Melitaea acentria Lukhtanov

Lukhtanov, Vladimir A., 2017, A new species of Melitaea from Israel, with notes on taxonomy, cytogenetics, phylogeography and interspecific hybridization in the Melitaeapersea complex (Lepidoptera, Nymphalidae), Comparative Cytogenetics 2, pp. 325-357 : 327-336

Melitaea acentria Lukhtanov sp. n.

Holotype

(Fig. 2a, b), male, BOLD process ID BPAL2191-13, field # CCDB-17949_A06, GenBank accession number # KY777529; Israel, Mt. Hermon, 33°18'45.6"N; 35°47'11.9"E, 2050 m, 01 June 2013, A. Novikova leg., deposited in the Zoological Institute of the Russian Academy of Science (St. Petersburg).

COI barcode sequence of the holotype (BOLD process ID BPAL2191-13; GenBank accession number # KY777529): ACTTTATATTTTATCTTTGGAATTTGAGCAGGTATATTGGGAACTTCTTTAAGACTTTTAATTCGAACTGAATTAGGAA

ATCCAGGATCTTTAATTGGTGATGATCAAATTTATAATACTATTGTTACAGCTCATGCTTTTATTATAATTTTTTTTATAGT

TATACCTATTATAATTGGAGGATTTGGAAATTGATTAGTTCCTTTAATGTTAGGAGCCCCTGATATAGCATTCCCACGAATA

AATAATATAAGATTTTGATTGCTCCCCCCCTCATTAATCTTATTAATTTCTAGAAGAATTGTAGAAAATGGTGCAGGTACAG

GATGAACAGTTTACCCCCCACTTTCATCCAATATTGCTCATAGAGGATCATCTGTTGATTTAGCAATTTTTTCTCTTCATTT

AGCTGGAATTTCTTCAATTTTAGGGGCTATTAATTTTATTACCACTATTATTAACATACGCATTAATAATATATCATTCGAT

CAAATACCTTTATTTGTTTGAGCTGTAGGTATTACAGCTCTTTTATTATTATTATCTTTACCAGTTTTAGCAGGAGCAATTA

CAATACTTCTTACTGATCGAAATATTAATACTTCATTTTTTGACCCTGCTGGAGGAGGAGATCCTATTTTATACCAACATTTA

Paratypes.

26 males and 10 females collected on Mt. Hermon, Israel.

Four males with codes CCDB-17949_E01, BPAL2234-13; KT874736, BPAL2236-13, CCDB-17949_E03; CCDB-25452_C10, BPAL3359-16 and BPAL3360-16, CCDB-25452_C11. Two females with codes BPAL3361-16, CCDB-25452_C12; CCDB-17949_E02, KT874697, BPAL2235-13. Two females without codes. Israel, Mt. Hermon, 33°18'45.6"N; 35°47'11.9"E, 2040 m, 22 June 2013, V.A. Lukhtanov & A. Novikova leg.

Four males with codes CCDB-25453_E10, BPAL3193-16; CCDB-25453_E08, BPAL3191-16; CCDB-25453_E09, BPAL3192-16; CCDB-25454_C03, BPAL3257-16; CCDB-25453_E11, BPAL3194-16. Six males and one female without codes. Israel, Mt. Hermon, 33°18'20"N; 35°47'09"E, 2030 m, 17 May 2014, A. Novikova leg.

One male with codes CCDB-17969_A04, BPAL2759-15. Israel, Mt. Hermon, 33°18'45.6"N; 35°47'11.9"E, 2040 m, 03 July 2014, V.A. Lukhtanov & A. Novikova leg.

Ten males with codes CCDB-25458_C06, BPALB125-16; CCDB-25458_C07, BPALB126-16; CCDB-25458_C08, BPALB127-16; CCDB-25458_C09, BPALB128-16; CCDB-25458_C10, BPALB129-16; CCDB-25458_C11, BPALB130-16; CCDB- 25458 _C12, BPALB131-16; CCDB-25458_D01, BPALB132-16; CCDB-25458_D02, BPALB133-16; CCDB-25458_D07, BPALB138-16. One male and three females without codes. Israel, Mt. Hermon, 33°18'41"N; 35°46'49"E, 1750-1900 m, 03 May 2016, V.A. Lukhtanov & E. Pazhenkova leg.

Two females with codes 25458 E06, BPALB149-16; 25458_E08 BPALB151-16. Israel, Mt. Hermon, 33°18'51"N; 35°46'31"E, 1800 m, 07 May 2016, V.A. Lukhtanov, A. Novikova & E. Pazhenkova leg.

All paratypes are deposited in the Zoological Institute of the Russian Academy of Science (St. Petersburg).

Males

(Fig. 2 a–c). Forewing length 16-19 mm. Forewing is roundish.

Upperside: ground color orange-red; the wing markings are small and delicate when compared to those in M. didyma and M. persea . Forewings with very narrow black marginal border fused with internervural marginal black spots. Submargimal series formed by black triangular spots on the forewings and by fine lunules on the hindwing. Forewing postdiscal series formed by small 1-3 black spots. Forewing discal series is complete or nearly complete, formed by black spots of variable size, the first four spots near costa are often enlarged. Hindwing discal series reduced or absent. Basal marking of the fore- and hindwings is delicate. Black basal suffusion is developed only near the base of hindwings. Fringe is white, checkered by black dots.

Underside: forewing ground color orange-red except for the apical part which is yellowish. Black markings delicate, reduced as compared with those of the upperside of the wing. Hindwing ground colour yellowish-white with two orange-red fascias. The red-orange submarginal fascia shows segmentation as the yellowish-white ground color spreads along the nervures. The orange-red macules of this fascia are bordered by back lunules from the outer side. From the inner side these macules are edged by black scales and additionally bordered by black lunules, giving the appearance that the proximal border of the submarginal fascia is doubly edged. Fringe white, checkered by black dots.

Females

(Figs 2d, 3). Forewing length 17-20 mm. Forewing is roundish. Ground color of the upperside is slightly lighter and black markings heavier than in males. Costal area of the wing apex yellow-orange. Underside of the forewings as in males but black markings are heavier and there are additional yellowish maculae between discal and postdiscal spots. Underside of the hindwings as in males. Fringe white, checkered by black dots.

Karyotype.

The genus Melitaea is known for its relatively low interspecific chromosome number variation ( Pazhenkova and Lukhtanov 2016). However, in certain cases, the chromosome numbers are key to distinguish between closely related Melitaea species. For example, karyotype differences in combination with information about parapatric distribution were the main argument for the non-conspecificity of M. didyma and M. latonigena Eversmann, 1847 ( Lukhtanov and Kuznetsova 1989). Therefore, chromosomal analysis is highly desirable in any taxonomic study of Melitaea . Here I conducted the chromosomal analysis of the high altitude population from Mt. Hermon ( M. acentria sp. n.). The haploid chromosome number n=27 was found in prometaphase I, MI and MII cells of three studied individuals (2016-006, CCDB-25458_C11; 2016-008, CCDB-25458_D01; 2016-009, CCDB-25458_D02) (Fig. 4). The MI karyotype contained one chromosome bivalent that was significantly larger than the rest of the bivalents.

The same chromosome number (n=27) was previously reported for M. persea from Iran ( de Lesse 1960). A karyotype characterized by n=27 including one large chromosome element was also found in M. didyma neera Fischer de Waldheim, 1840 from the North Caucasus (Russia), although in some other studied populations of M. didyma n =28 was found ( de Lesse 1960, Lukhtanov and Kuznetsova 1989). The chromosome number n=27 was also mentioned for " M. didyma libanotica " from Lebanon ( Larsen 1975), but the vouchers for this chromosomal analysis were larvae, and in my opinion their identification was not certain. They could represent M. didyma liliputana, but also M. acentria as well as M. persea (but certainly not M. deserticola in which n=29 was found and not M. trivia in which n=31 was found, Larsen 1975). Finally, n=27 was reported for " M. montium " from Lebanon ( de Lesse 1960), but in the last case the identity of the studied samples was also not clear because the identification was not supported by genitalia analysis.

Thus, no fixed karyotype difference is known to exist between M. acentria and M. persea as well as between M. acentria and M. didyma . Therefore we cannot use the available chromosomal data for delimitation between these species.

Male genitalia structure.

M. didyma from Israel (Mt. Hermon) and M. persea from Iran and Azerbaijan were analyzed and were found to possess typical characters described previously ( Higgins 1941, Oorschot and Coutsis 2014).

In M. persea all the main structures (ring-wall, tegumen, saccus, valvae) are elongated (Fig. 5a, b), longest in the genus Melitaea ( Oorschot and Coutsis 2014). The valva is elongated from lateral view (Fig. 6a) and the valval distal process is massive (Fig. 6b). The dorsum of the valval distal process lies nearly in line with the remainder of the valval dorsum (Fig. 6a). The ventrum of the valval distal process possesses a keel bearing strong teeth (Fig. 6b). The saccus is bifurcate, with long, distally pointed branches (Fig. 5a, b). The aedeagus is curved, with a pronounced dorso-lateral ridge (Fig. 7a). The lateral sclerotized element of the tegumen is massive and its distal half is shaped like a smoker’s pipe (Fig. 7b).

In M. didyma liliputana from Mt. Hermon all the main structures (ring-wall, tegumen, saccus, valvae) are significantly shorter than in M. persea (Fig. 5e, f). The valva is trapezoidal from lateral view (Fig. 6e). The valval distal process is delicate (Fig. 6f) and the dorsum of the valval distal process forms a clear angle with the remainder of the valval dorsum (Fig. 6e). The ventrum of the valval distal process is smooth, without a keel and/or teeth (Fig. 6f). The saccus is bifurcate, with short, distally rounded branches (Fig. 5e, f). The aedeagus is curved, without a pronounced dorso-lateral ridge (Fig. 7e). The lateral sclerotized element of the tegumen is delicate and its distal half is T- or Γ-shaped (Fig. 7f).

In M. acentria genitalia are clearly different from both M. persea and M. didyma , but at the same time are intermediate in some aspects. All the main structures (ring-wall, tegumen, saccus, valvae) are similar to those in M. persea but shorter (however, longer than in M. didyma ) (Fig. 5c,d). The valva is cylindrical from lateral view (Fig. 6c). The valval distal process is intermediate in its shape between M. persea and M. didyma (Fig. 6c, d). Its dorsal and ventral borders are roughly parallel from lateral view (Fig. 6c). The dorsum of the valval distal process forms a clear angle with the remainder of the valval dorsum (similarly to M. didyma ) (Fig. 6c). At the same time, the ventrum of the valval distal process possesses a keel bearing teeth (similarly to M. persea ) (Fig. 6d). However, this keel and teeth are smaller and more delicate than in M. persea (Fig. 6d). The saccus is bifurcate, with relatively long, distally pointed branches; however, these branches are shorter than in M. persea , but longer than in M. didyma liliputana, where they are almost absent (Fig. 5c, d). The aedeagus is curved, with a dorso-lateral ridge (Fig. 7c); thus the aedeagus of M. acentria is not intermediate between M. persea and M. didyma , but similar to M. persea . The lateral sclerotized element of the tegumen is massive and its distal half is shaped like a smoker’s pipe (Fig. 7d). This type of male genitalia was found in all seven studied samples including two samples (25453_E08 and 25458_C09) that were characterized by the mitochondrial haplogroup P2 (Figs 8 and 9).

COI barcode analysis.

The COI barcode analysis revealed five major clusters represented by (1) M. trivia syriaca, (2) M. deserticola , (3) M. didyma liliputana, (4) M. casta and (5) taxa of the M. persea group (haplogroups A, H, P1, P2 and P3) (Fig. 8). Interestingly, this analysis showed that the phenotypically similar species M. trivia syriaca, M. deserticola and M. didyma liliputana, can be easily separated by their DNA barcodes.

The analysis recovered the M. persea group ( M. acentria + M. persea + M. higginsi ) as a strongly supported monophyletic clade sister to M. casta (Fig. 9). This clade was divided into five lineages.

The first lineage (haplogroup P1) includes a huge range of M. persea populations from Daghestan (Russia) in the north to Shiraz province (Iran) in the south, including samples from Shiraz in SW Iran, which represents the type locality of M. persea . Across this range, M. persea shows various degrees of localized morphological diversification, and from this territory several taxa, currently attributed to M. persea , were described: M. didyma caucasica Staudinger, 1861; M. didyma kaschtschenkoi Christoph, 1889; M. didyma araratica Verity, 1929; M. didyma magnacasta Verity, 1929; Melitaea tauricus Belter, 1934; M. pesea hafiz Higgins, 1941; M. hafiz darius Gross & Ebert, 1975 and M. jitka D.Weiss & Major 2000. The taxonomy of these taxa was studied in more detail by Oorschot and Coutsis (2014), who found that they are closely related and should be considered no more than synonyms of M. persea persea. My DNA barcode results are consistent with this conclusion (Fig. 9).

The haplogroup P1 includes also a female sample 17966_F12 possessing intermediate morphological characters between M. interrupta and M. persea (Fig. 10b). In this specimen wing upperside is similar to that in M. interrupta , whereas the wing underside is without black scales along the veins which are typical for M. interrupta (Fig. 10h), but with orange-red submarginal spots edged by black scales typical for M. persea (Fig. 2 e–h, 10a). This sample was collected at the same place with three typical M. interrupta males (samples17966_F09, 17966_F10 and 17966_F11) possessing typical M. interrupta phenotype (Fig. 10h) and COI haplotypes (GenBank # KT874702, KT874740 and KT874741), which were very different from those of M. persea (see Fig. 4 in Pazhenkova and Lukhtanov 2016). It is thus probable that the female 17966_F12 is a result of a more or less recent hybridization between M. interrupta and M. persea . Thus, it likely represents a first molecular evidence for sporadic interspecific hybridization in Melitaea .

The second lineage (haplogroup P2) is represented by three specimens from north Lebanon originally identified as M. persea (Wahlberg et al. 2005) and by three samples of M. acentria from Mt. Hermon: two males (25453_E08 and 25458_C09) that were indistinguishable in their genitalia from M. acentria of the haplogroup A and a single female (25458_E08). This lineage was found to be closest to P1 ( Melitaea persea persea). It differed from P1 by 7 fixed DNA substitutions in the studied 658 bp fragment of the mitochondrial COI gene. The minimal uncorrected COI p-distance between the representatives of these two haplogroups was calculated using both fixed and non-fixed substitutions and was found to be 2.0 %.

The third lineage (haplogroup P3) includes samples from NE Iran ( M. persea paphlagonia). It differed from P1 ( M. persea persea) by 10 fixed DNA substitutions in the studied 658 bp fragment of the mitochondrial COI gene. The minimal uncorrected COI p-distance between these two haplogroups was found to be 2.3 %. They were also distinct in wing pattern: on the upper surface all the markings were well developed and the first four spots of the discal series were nearly fused to form a prominent costal bar (Fig. 10c, d). The male genitalia of M. persea paphlagonia were similar to those found in M. persea persea ( Higgins 1941). This lineage was not recognized as a taxon by Oorschot and Coutsis (2014). However, it was recognized as a distinct subspecies by Higgins (1941), and my DNA barcode results corroborate this conclusion. The level of COI differentiation between M. persea paphlagonia and M. persea persea (10 fixed DNA substitutions) was found to be equal to that found between M. persea persea and M. higginsi (10 fixed DNA substitutions).

The forth lineage (haplogroup A), one of the most diverged lineages, is represented by samples from Mt. Hermon ( M. acentria ). It differed from P1 ( M. persea persea) by 11 fixed nucleotide substitutions in the studied 658 bp fragment of the mitochondrial COI gene. The minimal uncorrected COI p-distance between these two haplogroups was found to be 2.4 %.

The fifth lineage (haplogroup H) includes samples of M. higginsi (Fig. 10e, f). This taxon is very rare in collections, and I have been lucky to find two specimens in the McGuire Center. It differed from P1 ( M. persea persea) by 10 fixed DNA substitutions in the studied 658 bp fragment of the mitochondrial COI gene. The minimal uncorrected COI p-distance between these two haplogroups was found to be 2.4 %. This taxon is similar to M. persea with respect to male genitalia structure ( Oorschot and Coutsis 2014), but quite different in wing pattern. Particularly, in males the hindwing uppersurface is without black spots which are always present in M. persea , and in both sexes hindwing underside veins are scaled with black, similar to M. interrupta and different from M. persea . My DNA barcode results confirm the distinctness of this high altitude very local Afghani taxon. They also confirm that this taxon is a member of the M. persea species group as suggested by Oorschot and Coutsis (2014), and not related to the Mongolian M. didymina Staudinger, 1895 as was supposed by Sakai (1978), as well as not related to M. didyma as was supposed by Kolesnichenko and Churkin (2004).

Diagnosis.

Butterfly wing pattern and male genitalia morphology, as well as DNA barcodes certaintly indicate that Melitaea acentria belongs to the M. persea species complex. After Oorschot and Coutsis (2014) this complex includes three closely related species: M. persea , M. eberti and M. higginsi . Male genitalia of these three species were analysed by Oorschot and Coutsis (2014) and were found to be virtually indistinguishable. Melitaea acentria differs from these most closely related species by several characters in male genitalia. In M. acentria main genitalia structures (ring-wall, tegumen, saccus, valvae) are significantly shorter. The valva is cylindrical from lateral view, not elongated (Fig. 6c). The valval distal process is intermediate in its form between M. persea and M. didyma (Fig. 6c, d). The dorsum of the valval distal process forms a clear angle with the remainder of the valval dorsum (in similar way as in M. didyma ) (Fig. 6c), but very different from M. persea . The keel and teeth of the valval distal process are smaller and more delicate than in M. persea (Fig. 6d). On average, in M. acentria the ground color of the wing upperside is more orange-red (Fig. 2 a–d). In other species of the M. persea complex it is yellowish-orange (Fig. 2e-g, 10a, c–f). However, this character is not constant (e.g. see M. persea with orange-red wing color on Fig. 2h). The great majority of M. acentria samples significantly differ from all other taxa by their DNA barcodes; however, probably due to mitochondrial introgression, a minor part of the samples cluster with the haplogroup P2 of M. persea .

Melitaea acentria significantly differs from the distantly related but phenotypically similar species M. didyma , M. deserticola and M. trivia by DNA barcodes and male genitalia structures. Particularly, it differs from M. didyma by the ventrum of the valval distal process possessing a keel bearing teeth and by the elongated shape of the ring-wall, tegumen, saccus and valvae. Melitaea acentria mostly differs from M. didyma by the hindwing underside with submarginal macules that are edged by black scales and then bordered by black lunules, giving the impression that the proximal border of the submarginal fascia is doubly edged; M. acentria shares this character with M. persea . In M. didyma submarginal macules of the hindwing underside are usually not edged by black scales and simply bordered by black strokes (Fig. 10g). However, elements of the black scaling of the submarginal macules can be found in few M. didyma samples, and sometimes this black scaling is strongly reduced in species of the M. persea complex.

Distribution.

Melitaea acentria is known to occur at high altitudes (1730-2060 m above the sea level) of Mt. Hermon (Fig. 11). Within these altitudes it is sympatric and syntopic with M. trivia syriaca, M. deserticola and M. cinxia . At the altitudes 1730-1780 m there is an essential overlapping of the M. acentria and M. didyma liliputana ranges where both species were found to fly together in early May 2016. Two other Melitaea species known from Mt. Hermon, M. collina and M. telona , were found to fly mostly at lower altitudes 1000-1600 m.

Habitat and phenology.

Three main vegetation belts have been described from Mt. Hermon: (i) evergreen Mediterranean maquis (300-1250 m); (ii) xero-montane open forest (1250-1850 m) and (iii) subalpine mountain steppe, or ‘‘ Tragacanthic belt’’ (1850-2814 m) ( Kent et al. 2013). Adults of M. acentria were found to fly in open grassy (Fig. 12) and stony (Fig. 13) areas of the upper part of the xero-montane open forest belt (1750-1850 m) (Fig. 14) and of the subalpine mountain steppe belt (1850-2060 m) (Fig. 15). Butterflies were observed from 3 May to 3 July. On the 3rd of May 2016 they were abundant at altitudes from 1780 to 1900 m, therefore I conclude that they can start to fly at the end of April and continue to fly at least until mid-July.

Etymology.

The name acentria is a noun of the feminine gender. This name originates from the Greek prefix “a” that means “not” and from the Latin word “centrum” (centre) derived from the Greek “κέντρον” (kentron, a sharp point). Acentria is the Internet nickname of Asya Novikova who collected the samples initiated this research.This name indicates also the peripheral position of the new species within the distribution range of the M. persea species complex.