Liolaemus anqapuka Huamaní-Valderrama, Quiroz, Gutiérrez, Aguilar-Kirigin, Chaparro, Abdala, 2020

Huamaní-Valderrama, Ling, Quiroz, Aarón J., Gutiérrez, Roberto C., Aguilar-Kirigin, Álvaro, Huanca-Mamani, Wilson, Valladares-Faúndez, Pablo, Cerdeña, José, Chaparro, Juan C., Cruz, Roy Santa & Abdala, Cristian S., 2020, Some color in the desert: description of a new species of Liolaemus (Iguania: Liolaemidae) from southern Peru, and its conservation status, Amphibian & Reptile Conservation (e 250) 14 (3), pp. 1-30 : 8-22

Liolaemus anqapuka Huamaní-Valderrama, Quiroz, Gutiérrez, Aguilar-Kirigin, Chaparro, Abdala sp. nov.

( Figs. 2–5 ).

urn:lsid:zoobank.org:act:EF6ABFF4-97BC-4C8F-83E7-79D2B3FE7171

1885 Ctenoblepharis adspersus —Boulenger, Catalogue of the Lizards in the British Museum (Natural History). Second Edition 2: 136–137.

1978b “ Ctenoblepharus sp.” Péfaur et al. Bulletin de l'Institut Français d'Études Andines VII (1–2): 129–139. 1982 Liolaemus insolitus Cei and Péfaur, In Actas 8vo Congreso Latinoamericano de Zoología. Pp. 573–686.

1995 Ctenoblepharys adspersa —Etheridge, American Museum Novitates 3142: 1–34.

2004 Phrynosaura [sp.] Nuñez, Noticiario Mensual Museo de Historia Natural 353: 28–34.

2010 Liolaemus cf. insolitus, Gutierrez and Quiroz, Herpetofauna del Sur del Perú, Available: http:// herpetofaunadelsurdelperu.blogspot.com [Accessed: 13 June 2020].

2011 Liolaemus species 2 , Langstroth, Zootaxa 2809: 32. 2020 Liolaemus aff. insolitus 7, Abdala et al., Zoological Journal of the Linnean Society 189: 1–29.

Holotype. MUSA 5573 , an adult male ( Figs. 2–3 ), from between Quebrada San Jose and Quebrada Tinajones, District of Uchumayo, Province of Arequipa, Department of Arequipa, Peru (16°31’47”S, 71°39’04”W) at 2,460 m asl, collected on 10 November 2013, by C.S. Abdala, R. Gutiérrez, A. Quiroz, L. Huamani, and J. Cerdeña.

Paratypes. Six adult females: MUSA 5574–75 , same data as holotype. MUSA 1766 , from Quebrada Tinajones, 300 m southeast of holotype (16°31’54.29”S, 71°38’57.547”W) at 2,492 m asl, collected on 9 October 2010, by A. Quiroz and J. Cerdeña. MUBI 13522 , MUSA 1767 , from Quebrada Tinajones, 600 m southeast of holotype (16°31’54.207”S, 71°38’46.187”W) at 2,528 m asl, collected on 9 October 2010, by A. Quiroz and J. Cerdeña. MUBI 14680 , from Quebrada Tinajones (16°31’22.705”S, 71°37’35.666”W) at 2,561 m asl, collected on 27 July 2007, by R. Gutiérrez and A. Quiroz . Two adult males: MUBI 13521 , from Quebrada Tinajones, 300 m southeast of holotype (16°31’54.29”S, 71°38’57.547”W) at 2,492 m asl, collected on 9 October 2010, by A. Quiroz and J. Cerdeña. MUBI 14417 , from Quebrada Tinajones (16°31’22.705”S, 71°37’35.666”W) at 2,561 m asl, collected on 27 July 2007, by R. Gutiérrez and A. Quiroz .

Diagnosis. We assign Liolaemus anqapuka sp. nov. to the L. montanus group because it presents a blade-like process on the tibia, associated with the hypertrophy of the tibial muscle tibialis anterior ( Abdala et al. 2020; Etheridge 1995) and its placement in the morphological and molecular phylogenies ( Fig. 11 ). Within the L. montanus group, Liolaemus anqapuka sp. nov. differs from L. andinus , L. annectens , L. aymararum , L. cazianiae , L. chlorostictus , L. dorbignyi , L. fabiani , L, forsteri , L. foxi , L. gracielae , L. huayra , L. inti , L. jamesi , L. melanogaster , L. multicolor , L. nigriceps , L. orientalis , L. pachecoi , L. pantherinus , L. patriciaiturrae , L. pleopholis , L. polystictus , L. puritamensis , L. qalaywa , L. robustus , L. scrocchii , L. signifer , L. vallecurensis , L. victormoralesii , L. vulcanus , and L. williamsi , for being species of larger size (SVL greater than 75 mm) unlike L. anqapuka sp. nov., which has a maximum SVL of 73.5 mm. Liolaemus anqapuka sp. nov., has between 58 and 72 (mean = 64.8) scales around the body, which differentiates it from species of the group with more than 80 scales, such as L. cazianiae , L. duellmani , L. eleodori , L. erguetae , L. forsteri , L. gracielae , L. molinai , L. multicolor , L. nigriceps , L. patriciaiturrae , L. pleopholis , L. poecilochromus , L. porosus , L. pulcherrimus , L. robertoi , L. rosenmanni , L. ruibali , and L. vallecurensis ; and also from species with less than 55 scales, like L. aymararum , L. jamesi , L. pachecoi , and L. thomasi . Liolaemus anqapuka sp. nov. have 60–72 dorsal scales (mean = 65.5), and differs from L. andinus , L. cazianiae , L. eleodori , L. erguetae , L. forsteri , L. foxi , L. gracielae , L. halonastes , L. molinai , L. multicolor , L. nigriceps , L. patriciaiturrae , L. pleophlolis , L. poecilochromus , L. porosus , L. pulcherrimus , L. robertoi , L. rosenmanni , L. ruibali , L. schmidti , and L. vallecurensis , which have between 75–102 dorsal scales. The number of ventral scales between 73–87 (mean = 81.3) differentiates it from species with more than 90 ventral scales, such as L. andinus , L. cazianiae , L. erguetae , L. eleodori , L. foxi , L. gracielae , L. halonastes , L. hajeki , L. molinai , L. nigriceps , L. patriciaiturrae , L. pleopholis , L. poecilochromus , L. porosus , L. robertoi , L. rosenmanni , and L. vallecurensis . Liolaemus anqapuka sp. nov. has juxtaposed or subimbricate dorsal scales, without keel or mucron, this differentiates it from species with conspicuous keel and mucron, as L. aymararum , L. etheridgei , L. famatinae , L. fittkaui , L. griseus , L. huacahuasicus , L. montanus , L. orko , L. ortizi , L. polystictus , L. pulcherrimus , L. qalaywa , L. signifer , L. tajzara , L. thomasi , L. victormoralesii , and L. williamsi . Females of L. anqapuka sp. nov. present 1–4 (mean = 2.6) precloacal pores, this character differentiates it from species like L. andinus , L. balagueri , L. fittkaui , L. multicolor , L. ortizi , L. polystictus , L. puritamensis , L. robertoi , L. robustus , L. rosenmanni , L. ruibali , L. thomasi , and L. vallecurensis , because they do not present precloacal pores in females.

Liolaemus anqapuka sp. nov. belongs to the clade of Liolaemus reichei sensu Abdala et al. (2020) . The color pattern of Liolaemus anqapuka sp. nov. has a combination of characteristics in males and females that distinguish it from the rest of the Liolaemus of the group. The number of scales around the body is between 58–72 (mean = 64.8), which differentiates it from L. audituvelatus , L. balagueri , L. insolitus , and L. reichei ( Table 3). The number of dorsal scales varies between 60–72 (mean = 65.5), which is lower than the number in L. audituvelatus , higher than in L. nazca , and has a variation in range of scales different than L. chiribaya , L. reichei , and L. torresi ( Table 3). The numbers of ventral scales of Liolaemus anqapuka sp. nov. vary between 73–87 (mean = 81) which are different from L. audituvelatus , L. nazca , and L. torresi ( Table 3). The presence of precloacal pores in females 1–4 (mean = 2.6), is different from L. audituvelatus , L. balagueri , and L. reichei , whose females do not have precloacal pores ( Table 3). Coloration patterns on lateral sides have light blue scales, which are different from L. audituvelatus , L. balagueri , L. nazca , L. torresi , and L. reichei ( Table 3). The existence of dorsal body scales with a keel differentiate it from L. nazca which have dorsal body scales without keel. Ventral thigh scales with keel are present in 100% of individuals of L. anqapuka sp. nov. but they are less evident than those present in L. chiribaya , where only 35% of individuals present this character ( Table 3). The maximum SVL is greater than in L. audituvelatus , L. poconchilensis , L. reichei , L. stolzmanni , and L. torresi ( Table 3).

Description of the holotype ( Figs. 2–3 ). Adult male (MUSA 5573), SVL 73.53 mm. Head 1.20 times greater in length (16.47 mm) than width (13.74 mm). Head height 10.48 mm. Neck width 14.37 mm. Eye diameter 3.67 mm. Interorbital distance 10.96 mm. Orbit-auditory meatus distance 6.55 mm. Auditory meatus 2.0 mm high, 0.97 mm wide. Orbit-commissure of mouth distance 5.77 mm. Internasal width 1.58 mm. Subocular scale length 4.09 mm. Trunk length 31.81 mm, width 24.37 mm. Tail length 63.91 mm. Femur length 14.65 mm, tibia 14.47 mm, and foot 18.01 mm. Humerus length 11.01 mm. Forearm length 9.31 mm. Hand length 10.82 mm. Pygal region length 5.95 mm, and cloacal region width 7.97 mm. Dorsal surface of head rough, with 17 scales, rostral 3.09 times longer (2.78 mm) than wide (0.9 mm). Mental as long (2.78 mm) as rostral, trapezoidal, surrounded by four scales. Nasal separated from rostral by one scale. Two internasals slightly longer than wide. Nasal surrounded by eight scales, separated from canthal by two scales. Nine scales between frontal and rostral. Frontals divided into three scales. Interparietal smaller than parietal, in contact with six scales. Preocular separated from lorilabials by one scale. Five superciliaries and 15 upper ciliaries scales. Three differential scales at anterior margin of auditory meatus. Ten temporary scales. Four lorilabials scales, in contact with subocular. Seven supralabials, which are not in contact with subocular. Five supraocular. Eight lorilabials. Six infralabials. Five chin shields, 4 th pair separated by five scales. Seventy scales around half a body.

Sixty-two rounded dorsal body scales, juxtaposed, and without a keel or mucron; laminar anterior on members, imbricate and slightly keeled; laminar on hind limbs, imbricate and slightly keeled; tail with dorsal scales in the first third juxtaposed, and the remaining two-thirds imbricate, presence of some scales keeled. Eighty-six ventral scales, from the mental to the cloacal region, following the ventral midline of the body, laminar, imbricated. Thirty-two imbricate gulars, smooth. Neck with longitudinal fold with 36 granular, not keeled scales, ear fold and antehumeral fold present. Gular fold incomplete. Forelimbs ventrally laminar, subimbricate to imbricate, not keeled; hind legs laminar, imbricate, with some keeled scales ( Figs. 2–3 ). Seventeen subdigital lamellae on the 4 th finger of the hand. Twenty-one subdigital lamellae of the 4 th toe, with four keels, plantar scales with keels and mucrons. Lamellar ventral scales on tail, imbricate, not keeled. Five precloacal pores. Supernumerary pores absent.

Color of holotype in life ( Fig. 3 ). Dorsal and lateral color of the neck is light gray with few light blue scales, with dull orange scales, and spots on side. Dorsum, limbs, and tail light gray. Vertebral region delimited, vertebral line and spots absent, but dotted with sky blue scales. Paravertebral and dorsolateral region of the body, large orange spots of irregular shape and size stand out. These orange spots are surrounded and dotted with numerous sky-blue scales, with thin design or undulating edges. The orange spots with light white irregular spots. There are no dorsolateral bands, antehumeral arch, or scapular spots. On the sides of the body the pattern of orange spots and light blue scales is repeated, but the gray color of the body is darker. This design extends to the first third of the tail. Tail with dark semi-complete rings with white back spots. Midline of the body with orange scales and spots. Back of the limbs with numerous light white spots unevenly distributed. Hands and feet dorsally white. Ventrally white from mental region to the tail. Gular and femoral regions light yellow. Flanks of the body with a thin orange border from the armpits to the groin.

Morphological variation. Twenty-two specimens (six males and 16 females). Dorsal surface of head rough with 14–21 scales (mean = 16.82; STD = 1.71). Nasal surrounded by 6–9 scales (mean = 7.41; STD = 0.73). Supralabials 7–10 scales (mean = 8.18; STD = 0.8), lorilabials 8–11 scales (mean = 9.32; STD = 0.89). A line of lorilabial scales. Supraoculars 4–6 (mean = 5.45; STD = 0.6). Interparietals smaller than parietals, surrounded by 4–8 scales (mean = 6.32; STD = 1.09). Infralabials 6–9 (mean = 7.14; STD = 0.77). Gulars 28–39 (mean = 33.41; STD = 2.99). Temporals smooth, 7–10 scales (mean = 9.09; STD = 0.97). Meatus auditory higher 1.37–2.47 mm (mean = 2.05; STD = 0.26), than wide 0.20–1.20 (mean = 0.81; STD = 0.25). Head longer 12.32–17.20 (mean = 14.91; STD = 1.31) than wide 9.15–15.92 (mean = 12.77; STD = 2.03). Head height 6.84–10.48 (mean = 8.38; STD

. anqapuka

.

sp

L

nov. scales

Light blue can that form irregular thin lines or clump forming together conspicuous spots light Sometimes lateral scales blue paravertebral on spots absent 4.22 (3 5)) (4 6 5

nazca

. L absent absent present 1 2 3) (3 () 6 4

chiribaya

, Regular associated with and paravertebral

. L lateral absent absent 3 2 () 4) 6 4.14 (3

torresi absent absent absent) 1.5

2

. (3 L 0

poconchilensis sides of Both the body absent absent 0 1.3 2 () 4 5) (4.6. L

clade

.

reichei L.

insolitus, from Regular the to occiput approximately two tail the thirds of next, to Scales paravertebral spots absent (1.3 3) 0 6.2 5) (7

Liolaemus the

.

reichei absent absent absent 0 4

of L

characters morphological L. balagueri absent absent present 0 5.08 ()

7 3

in

Differences auditvetulaus

absent absent absent 0 5

L.

.

)

continued

3

Table

(

Morphological character of Arrangement in celestial scales males of Arrangement scales in celestial females side spots Green in pores Precloacal females in Precloacal pores males

= 0.87). Underarm to groin length 21.61–32.8 (mean = 28.58; STD = 2.76). SVL males 56.23–73.53 mm (mean = 65.05 mm; STD = 7.08) and females 52.15–71.10 mm (mean = 62.9 mm; STD = 4.61). Femur length 10.11–14.65 mm (mean = 12.31 mm; STD = 1.06). Humerus length 7.56–11.01 mm (mean = 8.86 mm; STD = 0.99). Forearm length 7.65–11.56 mm (mean = 9.59 mm; STD = 1.06). Hand length 8.03–11.25 (mean = 10.25; STD = 0.86). Scales around midbody 58–72 (mean = 65.09; STD = 3.7). Dorsal 60–72 (mean = 65.59; STD = 3.5), juxtaposed to sub-juxtaposed, and smooth scales. Infradigital lamellae of the 4 th finger of the hand 15–21 (mean = 17.73; STD = 1.45) and of the 4 th toe 20–26 (mean = 21.67; STD = 1.5). Ventral 73–87 (mean = 81.32; STD = 3.37) larger than dorsal scales. Tail length 46.77–67.16 mm (n = 17, mean = 56.83 mm; STD = 5.91). Males with 4–6 (mean = 4.67; STD = 0.82) precloacal pores, and females with 3–5 (mean = 4.22; STD = 0.83) precloacal pores. Body measurements, males (mean = 66.62 mm) slightly larger than females (mean = 62.90 mm), tail length in males slightly larger (mean = 61.74 mm) than females (mean = 54.80 mm) [ Table 4].

Color variation in life ( Figs. 4–5 ). Liolaemus anqapuka sp. nov. shows evident sexual dichromatism. In males, head is darker than the gray body. In some specimens, supralabial and infralabial scales are generally lighter gray than the rest of the head. The subocular is generally white with irregular dark spots. The dorsal color of the neck is gray, varying in its hue, and may be dotted with some light blue scales and orange spots. The body color is always gray. The vertebral region in most males is well delimited with some light blue scales. No vertebral line, dorsolateral bands, antehumeral arch, or scapular spots. Few specimens have diffuse gray paravertebral spots, and rounded shape. As in the holotype, in the paravertebral, dorsolateral, and lateral regions of the body, irregular orange spots stand out, surrounded and dotted with celestial scales. Orange spots can vary in intensity and size, as light blue scales that can form thin irregular lines or clump together to form more conspicuous spots. In some specimens the amount of light blue scales is so remarkable that they cover the orange spots. Orange spots and light blue scales are distributed on the sides of the tail. In some individuals, the celestial scales reach the distal end of the tail. In some specimens, light blue scales are replaced by dark, bluish-green scales. In some, irregularly shaped white spots are distributed among the orange spots. The fore and hind limbs, as well as the tail, have the same design as the body. In the tail, incomplete rings of dark spots with light edges are formed. Ventrally, the majority of males are similar. The predominant color is white, some have faint yellow and a yellow hue that can vary in intensity, highlighted in the gular region and the hind limbs. On the sides of the belly, a thin orange longitudinal line protrudes from the armpit to the groin ( Fig. 4 ).

Females have a totally different coloring pattern than males ( Fig. 5 ). The color of the head varies from brown to gray, with some dark red spots and scales. The supralabial, infralabial, and lorilabial scales are lighter in color than the dorsal surface of the head. The back of the body can be light gray or brown; with small paravertebral spots, gray or dark brown, and circular or sub-quadrangular; with a small white spot on the back which can be the same size as the paravertebral; and with meager orange spots between the paravertebrals. A few females have light blue scales on paravertebral spots. On the sides of the body, there may be lateral spots of the same design as the paravertebral ones. The tail and hind limbs have the same design and color as the body, without dorsolateral bands. Ventrally they are white or faint yellow immaculate throughout the body. In some females, the tail has more intense yellow throughout its extension ( Fig. 5 ).

Etymology. The specific name refers to the coloration patterns of males. The word “anqapuka ” is an original word in the Quechua language (spoken currently in the Peruvian Andes), corresponding to a complex word between “anqa” assigned to the blue color, and “puka” which means orange or red color.

Distribution and natural history. Liolaemus anqapuka sp nov. is restricted to the western slopes of the La Caldera batholith, Arequipa, Peru, between 1,800 and 2,756 m asl, which includes the upper altitude limit of the La Joya desert ( Fig. 6 ). The distribution is within the Desert biogeographic province (sensu Morrone 2014). Liolaemus anqapuka sp. nov. inhabits arid environments, characteristic of the desert of southern Peru, with sandy-stony substrates and little slope, seasonal herbaceous vegetation, and columnar and prostrate cacti. This species also inhabits sectors without vegetation ( Fig. 7 ). It takes refuge mainly under stones, and in burrows that surround the roots of small bushes, prostrate cacti, and in cavities underground or in hardened sand. Some specimens of Liolaemus anqapuka sp. nov. were observed feeding on coleopteran larvae, as well as larvae and notably adults of Lepidoptera belonging to the Sphingidae family ( Fig. 8 ). Feeding on beetles is very similar to that reported for the closelyrelated species Liolaemus insolitus , which is specialized in feeding on so-called “flea beetles” of the subfamily Halticinae ( Coleoptera : Chrysomelidae ) [ Cei and Péfaur 1982]. The adults and larvae of the family Sphingidae are most abundant in the summer months, when the local rainfall is complemented by abundant ephemeral surface watercourses whose flow is derived from rainfall on the western slopes of the Andes, and these insects can display unusual and explosive development. During years when there is exceptionally high accumulated rainfall, a biological phenomenon known as a “blooming desert” can occur ( Chavez et al. 2019), and some phytophagous insects would be expected to be able to use the abundant plant resources that suddenly become available in these events, as reported for Sphingidae in northern Chile ( Vargas and Hundsdoerfer 2019). Liolaemus anqapuka sp. nov. was found in syntopy with other reptile species, such as Microlophus sp. and Phyllodactylus gerrhopygus .

Endemism, threats, and conservation status. Liolaemus anqapuka sp. nov. is considered as an endemic species with a restricted-range of geographical distribution, because the species occupancy is less than 10,000 km 2 ( Bruchmann and Hobohm 2014; IUCN 2016; Kier and Barthlott 2001; Noguera-Urbano 2017). Using the Geocat tool, and based on records of the species, we estimate the extent of occurrence (EOO) at 147.2 km 2 and the area of occupancy (AOO) at 80.0 km 2. The restricted range might be caused by their climatic tolerance, and the ecological adaptation to extreme environmental conditions found on the Desert biogeographic province. The main threats are the loss of habitat, because of the large-scale mining activities, urban expansion, and contamination by chemicals and metals; and also because of the presence of highways that cut through their natural habitat, and the opening of new secondary roads. Following the IUCN (2020) criteria, and using the actual knowledge of the new species, we evaluated the conservation status of L. anqapuka sp. nov. to be in the category of endangered EN [A2cde; A3cde; A4cde] [B1ab (i, iii) + 2abc (ii, iii, iv)], based on the area of occupancy (AOO) <500 km 2, the extent of occurrence (EOO) <5,000 km 2, the number of localities are ≤ 5; and we consider it as a species with restricted range because L. anqapuka sp. nov. has a global range size less than or equal to 10,000 km 2 ( IUCN 2016). Statistical analysis ( Figs. 9–10 ). The summary statistics for all the non-transformed, continuous, and meristic characters taken from five species of Liolaemus are shown in Appendix II. The homogeneity of variance was not supported for either continuous or meristic characters by the Levene’s test in some groups. Therefore, the results of the Principal Component Analyses (PCA) should be preferred for deriving linear combinations of the variables that summarize the variation in the data set. The results of the PCA for continuous and meristic characters are presented separately ( Tables 5–6).

The first four components of continuous characters explained 55.51% of the variation, and a screen plot test of the PCs indicated that only the first three components contained nontrivial information. The first axis represents body size, loading negatively for most variables, and accounts for 23.46% of the variation, with strong loading for width of the base of the tail. The second axis represents morphological variation and accounts for most of the remaining variation, with strong loadings for mental scale width, length of the 4 th supralabial scale, and upper width of the pygal area. The next axes account for the remaining variation.

The first four components of meristic characters explained 54.59% of the variation, and a screen plot test of the PCs indicated that only those components contain relevant information. The four axes represent morphological variation, loading strongly for number of paravertebral spots in the right side, number of scales around midbody, number of ventral scales, and number of gular scales. The four axes account for the remaining variation, albeit with values below 0.70 for subdigital lamellae of the 4 th finger of the forelimb, number of auricular scales, projecting scales on anterior edge of auditory meatus, and number of organs in the postrostral scales.

The positions of species based on their scores for the two morphological principal components axes are illustrated in Figs. 9–10 . The spatial distribution of the continuous characters indicates that they are sufficient to virtually separate the five Peruvian Liolaemus species of the L. reichei group. These species can also be distinguished by their position in the analysis of meristic characters only. In both analyses, Liolaemus anqapuka sp. nov. can be differentiated from other phylogenetically related species by its body size and morphological variation.

To further clarify the position of the Liolaemus species in the morphospace of both continuous and meristic characters, a DFA was carried out, where the group membership was determined a priori. The result obtained through the DFA for the five species of Liolaemus was not significant for continuous morphological characters (Wilk’s Lambda = 0.85, F = 0.71, P = 0.60), and the jackknife classification was 100% satisfactory. The DFA of operational taxonomic units for meristic characters was not significant either (Wilk’s Lambda = 0.69, F = 1.58, P = 0.23); however, the jackknife satisfactory classification was developed at a 100% rate. These results show L. anqapuka sp. nov. can be reliably distinguished from the other species by a combination of morphological characters.

Phylogenetic analysis ( Fig. 11 ). The objective of the phylogenetic analyses carried out (morphological, molecular, and Total Evidence) is not to resolve the relationships of the L. montanus group, which is far beyond the scope of this study. The main objective of these analyses is to obtain some approximation of the phylogenetic relationships of L. anqapuka sp. nov. and the rest of the L. reichei group sensu Abdala et al. (2020). The new taxon was recovered in three analyses, within the L. montanus group. In the morphological and Total Evidence analyses, under parsimony methodology, the L. reichei group is monophyletic; within this, L. anqapuka sp. nov., through molecular analysis of ML, the L. reichei group is paraphyletic.

Molecular analysis. The three DNA (cyt- b) obtained for L. anqapuka sp. nov. fall within the same clade, supporting the identification of the new species. The nearest terminal is L. aff. insolitus 4, a population innominate from Department of Arequipa, and it is grouped in the same clade with L. chiribaya , a species from Department of Moquegua, with node support (BS = 99). The clade that contains these three species is deeply separated from its sister clade, ( L. poconchilensis + L. aff. insolitus 8). The analysis does not recover the clade of L. reichei group sensu Abdala et al. (2020) as monophyletic.

Morphological analysis. The result of the morphological phylogenetic hypothesis shows that Liolaemus anqapuka sp. nov. belongs to the group of L. montanus , within the clade of L. reichei sensu Abdala et al. (2020) , together with L. audituvelatus , L. balagueri , L. chiribaya , L. insolitus , L. nazca , L. poconchilensis , L. reichei , L. torresi , and eight unnamed populations so far. Liolaemus reichei sensu Abdala et al. (2020) , is supported by 13 synapomorphies, of which four are continuous characters (lower number of scales from rostral to occiput, lower number of scales around midbody and lower ratio of tail length/SVL) and eight are discrete (ventral scales of the body equal to, or slightly larger than the dorsal; sides of the body not conspicuously colored, with little or no ventral sexual dichromatism; absence of white line in the temporal region; diameter of the eye, larger than the distance between the anterior margin of the eye, and the rostral scale; isognathic profile, substrate where they occur predominantly sandy).

This clade is divided into two large subclades, one with unnamed species and populations from Chile ( L. audituvelatus , L. poconchilensis , L. reichei , and L. torresi ) and the other with species and populations from central and southern Peru ( L. balagueri , L. chiribaya , L. insolitus , and L. nazca ). This last subkey is where the new species is recovered, supported by 19 synapomorphies, several of which stand out: ratio of auditory meatus height/head height, number of pygals, number of lorilabials contacting the subocular, number of supraoculars, dorsal surface of head (rugouse), scales on external edge of forelimbs (subimbricate), scales of dorsal hind limbs (subimbricate), with notch in edge of scales of gular fold, scales of pygal region (subimbricate), with dark line through the eye; white posterior edge of paravertebral spots in both sex (present), black dots scattered on dorsal region of hind limbs in males (absent), and dark line through the eye in females (present). Liolaemus anqapuka sp. nov. have populations of close relatives which also occur in Department of Arequipa, Peru, with particular morphological characteristics, and these are currently under description. Liolaemus anqapuka sp. nov. is recovered as a sister species of L. aff. insolitus 4, a population related to L. insolitus near the coasts of the Department of Arequipa, which occupies elevations of 1,000 m asl. This relationship is supported by six synapomorphies. Liolaemus anqapuka sp. nov. is supported by seven autopomorphies in the phylogenetic tree ( Fig. 11 ).

Total Evidence analysis ( Fig. 11 ). The L. reichei clade is recovered as monophyletic, and L. anqapuka sp. nov. belongs to this clade, as do the sister species of L. aff. insolitus 4, as well as in the morphological and molecular phylogenetics analyses. This relationship is supported by 14 synapomorphies, six of which are continuous characters and the support of this relationship is high (89%). This relationship is recovered within the clade ( L. aff. insolitus 5 ( L. aff. insolitus 4 + L. anqapuka sp. nov.)), and is supported by three morphological and 11 molecular synapomorphies. Likewise, a total of seven autopomorphies support the new species of Liolaemus . In this hypothesis, as in the morphological one, two sub clades are recovered within the L. reichei clade—on the one hand are the species that are distributed in northern Chile, and on the other are those in southern Peru.

Taxonomic history. Boulenger (1885) identified a male specimen (BMNH 65–5–3–3) from “Arequiba, 7,500 ft ” as Ctenoblepharis adspersus (an unjustified emendation of Ctenoblepharys adspersa Tschudi 1845 ) in his catalogue of the lizards in the British museum. Péfaur et al. (1978b) mentioned the distribution and classification of the reptiles from Department of Arequipa, noting that the specimens collected by Duellman (1974) from the “La Caldera batholith” located approximately 10 km southwest of Uchumayo town would be “ Ctenoblepharus sp.” (= Ctenoblepharys ). But this was not the only mistake. Years later, Cei and Péfaur (1982) wrote the original description of Liolaemus insolitus , considered to be a widely distributed coastal species which reached altitudes above 2,000 m asl, including the populations of the “La Caldera batholith” from Department of Arequipa. Etheridge (1995), from the specimens considered by Boulenger (1885), identified the possible existence of a different species of Liolaemus from Department of Arequipa, which shows the characteristics of the specimens collected by Duellman (KU 163589, 3 km SW Uchumayo, at 2,150 m asl). During the following years, the regional museums of Peru considered the population from “La Caldera batholith” as an undescribed form associated with Liolaemus insolitus ( Zeballos et al. 2002) , which they called Liolaemus cf. insolitus . Nuñez (2004) identified the specimen considered by Boulenger (1885) as a new species of the genus Phrynosaura (synonym of Liolaemus ). Gutiérrez and Quiroz (2010), based on photographic material, presumed that the population belonged to L. cf. insolitus . Later, Langstroth (2011) reviewed the field notes written by Duellman, Simmons, and Pefaur (unpublished) and their specimens cataloged as Phrynosaura stolzmanni from the University of Kansas (KU 163589, KU 163592, and KU 163594; collected from “ 10 km SE of the town of Uchumayo, in the La Caldera batholith”), and indicated that these lizards are not Liolaemus stolzmanni . Based on fieldnotes, which indicate that these specimens are individuals found in habitats of gray sand with granitic rocks and the coloration is cryptic with the habitat, he also highlights the mottled black, orange, and metallic blue back, and the lateral sides of the belly are orange; and these characters are corroborated with the photography of the individual KU163589; citing this population in his work as Liolaemus species 2 (KU 163589, KU 163592, and KU 163594). Finally, Abdala et al. (2020) corroborate through analysis of Total Evidence of the L. montanus group that the population from “La Caldera batholith” ( L. aff. insolitus 7) is an independent terminal, because it presents morphological characteristics different from the rest of the known species of Liolaemus . Therefore, we corroborate the hypothesis presented by Abdala et al. (2020), based in morphological and molecular phylogenetic evidence, which they named as L. aff. insolitus 7.

Acknowledgments.— We are grateful to Evaristo López [Museo de Historia Natural, Universidad Nacional San Agustín, Arequipa, Perú ( MUSA)], the staff of the Museo de Biodiversidad del Perú ( MUBI, Cusco, Perú), Sonia KretzschmarandEstebanLavilla [FundaciónMiguelLillo, Tucumán, Argentina ( FML), César Aguilar-Puntriano y Alejandro Mendoza (Museo de Historia Natural, Universidad Nacional Mayor de San Marcos, Lima, Perú ( MUSM)], for allowing the review of specimens from their museum collections and for facilitating access to the collection under his care. Comments of Roberto Langstroth, Emma Steigerwald, Matt King, and one anonymous reviewer improved our manuscript considerably. Collection permits for specimens were issued by Ministerio de Agricultura, through Resolución Directoral N ° 0399–2013–MINAGRI-DGFFS/DGEFFS and Resolución Directoral N °0112-2012-AG-DGFFS- DGEFFS; and additionally, the Resolución de Dirección General N ° 509-2018-MINAGRI-SERFOR-DGGSPFFS. We are grateful to Yovana Mamani Ccasa, from Cusco, for the support of the epithet in the original language of the Incas. LHV thanks Luis Arapa and Jeitson Zegarra for their help, in part, in obtaining morphological data; Mg. Sandro Condori and Dr. Sebastían Quinteros for their comments on the sequence alignment process; Dra. María Valderrama for access to the environments of the Genetics Laboratory of the National University of San Agustín; and finally, he especially thanks the researchers and others who help him during his stay on his trip to Argentina: Romina Sehman, AnaLu Bulacios, Marco Paz, the Abdala family in Mendoza, Lisseth Montes and family in Tacna, Valladares family in Chile, Carlos Valderrama and family in Lima. CSA thanks the Cerdeña family and the Hotel Princess from Arequipa. Thanks to CONICET and the Agencia y Técnica ( PICT 2015- 1398 , Argentina) and “Convenio de Desempeño Regional UTA-1795.”

Figures

Fig. 2. Details of the holotype of Liolaemus anqapuka sp. nov. (MUSA 5573; SVL = 73.5 mm, Tail = 63.9 mm): (A) dorsal and (B) ventral views of body; (C) ventral, (D) dorsal, and (E) lateral views of head; (F) ventral view of precloacal pores. Scale = 10 mm.

Fig. 3. Adult male of the holotype, Liolaemus anqapuka sp. nov. (MUSA 5573; SVL = 73.5 mm, Tail = 63.9 mm), from the Depart- ment of Arequipa, 2,460 m asl. Photos by C.S. Abdala.

Fig. 4. Male specimens of the Liolaemus anqapuka sp. nov. Photos by A. Quiroz (A–D) and C.S. Abdala (E).

Fig. 5. Female specimens of the Liolaemus anqapuka sp. nov. Photos by A. Quiroz.

Fig. 11. Phylogenetic trees showing the relationships between Liolaemus anqapuka sp. nov. and species within the L. montanus group by (A) Total Evidence analysis, (B) molecular phylogenetic analysis, and (C) morphological phylogenetic analysis. The values correspond to the support measure with symmetric resampling.

Fig. 6. Geographic distribution of Liolaemus montanus group species from Peru. Symbols with a black dot in the middle represent the type locality of each species. Species with quotation marks in the names belong to the candidate species listed in Aguilar et al. (2016).

Fig. 7. Habitat of Liolaemus anqapuka sp. nov. in (A) dry season and (B) wet season. Photos by A. Quiroz (A), C.S. Abdala (B).

Fig. 8. Liolaemus anqapuka sp. nov. eating a moth of the Sphingidae family. Photo by A. Quiroz.

Fig. 9. Plot of principal component scores for continuous characters for L. balagueri (yellow stars, n = 12), L. chiribaya (purple circles, n = 10), L. insolitus (red triangles, n = 15), L. nazca (sky blue triangle, n = 7), and L. anqapuka sp. nov. (green squares, n = 7). Eigenvectors, eigenvalues, and percent of variation explained for the first two principal components are summarized in Table 5.