Darevskia tuniyevi, Arribas & Candan & Kurnaz & Kumlutaş & Caynak & Ilgaz, 2022

Arribas, Oscar, Candan, Kamil, Kurnaz, Muammer, Kumlutaş, Yusuf, Caynak, Elif Yıldırım & Ilgaz, Çetin, 2022, A new cryptic species of the Darevskia parvula group from NE Anatolia (Squamata, Lacertidae), Organisms Diversity & Evolution (New York, N. Y.) 22 (2), pp. 475-490 : 481-483

Description of Darevskia tuniyevi sp. nov.

ZooBank registration (http://zoobank.org): urn:lsid:zoobank.org:pub:DFA82B86-D97A-4A5D-B11D-CDF14CB8489C .

Holotype ZDEU3 /2020, ♂, Pırnallı, Artvin, Turkey. leg. Muammer Kurnaz. 11.09.2020.

Paratypes 1- 10 ♂♂, 6 ♀♀, same locality, date and collectors as holotype. 2- ZDEU149 /2001, 6 ♂♂, 4 ♀♀, 15 km W of Şavşat , Turkey. leg. Yusuf Kumlutaş, Kurtuluş Olgun, Çetin Ilgaz, Aziz Avcı, Fatma İret. 04.07.2001. 3- ZDEU152 /2001, 4 ♂♂, 7 ♀♀, 19 km W of Ardahan, Turkey. leg. Yusuf Kumlutaş, Kurtuluş Olgun, Çetin Ilgaz, Aziz Avcı, Fatma İret. 06.07.2001. 4- ZDEU218 /2014, 5 ♂♂, 3 ♀♀, Çağlıyan , Artvin, Turkey. leg. Yusuf Kumlutaş. 23.07.2014 .

Etymology The specific epithet refers to Dr. Boris S. Tuniyev, for his remarkable work about the knowledge of Caucasian herpetofauna and its remarkable diversity.

Diagnosis A medium-small sized Darevskia , morphologically very similar to D. parvula (cryptic species, are not sister taxa), characterised by greater values of submaxyllary type (SBMXTyp) (nearly half of the specimens attributed to D. tuniyevi sp. nov. have submaxillary scales of types characteristic of D. adjarica and the other half are similar to D. parvula ), greater ventralia (TVP), greater number of temporal scales among masseteric and tympanic (TS1), higher dorsalia (DS), greater number of 4 th digit subdigital lamellae (SDL), and smaller values for ventral side tight scales (LS). Also, slightly longer pilei (PLI). See the “Comparison among species” section below and Table S2 for detailed differences. Genetically it is reciprocally monophyletic in respect to the other two D. parvula group species ( D. parvula and D. adjarica ).

Description of the holotype Adult male with an intact tail, preserved in ethanol ( Fig. S2 ). Coloration and pattern (in alcohol): dorsum background brownish (Beige -254- in life). Pattern faintly reticulated, composed of white ocelli surrounded by black, that appears well developed in the sides (costal or temporal bands), a last faint row of clear ocelli just in the upper border (more or less scalloped) of the costal bands (especially in the forebody), leaving the dorsal tract mainly brownish colour (Beige -254- in life), almost without a trace (or barely visible) of these clear ocelli, and dark pattern (derived from the black reticulation) reduced to two paravertebral rows of irregular spots, well visible, that continue and coalesce in the tail base. Legs brownish (Beige -254- in life) tone with faint reticulate black and (less visible) clear spots. Tail with scale rings of alternating width, usually the distal part of the scales are darker than the medium and proximal parts. Pileus with tiny well-defined spots above, the temporal bands faintly extending into the temporal area of the head. No trace of blue axillar ocelli. Supra, infralabials and subocular with more or less well-defined rounded spots. Gular area, throat, belly and underside tail without a dark pattern (white in alcohol, Salmon -83- in life). Blue spots (Light Caribbean Blue -163-) on the outer ventral scales.

Scalation Contact between rostral and internasal plates absent; Supraciliary granules in the right side, 12; Supraciliary granules in the left side, 12; Supraciliary plates_right, 6; Supraciliary plates_left, 6; Supralabial plates_right, 4; Supralabial plates_left, 4; Sublabial plates_right, 6; Sublabial plates_left, 6; Type of submaxillars, F (from Darevsky & Eiselt, 1980, Fig. 5); Gulars, 27; Collaria, 9; Supratemporals_right, 1; Supratemporals_left, 1; Scales between tympanic and parietal_right, 2; Scales between tympanic and parietal_left, 1; Scales between masseteric and tympanic_right, 3;Scales between masseteric and tympanic_left, 3; scales between masseteric and supratemporal_right, 2; Scales between masseteric and supratemporal_left, 2; Posttemporals_right, 6; Posttemporals_left, 6; Ventral plates (transversal rows), 25; Ventral plates (longitudinal rows), 6; Contact between Postorbital and Parietal, absent; Preanal 1, one big and enlarged plate; Preanal 2, six scales; Femoral pores_right, 23; Femoral pores_left, 24; Scales between femoral pores and outer plates_right, 3; Scales between femoral pores and outer plates_left, 3; Subdigital lamellae_right, 28; Subdigital lamellae_left, 28; Tibials, 16; Dorsalia, 56. Tail proximal rows of scales fairly keeled and outwards rose. Biometry Snout-vent length, 52.44 mm; Tail length, 122.82 mm; Total length, 175.26 mm; Pileus length, 12.85 mm; Pileus width: 6.06 mm; Head width, 6.06 mm; Head length: 6.85; Pileus Index, 2.12; Head Index, 1.94.

Intraspecific variation Males can have background tone more greyish than the holotype (from Beige -254 -, Drab Gray -256- to Pale Neutral Grey -296-) ( Fig. S3 ). Paravertebral rows of dark spots can be wider, occupying a good part of the dorsal tract, or less and more centred (as in the holotype), or even substituted by a fine stipple widespread across the entire dorsal tract, without defined paravertebral rows. The white ocelli in the dorsolateral area (upper limit of the costal (temporal) bands can be more marked or almost faint. In some rare specimens the paravertebral and costal bands connect and form a near-reticulated pattern. Underside, out of the reproductive period, the salmon or orange colour is conserved especially in the aged males (bigger ones). Throat, chest, and bellies are completely immaculate in all the specimens. The bigger ones have blue spots in the outer ventral scales.

Females less variable than males, with more or less developed spots in the paravertebral rows. In general, both these paravertebral dots, as with the costal bands, are clearly less developed than in males. Old females conserve the ventral salmon coloration outside the breeding period. No blue spots in outer ventral scales. In all the other aspects, similar to males.

Juveniles have the pattern and coloration of adults, are white bellied, and with big scales of the underside of the hind legs completely pigmented except a clear flange. Tail not brightly coloured, with the same colour and tone as the dorsum.

Osteology Darevskia tuniyevi sp. nov. (from Ardahan, Turkey) has 7 premaxillary teeth, from 14 to 18 maxillary teeth, from 18 to 23 dentary teeth, 28 presacral vertebrae in males and 29 in females (with 6 and 7 short ones in the lumbar area, respectively). There is no trace of short vertebrae associated with the third presacral vertebrae. Tail autotomic vertebrae are type A ( Arnold, 1973; Arnold et al., 2007). All the clavicles studied are open (marginated). The interclavicle is typically cruciform and oval sternal fontanelle. Costal formula is (3 + 2) with one or no inscriptional rib with a similar proportion of occurrence. Postorbital bone is shorter than the postfrontal. Anteromedial process of postorbital is present but small and frequently much reduced and barely discernible. Anterodistal process of the postfrontal is present. Squamosal and postorbital overlap approximately along one third of the length of the latter.

Habitat and ecology Artvin province receives average annual precipitation of approximately 700 mm, the air temperature ranges from -0.2 °C in winter to 26.3 °C in summer (see Fig. S4 ). The type locality (Pırnallı) is an area with dense forest at an altitude of 1258 m above sea level (coordinates: 41°15′13.66"N; 42° 3′57.16"E; 1258 m) ( Fig. S4-a ). Other localities like Çağlıyan (Artvin, Turkey) have similar traits for habitat and vegetation ( Fig. S4-b ).

Lizards were found to be active all day, with air temperatures between 25 and 33 °C. In general, specimens were caught in rocky, stony, or bare soil areas along the edge of the forest road in a mixed forest with Fagus orientalis (Oriental Beech), Picea orientalis (Caucasian Spruce) and Pinus sylvestris (Scots Pine) . The range is composed of volcanogenic flysch and sandstones ( Fig. S4 ). Other reptile species living in sympatry in the area are Mediodactylus sp. , Darevskia derjugini (Nikolsky, 1898) , Lacerta media Lantz and Cyren,1920 , Hemorrhois ravergieri (Menetries, 1832) and Zamenis hohenackeri (Strauch, 1873) .

Distribution In the Meskheti and Şavşat Ranges of the Lesser Caucasus (Şavşat, Kedi, Karçal Dağı, Yalnızçam Dağı) [centre of Artvin, northeast of Şavşat, Ardanuç and Ardahan] (see Fig. 1 ). The area seems to be encompassed in the interfluvium at the right side of the Çoruh (Coroch) river (that separates it from D. parvula and the Turkish area of D. adjarica ), and perhaps the Ajarisckali river at the north (separating from the Georgian area of D. adjarica ). D. tuniyevi sp. nov. is only known from Turkey but is very near the Georgian frontier (i.e. Camili, Artvin) and it is possible that enters Georgia in a narrow strip (perhaps up to the Ajarisckali river, as said above).

Comparisons among species Combining CDA and ANOVA results (in bold CDA characters also p <0.01 or 0.001 in ANOVA; see Table S2 for specific values of scalation and indexes of each species):

Darevskia tuniyevi sp. nov. males differ from D. parvula (cryptic species) with greater values for SBMXTyp (near half of the specimens attributed to D. tuniyevi sp. nov. have submaxillary scale similar to the types typical also in D. adjarica ), TVP (ventralia), TS1 (temporal scales among masseteric and tympanic), DS (dorsalia), and SDL (4 th digit of subdigital lamellae), and smaller LS (ventral side tight scales) [also in PLI (longer pilei) at p <0.01 in ANOVA, not included in CDA]. In females D. tuniyevi sp. nov. have greater values for TVP, SBMXTyp, SL, DS, TS1 and SDL, and smaller values for TS2 (longitudinal rows of temporal scales between tympanic and parietal), and LS (-0.34) [also POT (postemporal plates) in ANOVA at p <0.01)]. The sixth submaxillary morphology is very similar to the D. parvula one, but the smaller scale is comparatively greater than in this later ( Fig. S2-c ). Differences between D. tuniyevi sp. nov. and D. parvula are subtle and it can be considered cryptic species, despite not being a sister taxa.

Darevskia adjarica , the most different taxa in the group, is differentiated from D. parvula and D. tuniyevi sp. nov. by greater values of SbmxTyp and Sptmp, and smaller scores for POT, LS and TVP in males [also in ANOVA: SCG, SCP, SRL, MG, TS2, TS1, POT, FP, SDL, TS, DS, R-I, PLI and HWI]. In females, there are higher values of Sbmxtyp and smaller scores for POT, DS and TVP [also in ANOVA: SRL, FP, SDL, PLI and HWI]. + D. parvula (and D. tuniyevi sp. nov.) have contrary values for these characteristics. The sixth submaxillary has the higher scale subtriangular, more than twice the smaller one ( Fig. S2-c ). See also Arribas et al. (2018) for a description of D. adjarica .

Concerning pattern and pigmentation, D. adjarica is also the most different, with a generally more darker and contrasted pattern, a common greenish sheen (physical colour, not pigment) and sometimes axillary blue ocelli in males. Also, the belly is more frequently and more intensely pigmented with reddish tones than in the other two species ( D. parvula and D. tuniyevi sp. nov.). Variability in some of these characteristics (extension of black pattern and intensity of pigmentation) is parallel to the lizard’s bioclimatic characteristics and are more typical of moist places (own data).

Figures

Fig. 2 Canonical Discriminant Analysis (CDA) plot for MALES a and FEMALES b. Specimens, sample centroids and group perimeters are represented. For details, percentage of variance explained by axes and interpretation, see text

Fig. 3 BI tree inferred from the Cyt-b dataset. Numbers on branches indicate the bootstrap and posterior probability (pp) values (ML/BI, values lower than 50/0.70 are not shown). Orange stars represent species boundaries inferred by mPTP, while purple ones the results of bPTP. Divergence times between main clades are shown on the tree

Fig. 4 Haplotype networks corresponding to 16S rRNA a, Cyt-b b and Rag-1 c. Lines rep- resent a mutational step, black circles missing haplotypes, and coloured circles haplotypes. The circle area is proportional to the number of individuals

Fig. 1 Map showing the localities of the samples included in the morphological and molecular analyses. Blue: Darevskia parvula (Lantz & Cyren, 1913). Green: Darevskia adjarica (Darevsky & Eiselt, 1980). Red: Darevskia tuniyevi sp. nov