Liolaemus ubaghsi, Esquerré, Damien, Troncoso-Palacios, Jaime, Garín, Carlos F. & Núnez, Herman, 2014

Esquerré, Damien, Troncoso-Palacios, Jaime, Garín, Carlos F. & Núnez, Herman, 2014, The missing leopard lizard: Liolaemus ubaghsi sp. nov., a new species of the leopardinus clade (Reptilia: Squamata: Liolaemidae) from the Andes of the O’Higgins Region in Chile, Zootaxa 3815 (4), pp. 507-525 : 509-522

Liolaemus ubaghsi sp. nov.

1992 Liolaemus leopardinus leopardinus, Núñez. Smith. Herpet. Info. Serv. 91, 10.

2005 Liolaemus elongatus, Pincheira-Donoso and Núñez. Pub. Oc. Mus. Nac. Hist. Nat. , Chile, Santiago 59, 260.

Holotype. MNHNCL-3813 ( Fig. 1 ). Male, collected by Herman Núñez, Carlos Garín and Daniel Pincheira- Donoso at Chapa Verde, between the ski field and reservoir, Libertador Bernardo O’Higgins Region, Chile. 34°03'S 70°26'W, 2210 masl. On May 22-23, 2003.

Paratypes. MNHNCL-3812, 3814 and 3816 males; MNHNCL-3808, 3809, 3810, 3811, and 3815 females ( Fig. 2 ). Same collection data as the holotype. SSUC Re-491 and 492 females, collected by R. Thomson and G. Ugalde. Tranque Barahona, El Teniente Mine, Libertador Bernardo O’Higgins Region, Chile. On April the 15th, 2008. Measurements in Table 1 .

Etymology. This species is named after Prof. Georges Ubaghs (1916-2005), Belgian paleontologist, who made some of the greatest contributions to the knowledge of Palaeozoic echinoderms in the 20th century. He is also the great-grandfather of the first author of this paper ( DE). We propose the common name “Ubaghs’ leopard lizard” in English and “Lagarto leopardo de Ubaghs” in Spanish.

Diagnosis. Liolaemus ubaghsi sp. nov. belongs to the leopardinus clade. This group is composed by medium to large lizards (maximum SVL greater than 89 mm), with prominent neck folds, 72–93 scales around mid-body, juxtaposed or subimbricated rhomboidal keeled dorsal scales, with interstitial granular scales around them, nasal scales not in contact with the rostral scale (with some exceptions), undifferentiated auricular scale and parietal scales with a similar size as interpatietal (with some exceptions, especially in L. frassinettii ). Even though the pattern of each species is different, it always involves some kind of dark mottled pattern on a lighter (varying from beige, to olive brown and dark brown) background (see Figs 3–6 ). All species have leopard-like spots on the dorsum or at least on dorsal surface of the tail (with the exception of Liolaemus ubaghsi sp. nov)

Liolaemus ubaghsi sp. nov. is a medium to large, sturdy lizard. It has a mean SVL of 83.3 mm, and a maximum of 89.6 mm. The head is longer than it is wide, and the neck is as wide or wider than the head due to prominent transversal neck folds. On average, the tail is 1.46 times the length of the body (SVL). It has 72–87 scales around midbody, 62 to 83 dorsal scales (through the vertebral line, from the occiput to the level of the anterior surface of the femoral region) and 111 to 137 ventral scales (from tip of the snout to the cloaca).

It differs from L. valdesianus and L. frassinettii by its dorsal pattern. L. ubaghsi sp. nov. displays an occipital stripe or sometimes a vertebral line and has white dots dispersed on the dorsum, instead of the light green transversal stripes present in L. valdesianus and L. frassinettii . Also, the general background coloration in L. ubaghsi sp. nov. is ochre, while in L. valdesianus and L. frassinettii is olive brown and has a more gray or greenish hue. Furthermore, L. valdesianus and L. frassinettii have leopard-like spots on the dorsal side of the tail, whereas these are absent on L. ubaghsi sp. nov. It also differs from L. valdesianus by having more lamellae on the fourth toe, although there is a small overlap (28–35 in L. ubagshi sp. nov. vs. 25–28 in L. valdesianus ), and because the dorsal scales on L. valdesianus are juxtaposed or subimbricated, leaving large spaces between them with abundant interstitial granules (more than in any species of the leopardinus clade). These scales are subimbricated in L. ubagshi sp. nov., with less interstitial granules (see Fig. 7 ). Moreover, the infralabial scales in L. ubagshi sp. nov. and L. frassinettii are notoriously enlarged in relation to the supralabial scales, while is L. valdesianus the infralabials are barely larger than the supralabials (see Fig. 8 ). The dorsal scales on L. ubaghsi and L. frassinettii are notoriously more keeled than on L. valdesianus and L. leopardinus . It additionally differs from L. frassinettii because this species has a black spot that begins at the axilla and continues until about the middle of the flanks, instead of the complete dark flanks in L. ubaghsi sp. nov.. It strongly differs from L. leopardinus and L. ramonensis by its dorsal pattern. Although these two species may present a vertebral line, they never exhibit the occipital wide stripe often observed in L. ubaghsi sp. nov., and their flanks are not as dark as the flanks in L. ubaghsi sp. nov. Aditionally, L. leopardinus never has white dots scattered on the dorsal region, and it has big black leopard-like hollow spots on the dorsal region, which are absent in L. ubaghsi sp. nov. (as well as in the rest of the clade). In the same way, L. ramonensis has small black leopard-like spots absent in L. ubaghsi sp. nov.. In addition to all this diagnostic traits, L. ubaghsi sp. nov. is slightly smaller (maximum SVL= 89.6 mm) than L. valdesianus (95 mm), L. ramonensis (94.9 mm) and L. leopardinus (98.2 mm).

Description of the holotype. Male, with an SVL of 88.2 mm. Axilla-groin distance 39.3 mm. Tail complete, not regenerated, length, 113.8 mm. Left forelimb length, 32.1 mm. Left hind limb length, 51.3 mm. Head length, 21 mm. Head width, 17.1 mm. Head height, 11.3 mm. 4 small precloacal pores.

Rostral scale rectangular, 2.7 times wider than high, in contact with six scales, and not with the nasals. Nasal scales hexagonal, anteriorly elongated, with the nostril occupying the posterior half of the scale’s surface. Two postrostrals. Four internasal scales, the medial ones larger, and the side ones transversally divided. Two frontal azygous scales on the snout, the anterior one about double the size of the posterior surrounded by six frontonasals. Three prefrontals, half the size of the frontal. Undivided frontal scale. Two small postfrontals. Interparietal scale pentagonal, with a white “pineal eye” in the middle, in contact with six scales. Parietals hexagonal, about the same size as interparietal. Occipital scales polymorphic, juxtaposed and smooth. Supratemporals larger than occipitals, also polymorphic, juxtaposed and smooth. Four enlarged supraoculars. Eleven/twelve (left/right) scales forming the supraorbital semicircles. Scale organs abundant in all the head, however a much higher concentration on the anterior tip. Loreal length (from anterior tip of the eye to rostral), 6.8 mm. Eye length, 5.6 mm. Five/six loreal scales. Five elongated and overlapping superciliaries. Large subocular, which spans the whole length of the eye. Twelve/thirteen upper and twelve/thirteen lower squared palpebral scales, each one with a scale organ on it. One row of lorilabial scales. Seven/six supralabial scales, on both sides the fifth one posteriorly curved upwards, a diagnostic trait of the Liolaemus (sensu stricto) subgenus. Auditory meatus 4.3 mm high and 2.7 mm wide. Auricular scale not enlarged. Temporal scales polygonal, juxtaposed and smooth. Mental scale wider than rostral, in contact with four scales. Four pairs of postmental scales, the first pair in contact with each other, the second one separated by two scales. Four/five enlarged infralabial scales. Gular scales rounded, smooth and imbricate. Two strong transversal neck folds, on top of them a longitudinal fold. Neck scales granular, small, smooth and juxtaposed, with much smaller granular scales in the interstitial space.

Dorsal scales rhomboidal, subimbricated and keeled, without mucrons, around the same size as the ventral scales. There are some visible small granular scales in the interstitial space. Lateral scales become more rounded and with less developed keels, also more granular interstitial scales become visible. There is a longitudinal posthumeral fold on the flanks. Scales are granular at limb insertions. Ventral scales are rounded, smooth and imbricate. Scales around midbody: 80. Dorsal scales: 68. Ventral scales: 129. Dorsal scales of the arm vary from rhomboidal to rounded, and are imbricated and slightly keeled. Dorsal scales of the forearm are rounded, imbricate and smooth, with slight keels towards the posterior surface. Dorsal scales of the hand are rounded, imbricate and smooth. Ventral scales of the arm are granular and juxtaposed, with much smaller granular scales on the interstitial space. Ventral scales of the forearm are rounded, imbricate and slightly keeled. Palmar scales are rhomboidal with rounded tips, imbricate and slightly keeled, with jagged edges. The lamellae on the fingers are rectangular and imbricate, with three keels each. There are 24 lamellae on the third finger and 22 on the fourth finger. Dorsal femoral scales are rhomboidal with rounded tips, imbricate and slightly keeled. Dorsal tibial scales are similar to dorsal femoral ones but have stronger keels. Dorsal scales of the foot are rounded, imbricate and slightly keeled. There is a patch of granular scales on the ankle. Ventral femoral and tibial scales are rounded, imbricate and smooth. Plantar scales vary from rounded to rhomboidal with rounded tips, and are imbricate and keeled, with jagged edges. Lamellae of the toes are rectangular and have three keels each. There are 30 lamellae on the fourth toe. Dorsal caudal scales are rhomboidal, imbricate and slightly keeled. Ventral caudal scales are rhomboidal, imbricate and smooth.

Color and pattern in preservative. Background dorsal coloration is ochre with a very slight greenish hue due to loss of pigmentation. Dorsal head color is a bit darker and more brownish than the body. The dorsum has small dots the size of one scale scattered on it. More concentrated on the 10 to 12 mid-dorsal scales, forming some sort of occipital stripe. The white scattered dots observed on live specimens are not seen in preserved ones like the holotype. The flanks are dark brown. The tail is mottled on its dorsal surface. There are few black dots on the dorsal surface of the limbs. The ventral region is gray, with a slightly darker gular region.

Color in life. Based on photographs of live individuals. The general coloration of the dorsal surface of the body, head, tail and limbs is brown, with dark brown flanks. Black dots are scattered along the dorsum at regular intervals. The dots may form a wide occipital stripe or a vertebral line as described above. Tiny white dots, made of a single scale are also scattered and may form transversal lines in some specimens. The ventral region is bright lemon yellow or orange. The gular surface is white with dark spots on it (see Fig. 9 ).

Variation in the paratypes.

The body measurements for the four males (including the holotype) have the following mean and extreme values (in mm): SVL, 85 (77.5–89.6); axilla-groin distance, 37.4 (35.2–39.3); left forelimb length, 31.7 (30.5–33.1); left hind limb length, 49.8 (46.5–51.8); head length, 19.7 (17.8–21); head width, 15.7 (14–17.1); head height, 10.1 (8.9–11.3); loreal length, 6 (5.1–6.8); eye length, 5.3 (4.5–6.2); auditory meatus height, 4.5 (3.8–5.1); auditory meatus width, 2.7 (2.6–3.1). Tail length was not included because all the male paratypes had the tail cut off or regenerated. The same measurements for the six adult female paratypes are as follows (juvenile female SSUC Re-491 is excluded from these measurements): SVL, 85.6 (81.2–88.6); axilla-groin distance, 41.2 (37–45.6); tail length, 131.9 (119.4–141.4); left forelimb length, 29.5 (27.2–31.7); left hind limb length, 46.8 (43.1–51.2); head length, 19.3 (18.5–21); head width, 14.2 (13.6–15.1); head height, 9.5 (9.1–9.4) loreal length, 5.9 (5.3–6.9); eye length, 4.8 (4.3–6); auditory meatus height, 4.2 (3.7–5.5); auditory meatus width, 2.3 (1.6–2.8). Based on this measurements and the limited sample size it is difficult to determine whether this species has sexual dimorphism or not. There are no clear differences in body size, however males do have proportionally larger heads than females (head length, F=8.03, p =0.025; head width, F=9.6, p =0.017; head height, F=6.6, p =0.037).

The four males have 73–80 scales around midbody, 64–73 dorsal scales and 111–129 ventral scales. All males have four precloacal pores except for one that has no visible pores, however this one is not an adult and the pores might not have developed yet (MNHNCL-3814). The three male paratypes have the following variation in squamation in relation to the holotype: rostral scale 3.8–4.6 mm wide and 1.8–2 mm high. Nasals barely touching the rostral with the anterior tip in one specimen. Anterior frontal azygous scale of the snout smaller than posterior in one specimen, and only one frontal azygous scale in two specimens. Four frontonasals in one specimen. Prefrontals with similar size as the frontal in two specimens. Frontal divided transversally in one specimen. Three postfrontals in one and four in another specimen. Interparietal in contact with five scales in one and with seven scales in another specimen. Parietals with irregular shapes instead of hexagonal in all three male paratypes. Four to six supraoculars. Five to six superciliaries. Twelve to fifteen upper and thirteen to fourteen lower palpebrals. Mental as wide as rostral in one specimen. Five pairs of postmentals in two specimens. Smooth scales on the flanks in two specimens. Rhomboidal with rounded tips scales on the dorsal surface of the forearm in one specimen, and slightly keeled in two specimens. Smooth edge on the palmar scales in one specimen. Lamellae on the third finger: 21 to 23. Lamellae on the fourth finger: 22 to 25. A weak mucron on the dorsal femoral scales in one specimen. Rounded dorsal tibial scales in one specimen. Rhomboidal with rounded tips dorsal foot scales in two specimens and smooth in another specimen. Lamellae on the fourth toe: 28 to 34. The seven female paratypes have 72–87 scales around midbody, 62–83 dorsal scales and 114–137 ventral scales. All females lack precloacal pores. The female paratypes have the following variation in squamation in relation to the holotype: rostral scale 3.9–4.3 mm wide and 1.5–1.8 mm high. Nasals barely touching the rostral in three specimens. One frontal azygous scale on the snout in two specimens. Four frontonasals in two specimens. Prefrontals similar size as frontal in four specimens. Frontal divided longitudinally in one and transversally in two other specimens. Three postfrontals in two specimens. Interparietal in contact with five scales in one specimen, with seven scales in three specimens and with eight scales in another specimen. Interparietal smaller than parietals in one specimen. Parietals irregularly shaped in five specimens. Four to seven supraoculars. Four to seven loreal scales. Five to eight superciliaries. Five to eight supralabials. Nine to thirteen upper and ten to eleven lower palpebrals. Mental as wide as rostral in two specimens. Five pairs of postmentals in six specimens. Four to six infralabials. Smooth scales on the flanks in five specimens. Dorsal scales of the arm rounded in one specimen and smooth in three specimens. Ventral scales of the forearm smooth in four specimens. Palmar scales with smooth edges in one specimen. 20 to 27 lamellae on the third finger. 22 to 27 lamellae on the fourth finger. Dorsal femoral scales rounded in four specimens and smooth in two specimens. Dorsal tibial scales rounded in four specimens. Dorsal scales of the foot rhomboidal with rounded tips in one specimen and smooth in another specimen. 30 to 35 lamellae on the fourth toe.

All preserved specimens have the basic ochre background coloration with black dots scattered along the dorsum and dark flanks. Some specimens, however, have a wide dark occipital stripe along the vertebral region, and others have a thinner black vertebral line instead. The dorsal surface of the head is mottled with black dots in some specimens. There is a high variation in ventral melanism. Some individuals, similar to the holotype, exhibit a low level of melanism, with a gray ventral region, while others exhibit varying degrees of mottled bellies, from separated small black dots scattered all over the ventral surface of the body, head, limbs and tail, to a more intense mottled pattern that appears as white reticulation between the dots. The gular region is always darker than the belly. In some specimens the gular region is only light gray, while in others it is uniformly black.

Distribution. Most probably endemic to the Andes East of Rancagua City in Chile (see Fig. 10 ). From museum collection data we know it is found between the type locality (34°03'32.8''S 70°26'05.3''W, 2210 masl) at the Chapa Verde Ski Center and the Sapos Reservoir (S 34 04’ W 70 22’ 1870 m 34°04'S 70°22'W, 1870 masl). All these localities and probably the distribution of L. ubagshi are part of El Teniente copper Mine, property of Codelco (Corporación Nacional del Cobre de Chile), a Chilean state mining company. For this reason public access and therefore the study of this species is very difficult. Another population from the mountains of Río Clarillo National Reserve might also belong to this species (see Discussion).

Natural history. Species with viviparous reproduction, determined by examining two pregnant female specimens, not included as paratypes (MNHNCL-605 and 608). A group of nine Liolaemus ubaghsi sp. nov. was found in brumation underneath a rock in winter along with other five L. cf. bellii ( Gray 1845) . Liolaemus ubaghsi sp. nov is found in sympatry with this species and with L. nigroviridis ( Müller & Hellmich 1932) . It lives in a high Andean environment, characterized by short shrubs and a rocky landscape ( Fig. 11 ).

Range 62.7– 27.1– 111.6– 23.9–3 41.7–51 17.8– 13.2– 8.9– 72–87 62–83 111– 28–35

89.6 45.6 141.4 3.1.8 21 16.8 11.3 137 Organisms adapted to high elevation and that are somewhat unable to live at lower elevations, like the leopard lizards, can be isolated, and thus undergo speciation through “sky islands”, where low elevation valleys serve as geographical barriers ( Knowles 2000; Masta 2000; Shepard & Burbrink 2008; Robin et al. 2010). Interglacial periods like the present one restrict high elevation species such as the leopard lizards to mountains. Such divergence among mountain-top populations is also observed in another liolaemid taxon, L. nigroviridis , with apparently deep divergence ( Cianferoni et al. 2013), which is mostly co-distributed with the species of the leopardinus group, and may share a similar evolutionary history. On the other hand, L. nigroviridis can also be found at lower elevations than the leopard lizards, therefore secondary contact is more likely to have happened in L. nigroviridis , which would explain why this species exhibits much less morphological divergence than the species of the leopardinus clade. The leopard lizards probably descend from a common ancestor belonging to the elongatus clade, a group mainly distributed on the eastern slope of the Andes. It would seem logical that the lineage underwent dispersal to the low elevation valleys of Central Chile during the glacial periods of the Pleistocene ( Rabassa & Clapperton 1990), and that different populations retreated to different high elevation locations during interglacial periods where they could diverge without gene flow and therefore undergo speciation ( Hellmich 1934; Fuentes & Jaksic 1979).

Pincheira-Donoso and Núñez (2005) mentioned that the population now described as L. ubaghsi displayed traits belonging to both L. valdesianus and L. elongatus . While this species is certainly not L. elongatus , mainly because its southern distribution is more than 400 km north of the northern range limits of L. elongatus , with its northern boundary at Neuquén and Agrio Rivers in Argentina ( Morando et al. 2003), it most certainly shares a close common ancestor with L. valdesianus . The phylogenetic analysis performed by Esquerre et al. (in prep.), which will be published elsewhere, recovers the leopardinus group as monophyletic, including L. ubaghsi , and this clade is in turn the sister group of the elongatus clade. The analysis however does not resolve the relationships within the clade except by locating L. frassinettii as the sister taxon of the rest of the group, which makes sense seen from a biogeographic scope. Liolaemus frassinettii represents a lineage that colonized the Cordillera de la Costa, a mountain range separate from the Andes by a low elevation valley, where no species of the elongatus (thus, also the leopardinus ) clade live. It would appear, by looking at the morphology, pattern and distribution of the remaining species of the leopardinus clade, that there is a morphological gradient from L. ubaghsi to L. leopardinus . Evidence from their morphology and DNA suggest they are recently diverging lineages, in comparison to other Liolaemus clades. The genetic distances for the mitochondrial marker cytb between the Andean leopardinus clade (all except for L. frassinettii ) are quiet shallow ( Table 3 ). Despite this, consistent morphological and color pattern differences in addition to great geographic barriers strongly suggest their evolutionary independence as lineages. First, even though the Andean habitat of L. ubaghsi is less than 50 km away from the Cordillera de la Costa habitat of L. frassinettii , the 15 km-wide valley between the base of both mountain ranges is at an elevation of merely 500 masl. On the other hand there is a mid-elevation connectivity between the two Cordilleras formed by Cuesta de Chada and Challay Hill that goes from 400 and almost 1200 masl. Nevertheless there is no record of any of the leopard lizards living bellow 1800 masl. Second, L. valdesianus or any other leopard lizard have not been found on the eastern side of the mountains west of the Maipo River Canyon, hence the 3500–4000 masl mountain tops and the 1500 masl Maipo River Canyon stand between L. ubaghsi and L. valdesianus . For these reasons, we think that gene flow has not occurred in recent times, and that these taxa are independently evolving populations, and the phenotypic differences serve as an operational species delimitation criterion ( De Queiroz 2007). We think this small species group is an excellent model for studying early speciation and that multiple loci molecular studies will shed more light on their phylogenetic and phylogeographic history.

There still remain a number of problematic populations of the clade that need to be studied, however, these issues are beyond the scope of the current article and should be part of further investigations. A population identified as L. leopardinus was found in the high parts of Río Clarillo National Reserve, about 20 km north of the terra typica of L. ubaghsi ( Díaz & Simonetti 1996; Díaz et al. 2002). No specimens of this locality are found in museum collections, and by examining photographs of live specimens it could be determined that this population does not belong to L. leopardinus ( Hamasaki & Troncoso-Palacios 2011) . The true identity of this population remains unresolved until some specimens can be examined, nevertheless judging only by the locality and a poor quality photograph taken by Mr. I. Díaz, it should be expected to be a population of L. ubaghsi , and we suggest provisionally naming this population L. cf. ubaghsi . There is another population in El Yeso Reservoir (33°41'S 70°07'W) standing between the distributions of L. valdesianus and L. ramonensis . Núñez (1992), assigned this population to L. l. leopardinus , however Núñez & Jaksic (1992) mentioned that this population may represent genetic connectivity between L. valdesianus and L. ramonensis , additionally noting that the specimens collected at that locality were more similar to the latter. In contrast, Hamasaki & Troncoso-Palacios (2011) found the population not distinct from L. valdesianus , proposing that it should be referred as L. cf. valdesianus for the moment. We examined specimens from this locality and we think they should be considered as L. valdesianus . There is also a population phenotypically similar to L. ubaghsi from which there are two specimens in the MNHNCL and photographs (provided by D. Demangel) from Alto Huemul, in the Andes south-east of San Fernando City, and near 100 km south of the type locality of L. ubaghsi . With that material it was not possible to determine the exact identity of the population, and it also remains an important aspect of our future research. Based on specimens observed and photographed in the wild, we would also like to extend the northern distributional limit of L. valdesianus to the Olivares River (33°24'S 70°08'W and 33°17'S 70°08'W), almost 40 km north of El Yeso Reservoir. Finally, L. leopardinus has been mentioned recently for two locations without listing specimens examined and therefore need confirmation: Riecillo Valley, Valparaiso Region ( Núñez et al. 2010) and Cerro Carpa, Metropolitan Region ( Hamasaki & Troncoso-Palacios 2011). We examined photographs from Riecillo Valley, Valparaiso Region (33°03'S 70°22'W) provided by C. Celedón, and have identified it as L. leopardinus . Close examination of specimens of that population is necessary, but we provisionally propose extending the northern distribution limit of L. leopardinus to that locality.

As an additional remark, close inspection of the L. frassinettii type material, revealed that one of the two paratypes was in fact a juvenile female of L. nigroviridis (MNHNCL-1598), but poor preservation condition of the specimen made it difficult to identify correctly. This species was described based only in three females (two by correction), even though its species status is supported by genetic evidence (Esquerré et al. in prep), geographical isolation and new examined specimens, descriptions of the male and more data about species variation are needed.

To conclude, we think that the biogeographic and phenotypic characteristics of L. ubaghsi are enough to determine that it is a divergent and independent evolutionary lineage of the leopardinus clade. Nevertheless there seem to be far more populations of the clade in the central Chilean Andes than the ones commonly known and need to be included in our further studies of the group. It also appears to be an excellent model to study several evolutionary processes since speciation is likely to be happening at different levels between the different lineages.

Figures

FIGURE 1. Holotype of Liolaemus ubaghsi. MNHNCL 3813.

FIGURE 2. Type series of Liolaemus ubaghsi. Going from left to right: MNHNCL- 3808 - 3816 (SSUC Re- 491 and SSUC Re- 492 not included). Top: dorsal view. Bottom: ventral view.

FIGURE 7. Comparison of the dorsal scales on some species of the leopardinus clade. All figures are 8 mm wide. A: Liolaemus ubaghsi sp. nov. (SSUC Re- 492). B: L. valdesianus (SSUC Re- 363). C: L. frassinettii (SSUC Re- 80). D: L. leopardinus (SSUC Re- 364). What this figure illustrates is that the dorsal scales of L. valdesianus tend to be much more juxtaposed and with more interstitial space filled with granular scales between them than in the rest of the leopardinus clade.

FIGURE 8. Comparison of supralabial and infralabial scales on some species of the leopardinus clade. A: Liolaemus ubaghsi sp. nov. (SSUC Re- 492). B: L. valdesianus (right, SSUC Re- 363). C: L. frassinettii (SSUC Re- 80). D: L. leopardinus (SSUC Re- 364). What can be seen here is that infralabial scales in L. ubaghsi sp. nov. and L. frassinettii are notoriously enlarged in comparison with L. valdesianus. Infrabalabials can be enlarged in L. leopardinus too, as in the specimen shown here.

FIGURE 9. Dorsal and ventral views of live adult specimens of Liolaemus ubaghsi. Because ventral view specimens were held in hands while being photographed, the complete body is not visible in them. Photos: Alejandra Alzamora.

FIGURE 10. Map of the Andes Cordillera surrounding Santiago City. Coordinates are in decimal system. Symbols depict known localities from museum specimens, literature and photographs (in the case for L. cf. leopardinus). Stars, L. ubaghsi; pentagons, L. frassinettii; triangles, L. valdesianus; squares, L. ramonensis; circles, L. leopardinus. Symbols with question marks represent localities for uncertain taxonomic identity, but probably belonging to the species of that symbol (e. g. star with question mark represents L. cf. ubaghsi). Names of localities are for reference only.

FIGURE 11. Habitat of L. ubaghsi. Photo: Alejandra Alzamora.

FIGURE 3. L. leopardinus from El Colorado, Metropolitan Region. Photo: Damien Esquerré.

FIGURE 4. L. ramonensis from San Ramon Mountain, Metropolitan Region. Photo: Jaime Troncoso-Palacios.

FIGURE 5. L. valdesianus from El Morado National Monument, Cajón del Maipo, Metropolitan Region. Photo: Damien Esquerré.

FIGURE 6. L. frassinettii from Altos de Cantillana, between Metropolitan and Valparaiso Regions. Photo: Bernardo Segura.

Tables

TABLE 1. Morphological, meristic and coloration characters in the species of the leopardinus clade. SAMB: Scales around midbody; 4 TL: Fourth toe lamellae. SVL is measured in mm. Examined juveniles are excluded from SVL measurements. L. valdesianus from El Yeso Reservoir were not included in these measurements.

SAMB	77.7±4.6 (72–87)	76.6±4 (72–81)	87.7±4.1 (85–93)	81±3.7 (77–85)	82±3.9 (76–89)
Dorsal scales	70.9±6.3 (62–83)	74.2±4 (68–78)	81±4.4 (76–89)	79.7±4.2 (75–84)	76.2±5.2 (68–84)
Ventral scales	121.1±8.1 (111–137)	108.2±6.9 (100–115)	124.1±10.2 (111–140)	120.7±9.7 (109–136)	116.1±5.6 (108–129)
4TL	31.5±2.2 (28–35)	26.5±1.1 (25–28)	29±2.2 (26–32)	28.5±2.4 (25–31)	28.6±2.2 (25–31)
Precloacal pores in males	4	3–4	3	5	0–4
Head color	Ochre, mottled	Light brown or olive brown to black, sometimes mottled	Brown to olive brown, mottled	Light brown to ochre, mottled	Beige, with big black spots
Dorsal color	Ochre	Grey to olive brown	Olive brown	Light brown to ochre	Beige
Dorsal pattern	Black and white one scale sized dots, which may form transversal lines along the dorsum. Sometimes dark occipital stripe or a discontinuous vertebral line	Light transversal stripes with black on anterior edge	Light transversal stripes with black on anterior edge	Black dots uniformly distributed on the dorsum that may form transversal stripes. Sometimes there is a discontinuous vertebral line	Thick, 2–3 scales wide black continuous or discontinuous vertebral line. At the sides of the dorsum, big black hollow leopardlike spot arranged in transversal lines
Flanks	Dark band	Dark band	Posthumeral dark spot or band	Slightly dark band	Slightly dark band

TABLE 3. Estimates of evolutionary divergence for the mtDNA loci cytb between species of the leopardinus clade. Number of base differences per site are shown.

Species	L. frassinettii	L. ubaghsi	L. valdesianus	L. ramonensis
L. ubaghsi	0.045	-
L. valdesianus	0.042	0.011	-
L. ramonensis	0.04	0.08	0.003	-
L. leopardinus	0.04	0.012	0.007	0.004
Discussion