Natural History Museum, Londres), CBF (Colección Boliviana de Fauna), FMNH (Field Museum of Natural History), KU (University of Kansas, Biodiversity Institute), MCN (Museo de Ciencias Naturales de la Universidad Nacional de Salta), MNHN-CL (Museo Nacional de Historia Natural, Chile), MNHN-Paris (Muséum National d’Histoire Naturelle, Paris), SDSU (San Diego State University Collection) and SSUC (Colección de Flora y Fauna Patricio Sánchez Reyes, Pontificia Universidad Católica de Chile). Finally, I discuss the records for these three species, ordered according to their geographical distribution (north to south).

Here, I review the records of three species of the montanus group.For this purpose, I examined museum specimens (Appendix I) and reviewed the literature on each species.Examination of scales was made using lenses of different magnifications, and measurements were taken with a digital vernier caliper (±0.02 mm precision).Dorsal scales were counted between the occiput and the level of the anterior border of the hind limbs.Acronyms mentioned in this publication are: BMNH (Natural History Museum, Londres), CBF (Colección Boliviana de Fauna), FMNH (Field Museum of Natural History), KU (University of Kansas, Biodiversity Institute), MCN (Museo de Ciencias Naturales de la Universidad Nacional de Salta), MNHN-CL (Museo Nacional de Historia Natural, Chile), MNHN-Paris (Muséum National d'Histoire Naturelle, Paris), SDSU (San Diego State University Collection) and SSUC (Colección de Flora y Fauna Patricio Sánchez Reyes, Pontificia Universidad Católica de Chile).Finally, I discuss the records for these three species, ordered according to their geographical distribution (north to south).Duméril & Bibron, 1837.This poorly known species was described for "Chili" based on a single specimen.Shortly after, this specimen was redescribed by Bell (1843) and illustrated by Duméril et al. (1854) (Figure 1).Since the type locality was poorly defined, several authors have considered this species Revision of the geographic distribution of three species of the montanus group of Liolaemus Wiegmann, 1834 (Reptilia: Squamata: Liolaemidae) as widely distributed in the highlands of Peru, Bolivia, Chile and Argentina (Koslowsky 1898;Donoso-Barros 1966).Later, Cei et al. (1980) rediscovered the holotype (MNHN-Paris 6860) and restricted the type locality to the "highlands of Peru and Bolivia".This decision is in accordance with D'Orbigny (1847), who collected the holotype and mentioned his route as follows: Valparaíso (Chile), Puerto de Cobija (Chile), Puerto de Arica (Chile), Tacna (Peru) and La Paz (Bolivia).In this trajectory, specimens of the montanus group have only been recorded between Tacna and La Paz (all current records are referred as L. signifer).Also, Cei et al. (1980) indicated the existence of one specimen from Huaracone, Puno department, Peru (MNHN-Paris 2503).However, Laurent (1984) highlighted notable differences between measurements of the body and tail indicated by Duméril and Bibron (1837) and Bell (1843), compared to the measurements reported by Cei et al. (1980) for the holotype of L. signifer.Although there are doubts regarding the specimen MNHN-Paris 6860 as really being the holotype of L. signifer, there is no a conclusive study and because Duméril and Bibron (1837) and Bell (1843) mentioned that there is only one type specimen, I consider the specimen MNHN-Paris 6860 as the holotype.On the other hand, despite the restriction of the type locality made by Cei et al. (1980), these authors keep L. signifer as a species distributed in the highlands of Peru, Bolivia, Chile and Argentina "considering the literature data" (my own translation, p. 923).Therefore, an analysis of the publications that mention L. signifer after 1980 is required.
Regarding Bolivia, Fugler (1989) reported a population from Pacajes Province, but did not list specimens.De la Riva et al. (1992) reported a population from Ulla-Ulla but also did not list specimens.Ibisch and Böhme (1993) recorded specimens from Palamani Pampa and Jachcha Kochi (Cochabamba).Aparicio (1993) recorded this species from Huaraco.Later, Dirksen and De la Riva (1999) reviewed all these locations and added more records based on museum collection catalogues, and also they provided coordinates  As for Peru, Abdala et al. (2008) listed specimens from Puno department: Lago Titicaca, Lago Titicaca road to Puno, 82 Km W Puno, Pichupichuin (8 km NW Huacallani), Caccahara (probably Cacachara), 12.9 km SW Limbani, 4 km NW of Juliaca, Huaylarco (88.5 km NE Arequipa; probably Huaylarco from Moquegua department) and 5 km W Tincopalca; Tacna department: 16 Km S Taratara and 5 km E Lago Suches; Moquegua department: Pampa Huatire; Arequipa department: Vicoyaca; Ayacucho department: 35 km NE Puquio; and Apurimac department: 40 Km SW Chalhuanca.Records from Arequipa department are very close to the distribution of Liolaemus annectens Boulenger 1901, a species previously considered a subspecies of L. signifer (Laurent 1992), but recently elevated to full species (with no data provided) by Pincheira-Donoso et al. (2008).It is necessary to carry out a study to clarify the relationships between these species and check the records of L. signifer from Arequipa.The same goes for the records from Apurimac and Ayacucho (Figure 2), which are close to the type locality of L. melanogaster Laurent 1998.Regarding Chile, Veloso and Navarro (1988), indicated that this species occurs in the highlands of the Antofagasta Region between 3600 and 4600 m, although they do not mention specific localities or specimens examined.Núñez and Jaksic (1992), included it for Chile without data.Interestingly, Núñez (1992) did not indicate locations for Liolaemus signifer in the catalog of geographic data of Chilean reptiles and snakes in the Museo Nacional de Historia Natural de Chile.However, Núñez and Veloso (2001) also mentioned that this species occurs in the highlands of Antofagasta Region without any data.Valladares et al. (2002) listed one specimen from Cariquima, Tarapacá (SDSU 1600), but this record is not from Cariquima since the specimen SDSU 1600 was collected in Puno (Peru) by W.E. Duellman (R. Langstroth, pers.comm.).Finally, Pincheira-Donoso and Núñez (2005) included L. signifer for Chile and provided a description.In the section Material Estudiado (p.450), these authors indicated that they examined the specimens BMNH-51.7.17.76 from "Chili" and BMNH-1902.5.29.63-73 from Uyuni, (Potosí, Bolivia).The locality "Chili" for the first specimen is too vague to sustain the presence of L. signifer in Chile.On the other hand, Uyuni (Potosí department) could represent an important distributional expansion.However, L. schmidti has been recorded in Potosí (FMNH 204525-26), and since Pincheira-Donoso and Núñez (2005) indicated that L. schmidti is a synonym of L. andinus without providing any data (Lobo et al. 2010) the status of the specimens from Uyuni should be revised.
Also, Pincheira-Donoso and Núñez (2005) indicate that they examined two specimens from Tacora (Visviri, Arica y Parinacota Region, Chile) (p.175), although these specimens were not included in the section about the material examined.I agree with Pincheira-Donoso and Núñez (2005) that the specimens MNHN-CL 2065-66 (two males) correspond to Liolaemus signifer, based on a comparison of these with several digital photographs of specimens of L. signifer from Bolivia, photographs of the holotype, and the examination of specimens of L. pleopholis Laurent 1998 (the most similar and geographically closest species).Laurent (1998) provides two diagnostic characteristics that distinguishe L. pleopholis from L. signifer: dorsal scales count and female Hellmich index (dorsal head scales).Table 1 provides the dorsal scale counts and one more scalation character that differs bewteen the species.I was unable to compare the female Hellmich index because the two specimens from Tacora are males.Additionally, Pincheira-Donoso and Núñez (2005) indicated that in Liolaemus signifer the scales of the dorsal dark spots have less developed keels than in L. pleopholis (p.168).However, I found that both species have developed keels on dorsal dark spots (Figure 3).
The dorsal pattern of coloration is similar in both species.The adult males of Liolaemus signifer have olivebrown, bright green or light brown dorsal color and may have a marked or diffuse pattern (Figure 4).Females have brown dorsal color and a defined pattern.This variability was previously described by Cope (1875), Pearson (1954) and Hellmich (1962).Color pattern in life can be seen in De la Gálvez and Pacheco (2009).On the other hand, Laurent (1998) indicated that the adult males of L. pleopholis have gray dorsal color with a diffuse pattern or uniform gray, and Veloso et al. (1982) indicated that males have uniform bright green or yellow coloration.Although in this study I have confirmed the different range of dorsal scales count proposed by Laurent (1998), it is necessary to conduct a study to clarify the relationship between both species and provide a clearer diagnosis, since apparently Liolaemus pleopholis is either a cryptic species or maybe a synonym of L. signifer.
Liolaemus signifer can be diagnosed from the geographically nearby species by the following combination of characters: 1) 61-87 dorsal scales (Laurent 1998); 2) rounded to subtriangular dorsal scales (smooth or slightly keeled); 3) juveniles, females and subadult males with dark spots over light brown dorsal color; 4) adult males with green shades and diffuse dorsal pattern.On the other hand, L. pleopholis has 87-98 dorsal scales (Laurent 1998), L. tropidonotus Boulenger 1902 has strongly keeled dorsal scales (Boulenger 1902;Núñez 2004) and the males of L. schmidti and L. forsteri Laurent 1982 never have green dorsal color.Diagnosis with respect to L. annectens is uncertain and must be provided in a future study.
Additionally, Pearson (1954) studied the ecology of L. signifer in a population from southern Peru (Lago Suche) and mentioned one specimen at 4800 m (Volcan Tutupaca, Tacna).De la Gálvez and Pacheco ( 2009) studied the effect of exploitation of L. signifer for traditional medicinal purposes.This species is listed as Least Concern (IUCN 2013).
In conclusion, this species has been recorded from Puno, Tacna and Moquegua departments (Peru), several locations from La Paz, Cochabamba and Oruro departments (Bolivia), and the Arica y Parinacota Region in Chile.All of these records are between 3700-4800 m (estimated with Google Earth).The records from Arequipa, Apurimac and Ayacucho departments (Peru) need to be reevaluated with an improved diagnosis of Liolaemus annectens and other species described from Peru.The records from the highlands of Antofagasta (Chile) and Potosí (Bolivia) are probably a misidentification of L. schmidti.Similar patterns can be seen in Pearson (1954) and Hellmich (1962) for specimens of L. multiformis (= L. signifer) from Peru and Bolivia, respectively.
Additionally, the species has been reported for Chile by Núñez and Veloso (2001) from Antofagasta Region, without indicating a specific location or specimens collected.Later, Pincheira-Donoso and Núñez (2005) indicated the existence of one specimen (MNHN-CL-2988) from Salar de Aguas Calientes at 4200 m, Antofagasta Region (Figure 5).Unfortunately, there are two places called Salar de Aguas Calientes in Antofagasta (northern and southern), both at 4200 m.These are separated by 30 Km, and both are approximately 70 Km NW of the nearest record of Liolaemus multicolor in Argentina (Río Rosario de Susques).Pincheira-Donoso and Núñez (2005) indicate that they are not sure which species this specimen belongs: "a comparative review of many adults of L. (Eulaemus) multicolor from Argentina (DBCGUCH) and the specimen from Chile, allows us establish some doubts on its real status, which however may be explained in that it is a specimen of small size (SVL = 55.7 mm), which would correspond to a juvenile; nevertheless, it is interesting to open the possibility that this specimen indeed corresponds to a female of L. (Eulaemus) andinus, problem that could be solved by a more detailed examination of available juveniles of L. (Eulaemus) multicolor from Argentina..." (my own translation, p. 164).
In fact, the record from Salar de Aguas Calientes (Chile) corresponds to a juvenile male of Liolaemus molinai Valladares, Etheridge, Schulte, Manríquez & Spotorno 2002; and not to L. multicolor or L. andinus, because: (1) The dorsal coloration pattern of the specimen from Aguas Calientes is the same as that of males of L. molinai (Figure 6); (2) The specimen MNHN-CL 2988 has smooth paravertebral scales, whereas L. multicolor has keeled paravertebral scales (Cei 1993); (3) The specimen MNHN-CL-2988 (SVL = 55.7 mm) has well developed precloacal pores, whereas in juvenile males of L. multicolor the precloacal pores begin to develop when males reach approximately 55 mm (Valdecantos and Lobo 2007) (Figure 7); (4) The specimen MNHN-CL 2988 has dark spots scattered on the belly like males of L. molinai, whereas L. multicolor specimens have an immaculate belly or dark lines (Cei 1993); (5) The northern Salar de Aguas Calientes is only 15 km NW of Los Farellones de Tara, type locality of L. molinai and the southern Salar de Aguas Calientes is only 40 km; therefore, in any case, the distance between these two populations is small; (6) The specimen MNHN-CL 2988 has precloacal pores, whereas L. andinus has no preclocal pores (Koslowsky 1895); ( 7) The specimen MNHN-CL 2988 has 86 scales around midbody, whereas L. andinus has 98-110 scales around midbody (Koslowsky 1895;Laurent 1982).
Another major problem here is the lack of diagnosis between Liolaemus molinai and L. erguetae Laurent 1995 (the most similar species), since Valladares et al. (2002) did not provide a diagnosis to distinguish them.Both species have similar color pattern and females have precloacal pores.It is necessary to conduct a comparative study of both species.It is possible that L. molinai may be a junior synonym of L. erguetae (R. Langstroth, pers.comm.).
In conclusion, Liolaemus multicolor should be considered endemic to Argentina and the record from Salar de Aguas Calientes (Chile) corresponds to L. molinai.
Liolaemus andinus Koslowsky, 1895.The species was described from the highlands of Catamarca province (Argentina), between 3000 and 4000 m.Koslowsky (1895) pointed out the absence of precloacal pores in both sexes (p.364) and that the species has 105-110 midbody scales (p.365).Laurent (1982) indicated that the female holotype is deposited in the Museo de La Plata (Argentina), without any museum number.The specimen lacks precloacal pores  and has 98 midbody scales.Additionally, Laurent examined a male of uncertain status (not a type) from Catamarca province that, according to him, could correspond to another species (p.90).According to Ferraro and Williams (2006) the type series is lost.The species is not listed in the IUCN Red List (IUCN 2013).
Currently, the true identity of Liolaemus andinus and the specific location where it was collected are unknown.Schulte et al. (2000) mentioned specimens from La Rioja and Jujuy provinces (Argentina).Lobo et al. (2010) pointed out that the record from Jujuy is probably L. multicolor (p.7), and since the type locality of L. andinus is not known, the record from La Rioja should be taken with caution.Additionally, Abdala et al. (2008) recorded specimens from Aguas Calientes, Las Grutas (20 Km from Paso San Francisco), near Río Chaschuil (42 Km SE from Paso San Francisco), 68 Km NW from Río San Francisco and Cazadero Grande, all in Catamarca province, Argentina.However, although L. andinus is a species of uncertain status, they did not provide data to support these records.Moreover, Abdala et al. (2008) pointed out that females of L. andinus have no precloacal pores (p.461), but based on the same specimens Abdala et al. (2013) pointed out that females of L. andinus have up to four precloacal pores (x = 2) (p.1568).Abdala et al. (2013) indicated that these specimens have 76-88 midbody scales, whereas Koslowsky (1895) and Laurent (1982) (they examined the currently lost type specimens of L. andinus) indicated 105-110 and 98 midbody scales, respectively, for L. andinus.Since the two major diagnostic characters of L. andinus do not match the information provided by Abdala et al. (2013), these specimens should not be considered conspecific with L. andinus.
The same applies to records from Chile.Halloy et al. (1991) indicated the presence of Liolaemus andinus in Cuesta Colorada (4500 m), 20 Km E of Maricunga, Atacama Region, and published a photograph of one specimen (p.64).They indicated that L. andinus has 95-110 scales around midbody, however they did not list specimens collected and did not indicated if males have precloacal pores or not.Later, L. andinus was included in Chile by Núñez and Veloso (2001).Thereafter, Pincheira-Donoso and Núñez (2005) included L. andinus for Chile, extended its distribution to the Antofagasta Region and indicated that L. molinai and L. schmidti are synonyms of L. andinus (without any data supporting their claim).These synonyms were rejected by Lobo et al. (2010).I agree with this because the males of L. schmidti and both sexes in L. molinai have precloacal pores, whereas L. andinus lacks precloacal pores in both sexes according to Koslowsky (1895).Later, Abdala and Quinteros (2008) listed one specimen from Maricunga (FML 2553).Finally, L. andinus was included for the Atacama Region by Valladares (2011) and Troncoso-Palacios and Marambio (2011) based on a literature review.Additionally, Troncoso-Palacios and Marambio (2011) published a photograph of one specimen from Laguna Verde (40 Km E of Cuesta Colorada), but this male specimen has precloacal pores and assigning it to L. andinus was incorrect.
The photograph of Liolaemus andinus published by Halloy et al. (1991) strongly resembles a female L.
rosenmanni Núñez & Navarro, 1992 (Figure 8), and Cuesta Colorada is only 20 Km E of Maricunga, where the allotype of L. rosenmanni (DBGUCH-0916) was collected, and 22 Km E of Laguna Santa Rosa where Moreno et al. (2000) recorded L. rosenmanni.Both records from Maricunga, Halloy et al. (1991) and Abdala and Quinteros (2008), are probably specimens of L. rosenmanni.On the other hand, it is necessary to collect specimens from Laguna Verde and carry out a formal study to clarify its taxonomic assignment.
Moreover, males of Liolaemus rosenmanni always have precloacal pores (Núñez and Navarro 1992).The two males of L. rosenmanni that I examined (from La Ola) also have precloacal pores.According to Núñez and Navarro (1992), L. rosenmanni has more than one hundred midbody scales (unspecified range or count).In a previous work I found a range of 86-92 midbody scales for L. rosenmanni from La Ola (Troncoso-Palacios and Ferri-Yáñez 2013) and according to Mr. D. Esquerré, the holotype of L. rosenmanni has 93 midbody scales (pers.comm.).Therefore, the midbody scale range is 86-93 for L. rosenmanni and 98-110 for L. andinus.These characteristics allow me to reject the hypothesis that L. rosenmanni coud be a synonym of L. andinus because males of L. rosenmanni have precloacal pores and both species have different midbody scale count ranges.
In conclusion, Liolaemus andinus should be considered endemic to Catamarca, Argentina.However, the type specimens are apparently lost, and if it is not possible to collect topotypic specimens matching the description of Koslowsky (1895) to designate a neotype, L. andinus should be considered a nomen dubium, since it is a name of unknown application (ICZN 1999).Records from Jujuy (Argentina) and Atacama (Chile) probably correspond to other species.
I hope this article will help to clarify the distribution of these species of the montanus group, whose study is hampered by their high altitude ranges and sometimes by the difficulty of accessing such extreme locations.Also, I hope this article will serve to call attention to the confusion that can be the result of records of poorly known species without voucher specimens or without defining the locality of the collection.At the same time, I hope this article helps to stimulate the study of some populations whose taxonomic status still remains uncertain.
for those they could identify: Road between Caracoles and Quime, Cordillera Tres Cruces, road between La Paz and Tiahuanaco, La Cumbre, 130 km NE of La Paz, 35 mi NW of La Paz, Potone, Tiahuanaco, Esperanza (Province Pacajes), 7 km E of Ulla Ulla, Caupolicán (Hacienda Corpaputu, 10 km E of Achacachi), Patacamaya, Altiplano (Oruro), 5 km NE of Oruro, 15 km NE of Oruro, 60 km NE of Oruro, Hacienda Huancaroma, 8 km S of Eucaliptos, Chocaya, 30 mi N of Potosí and 7 km S of Potosí.Later, Abdala et al. (2008) listed specimens from Confital (Cochabamba).Finally, De la Gálvez and Pacheco (2009) indicated the presence of L. signifer at 20 locations between Titicaca and La Paz and in southern La Paz.The authors also provide a map, but unfortunately did not refer to the names and coordinates for these locations.Even though many of these records are not specific (only indicating the province), they are all in the department of La Paz and the neighboring departments of Cochabamba and Oruro in areas bordering La Paz department, except the records from the vicinity of Potosí and Chocaya, located in Potosí department, which are more than 220 Km S (straight line) from any other location (Figure 2).I believe that these records are doubtful and need confirmation, especially considering that the similar species Liolaemus schmidti Marx 1960 has been recorded in Potosí department (Dirksen and De la Riva 1999).

Figure 3 .
Figure 3.Comparison between dorsal scales of Liolaemus signifer (left, MNHN-CL 2065) and L. pleopholis (right, MNHN-CL 3950).Note that both species have more developed keels on the dark spots on the dorsum.