﻿Integrative analysis reveals cryptic speciation linked to habitat differentiation within Albanian populations of the anomalous blues (Lepidoptera, Lycaenidae, Polyommatus Latreille, 1804)

﻿Abstract The Balkan Peninsula is one of the greatest hotspots for biodiversity in Europe. While the region has been investigated thoroughly, some parts remain understudied and may still harbour undiscovered diversity, even in well-studied organisms such as Lepidoptera. Here we investigated the group of the so-called anomalous blue butterflies, also known as ‘brown complex’ of the subgenus Agrodiaetus Hübner, 1822 including the taxa of the entire Polyommatusaroaniensis (Brown, 1976) species complex. This species complex is distributed in the southern part of the Balkan Peninsula and known to be represented by three closely related allopatric species, differentiated by their chromosome numbers (n) and mitochondrial (mt) DNA. These are P.aroaniensis sensu stricto (Southern Greece, Peloponnese, n=47–48; mt haplogroup aroa1), P.timfristos Lukhtanov, Vishnevskaya et Shapoval, 2016 (Central Greece, Attika, n=38, aroa2) and P.orphicus Kolev, 2005 (North-Eastern Greece, Southern Bulgaria, n=41–42, orph1). Based on an analysis of chromosomal, molecular and morphological markers, we demonstrate that a fourth taxon of this species complex exists in Albania. This taxon possesses the mt haplogroup aroa3, which is the most differentiated within the entire P.aroaniensis species complex, and the karyotype (n=42–43), which differs by one fixed chromosome fission from P.orphicus. The Albanian taxon seems to be ecologically specialised (habitat on dark-coloured, ophiolitic substrate soils) and differs in colouration (wing reflectance) from the others taxa of the P.aroaniensis species group. Based on the evidence here presented and following the current view of the taxonomy of the group, we propose considering the Albanian taxon as a new species, here described as Polyommatusluraesp. nov. At the contact zone between the new species and P.orphicus, in addition to typical ones, we detected specimens with haplogroup orph2, karyotype n=43 and intermediate morphology, which seem to represent P.lurae × P.orphicus hybrids.


Introduction
The so-called anomalous blues of the subgenus Agrodiaetus Hübner, 1822 constitute a distinct lineage within the species-rich genus Polyommatus Latreille, 1804 (Talavera et al. 2013a). The subgenus Agrodiaetus is distributed throughout the Western Palaearctic and Central Asia. With over 120 species described, this subgenus is a striking example of explosive radiation in the last three million years (Kandul et al. 2004;Kandul et al. 2007;Talavera et al. 2013b). Sexes are often dimorphic, with females usually brown in colour, and with males displaying colours such as blue, white, silver, orange, violet, or brown on the dorsal part of the wings, which is probably a signature of the reinforcement of pre-zygotic isolation (Lukhtanov et al. 2005;Dincă et al. 2017).
In Europe, most Agrodiaetus species are restricted to the southern warm parts of the continent, and most of them are endemic to relatively small areas. The Balkans are one of the richest European regions for subgenus Agrodiaetus (Kudrna et al. 2015) and, currently, the following taxa of Agrodiaetus with brown males are recognised as present in this region: Polyommatus admetus (Esper, 1783), Polyommatus ripartii (Freyer, 1830) and those under the so-called Polyommatus aroaniensis Brown, 1976 species complex. In this paper, we focus on the latter species complex, which remains still insufficiently studied, especially in underexplored regions like Albania.
The first species of "brown" Agrodiaetus described as endemic to the Balkans was P. aroaniensis (Brown, 1976). After Coutsis (1972) attracted attention to a population of "P. ripartii" on Mt. Chelmos in Southern Greece, noting that in "about half of the specimens the white streak on the hindwing underside, always present in ripartii, was completely absent". Brown (1976) wrote about this population from Chelmos as an "unrecognized Agrodiaetus sp. similar and often sympatric with ripartii [pelopi] in Greece" (Kolev and van der Poorten 1997). The haploid chromosome number initially identified by Brown (1976) as n=15-16 was later corrected to n=47-48 (Lukhtanov et al. 2003). Apart from the unique chromosome number, typical P. aroaniensis is characterized by a coffee-brown colour with a perceptible reddish hue and by the lack of darker marks along the margins. While P. aroaniensis was apparently first discovered based on the absent or vestigial white streaks on the hindwings, it was later discussed that this trait was only valuable for about 60% of the specimens in the population (Kolev and van der Poorten 1997). Using the white stripe state (absent, reduced or prominent presence) as an identification trait for P. aroaniensis turned out to be difficult, as has been recently shown by the misidentification of P. aroaniensis in Croatia. Indeed, some specimens lacking this white stripe were initially determined as P. aroaniensis based only on this morphological trait, and were later corrected into P. ripartii by the use of Generally, identification of species in the subgenus Agrodiaetus remains challenging because of considerable geographic and individual variability in habitus (Dincă et al. 2013), e.g. the state of the white stripe on the underside of hindwings. However, modern identification techniques based on genetic markers in combination with karyological data, often obtain reliable differentiation between Agrodiaetus taxa (Lukhtanov and Dantchenko 2002;Wiemers 2003;Kandul et al. 2004;Lukhtanov et al. 2006;Vila et al. 2010). Besides, at species level, subtle fixed variations in traits such as wing colour are still useful for identification in the Agrodiaetus subgenus and other Polyommatus sp(p)., especially male dorsal wing colour (Lafranchis 2004). Indeed, new morphometric techniques based on measurements of wing colour and its reflectance pattern have been tested for species identification and delimitation. For example, remarkably good correlations with genetic markers have been found for Lysandra bellargus and Polyommatus icarus (Kertész et al. 2021;Piszter et al. 2021) and this colour morphometric technique can also be useful for discriminating between species of the Agrodiaetus subgenus.
Thus, an integrative analysis combining multiple markers and techniques may be the best solution to resolve complex systematics and to uncover potential cryptic diversity in the Agrodiaetus subgenus (Lukhtanov and Dantchenko 2017;Kertész et al. 2021). By using such an integrative approach, based on combined molecular, cytogenetic, and colour morphometric analyses, we here demonstrate that a fourth taxon within the entire Polyommatus aroaniensis species complex is present in Albania.

Sample collection and storage
All butterflies were collected by L. Parmentier at the different biotopes investigated in Albania, in the provinces of Korçë, Elbasan, Dibër and Kukës. Collected samples were put in glassine envelopes in the field, a unique code assigned to each and stored in cooled plastic boxes. At different sites, a selection of fresh male samples was kept alive, until the posterior part of the abdomen was removed for karyological analysis. Taking into account the possibility of multiple cryptic species within a local area even in well-studied European butterflies (Dincă et al. 2011;Hinojosa et al. 2022), multiple individuals were collected in each place, paying special attention to the specimens with unusual or intermediate morphology (Vishnevskaya et al. 2016). Unless otherwise noted, all collected samples were imagoes that were spread and dried to be used for the analysis of habitus, and all are stored in L. Parmentier's private collection (LPAcoll, Zulte, Belgium). Legs used for DNA analysis are stored in R. Vila's Collection (RVcoll, Institute of Evolutionary Biology, Barcelona, Spain) and karyotype plates in the Zoological Institute of the Russian Academy of Sciences. Pictures of biotopes were taken with an Iphone 6 and butterflies with a Canon 70D and a 100mm macro lens.

DNA extraction and sequencing
In order to elucidate the genetics of the subgenus Agrodiaetus in Albania, we analysed DNA of 41 Albanian specimens (males and females). To put them in context, we mined 19 additional COI sequences from GenBank, a subset where the most similar sequences to the Albanian ones were included and overlapped at least 650 base pairs (bp) of the cytochrome c oxidase subunit (COI). DNA extraction was done following the protocol described in Dincă et al. (2017). The primers LepF1 and LepR1 (ATTCAACCAATCATAAAGATATTGG and TAAACTTCTGGATGTC-CAAAAAATCA respectively) were employed for the COI amplification, obtaining a full DNA barcode fragment of 658 base pairs (bp). Double-stranded DNA was amplified in 25 μl volume reactions: 14.4 μl ultra-pure (HPLC quality) water, 5 μl 5X buffer, 2 μl 25mM MgCl 2 , 0.5 μl 10 mM dNTP, 0.5 μl of each primer (10 μM), 0.1 μl Taq DNA Polymerase (Promega) and 2 μl of extracted DNA. Conditions for the PCR cycles were set as follow: first denaturation step at 92 °C for 60 s, then 92 °C for 15 s, 48 °C for 45 s and 62 °C for 150 s in 5 cycles and other 30 cycles changing the annealing temperature to 52 °C with the final extension step at 62 °C for 7 min. A 411 to 440 bp fragment at the 5' end of the nuclear ITS2 was amplified by polymerase chain reaction using the primers ITS3 (GCATCGATGAAGAACGCAGC) and ITS4 (TCCTCCGCTTATTGATATGC) (White et al. 1990). The reactions were prepared as for COI but, in this case, the typical thermal cycling profile was: 95 °C for 45 s, 51 °C for 60 s and 72 °C for 60 s, for 40 cycles. PCR products were purified and Sanger sequenced by Macrogen Inc. Europe (Amsterdam, the Netherlands). All new COI and ITS2 sequences have been deposited in GenBank (ON715895-ON715938 for COI and OP537924-OP537930 for ITS2) (Table 1).

Phylogenetic inference
The COI analysis involved 60 sequences (19 GenBank sequences and 41 own material). For the COI phylogeny, sequences of different length (from 647 to 657 bp) were included into the final dataset alignment. We used Geneious Prime 2019.0.3 (https:// www.geneious.com) software to align the sequences and then edited them manually. The final COI alignment included 657 sites, with 137 variable sites and 112 parsimony-informative sites. A phylogeny was reconstructed in BEAST v2.5.0 (Bouckaert et al. 2014). Parameters were estimated using two independent runs of 30 million generations each and convergence was checked with TRACER 1.7.1 (Rambaut, 2018). A burn in of 10% was applied. Samples of P. (A.) damon were used to root the tree. Besides, a haplotype network of the COI barcode region was created in POPART v1.7 (Leigh & Bryant, 2015) using the TCS method.
Based on Wiemers et al. (2009), ITS2 secondary structure improves phylogeny estimation of the subgenus Agrodiaetus and thus we combined mitochondrial and nuclear sequences to improve phylogenetic signal, in agreement with Vishnevskaya et al. (2016). This resulted in a concatenated COI + ITS2 alignment with a total of 1039 bp. Phylogeny on concatenated sequences was reconstructed in BEAST v2.5.0 (Bouckaert et al. 2014) on a subset of Albanian specimen, covering all study sites, and extra sequences mined from GenBank (9 own material and 45 extra sequences). Phylogenetic relationships were inferred using Bayesian Inference (BI), maximum likelihood (ML) and maximum parsimony (MP) analyses. The Bayesian analysis of the concatenated matrix COI+ITS2 was performed using the program MrBayes 3.2 (Ronquist et al. 2012) with default settings as suggested by Mesquite (Version 3.04. http://mesquiteproject.org): burn-in = 0.25, nst = 6 (GTR + I + G). Two runs of 10 million generations with four chains (one cold and three heated) were performed. The first 25% of each run was discarded as burn-in. The consensus of the obtained trees was visualized using FigTree 1.3.1. (http://tree.bio.ed.ac.uk/software/ figtree/). The samples of P. damon were used to root the tree. The NJ analysis of the concatenated matrix COI+ITS2 was performed using the program Mega X (Kumar et al. 2018) and Tamura3-parameter+G as the optimal model (also estimated by Mega X). The samples of P. damon were used to root the tree. The standard nonparametric bootstrap (Felsenstein, 1985) (100 replicates) was used to evaluate statistical nodal support of the tree.

Analysis of karyotype
Testes were removed within 1 hour after collection and were stored in the 3:1 fixative for several months at +4 °C and then stained with 2% acetic orcein for 30 days at 20 °C. In total we karyotyped a selection of 15 samples representative for the different species and biotopes handled in this paper. We used a two-phase method of chromosome analysis as described in Lukhtanov and Dantchenko (2002). In the first phase, the stained testes were placed into a drop of 40% lactic acid on a slide, the gonad membranes were torn apart using fine needles and intact spermatocysts were removed and transferred into another drop of 40% lactic acid. Intact spermatocysts were studied and photographed. The first phase was most useful for counting the number of chromosome bivalents and multivalents. In the second phase, different stages of chromosome spreading were studied using a slight, gradually growing pressure on the coverslip. The second phase was most useful for studying the chromosome structure and distinguishing between bi-and multivalents. By scaling up the pressure on the coverslip, we were able to manipulate chromosomes, e.g. change their position and orientation on the slide, and consequently to resolve controversial cases of contacting or overlapping bivalents. Haploid chromosome numbers were counted at metaphase I (MI) and/ or metaphase II (MII) of meiosis.
Leica DM2500 light microscope equipped with HC PL APO 100×/1,44 Oil CORR CS lens and S1/1.4 oil condenser head was used for bright-field microscopy analysis. Leica lens HC PL APO 100×/1,40 OIL PH3 was used for phase-contrast microscopy analysis.

Morphometric measurements of habitus (dorsal wing reflectance)
Wing colour is an important trait for identification of butterflies and a species-specific characteristic (Bálint et al. 2012), an indicator of genetic variation (Wasik et al. 2014), and evidence of a changing population (Hiyama et al. 2012;Kertész et al. 2021). Observing fixed differences in wing colour of butterflies of different population can serve as a reliable tool for species identification (Bálint et al. 2010). Here we used colour measurements of dorsal wings of male Agrodiaetus to generate standardized RGB measurements of set specimens. In our set-up a constant light source in a darkened room was used (3 Marbul® suspension light sources of 12 W, 955 lm, 3500 K) in a triangle position at 1 m above the specimen to obtain a reproducible and uniform light source and minimize shades. RGB pictures were made on specimens positioned at an angle of 20° to the equatorial to measure maximum light reflectance generated by wing scale structures. The spectral position of the reflectance maximum of such photonic nanoarchitectures depends on the nanoscale geometric dimensions of the elements building up the nanostructure and was based on earlier experience and method described by Kertész et al. (2021). To obtain a uniform colour zone, the intervein space -which showed most reflectance variability-of the inner postdiscal zone (only between M1 and CU2 cells) was used. Per measurement a circular area of the wing was blurred in Lightroom and the average colour obtained was mapped on a disc. On this uniform colour discs (3 samples per wing), colour measurements were done generating exact RGB and HUE values using the colour picker tool. Averaged values per specimen were used for statistics. Wing colour measurements were taken only from fresh samples (worn specimens and those with minimal damage on fringe were discarded) belonging to the Polyommatus aroaniensis species complex collected from different Albanian localities, with habitats harbouring ophiolitic, karst and mixed substrates, as detailed in Suppl. material 1.
Posterior statistics was done running a permutational multivariate analysis of variance, using distance matrices with the Adonis call (Vegan package) in R (Oksanen et al. 2016).

Biotopes and Albanian samples
In total, 251 Agrodiaetus samples belonging to the P. aroaniensis and P. ripartii species complexes were analysed. Specimens were collected from 94 Albanian different sites visited and distributed from Southern up to North-eastern Albania. Only in a minor part of the sites (7/94 sites), specimens were identified as belonging to the P. aroaniensis species complex, mostly in provinces Dibër and Korçë. Specimens with clear P. ripartii traits -based on habitus descriptions given by Tshikolovets (2011) and Vishnevskaya et al. (2016) -were excluded from the analyses, after some voucher samples were barcoded to confirm our identifications (data not shown). Especially the study sites in the Dibër province, Lurë region, attracted our attention because many individuals displayed a remarkably dark habitus and mostly were completely lacking the white stripe on the hindwing underside (Fig. 1a, b). Their biotope was atypical, with a dark ophiolitic substrate (Fig. 1d). Here, in the vicinity, Onobrychis alba, a known foodplant of different Agrodiaetus species, was found growing (Fig. 1e, f ). Besides, in another region, near Valikardhë village, a second cluster of study sites potentially harbouring orphicus/aroaniensis populations ( Fig. 1c) was identified, although the habitat soil type was visibly paler and consisted of typical karst (chalk) substrate.
To construct the haplotype network, we used 54 specimens that were collapsed in 28 haplotypes representing 7 haplogroups (Fig. 3) with P. damon as outgroup: 3 haplogroups for the P. ripartii species complex (including P. ripartii ripartii and P. ripartii pelopi), 2 for P. orphicus (orph 1, and orph 2 representing the Albanian lineage) and 3 for the P. aroaniensis s. l. clade including P. timfristos (aroa 2), P. aroaniensis (aroa 1) and P. lurae sp. nov (aroa 3). The latter haplogroup is clearly distinct from related taxa as can be seen in the number of nucleotide substitutions between them.
The phylogenies based on the concatenated COI+ITS2 sequences (54 specimens, same as the haplotype network) and using the BI and NJ methods are given in Suppl. material 2. The NJ analysis of the concatenated matrix (using Tamura3+G as the best model) (Suppl. material 2, Fig. 1) revealed lurae as a highly supported, differentiated lineage with the most basal position within the entire aroaniensis species complex. However, this basal position had a low support. The BI analysis of the concatenated matrix (using default settings) (Suppl. material 2, Fig. 2) revealed lurae as a highly supported lineage, sister to aroaniensis (but the support for this sister relationship was very low). Generally, the concatenated alignments revealed the same topology as in the case of the COI tree, with very good support of P. lurae sp. nov as a well differentiated lineage, and putative sister to P. aroaniensis, although with lower support.
In all cases using different phylogenetic reconstruction models, and based on COI or concatenated COI+ITS2 sequences, P. lurae sp. nov formed a monophyletic, well differentiated clade with very good support.

Polyommatus orphicus from Albania (populations in which the only orph2 haplogroup is present)
In five studied samples (K75, K76, K80, K84, K85) the number of countable elements was found to be n=42 at MI and MII cells. Bivalents at MI and univalents at MII were fairly well differentiated with respect to their size; however, it was difficult to subdivide them objectively into size groups because the sizes of the elements decrease more or less linearly (Fig. 4). Brown dots next to specimens within the orph2 clade indicate P. lurae × P. orphicus potential hybrids (n=43-44) found in the contact zone or dispersing.

Polyommatus lurae sp. nov from Albania (populations in which the only aroa3 haplogroup is present)
In five studied samples (K68, K71, K73, K81, K88), two different haploid chromosome numbers (n=42 and n=43) were observed at MI and MII cells of the 14 specimens studied. This variation was most likely caused by polymorphism for one chromosome Figure 3. Haplotype network of the P. aroaniensis species group in relation to P. damon and P. ripartii. Coloured circles represent different taxa, as indicated in the legend; coloured boxes delimitate the haplogroups discussed in the text. Each line segment represents a mutation step, and black small circles represent "missing" haplotypes. fusion/fission. This polymorphism resulted in three types of MI karyotype: n=42 (homozygous for chromosomal fusion/fission, one pair of fused chromosomes; 2n=84), n=43 (homozygous for chromosomal fusion/fission, two pairs of unfused chromosomes; 2n=86) and n=42 (heterozygous for chromosomal fusion/fission, resulting in 41 bivalents and one trivalent; 2n=85). Bivalents at MI and univalents at MII were fairly well differentiated with respect to their size; however, it was difficult to subdivide them objectively into size groups because the sizes of the elements decrease more or less linearly (Fig. 5).

Contact zone with Polyommatus lurae × P. orphicus potential hybrids (area where both orph2 and aroa3 haplogroups are present)
The contact zone between the two species was defined as the area where both orph2 and aroa3 haplogroups were found to coexist, and coincided with a mixed ophiolite/karst substrate. In this area, in addition to specimens typical to either species, a number of countable elements of n=43 at MI and MII cells was found for three samples with a mitochondrial haplogroup orph2 (samples K69, K83, and K90). Bivalents at MI and univalents at MII were fairly well differentiated with respect to their size; however, it was difficult to subdivide them objectively into size groups because the sizes of the elements decrease more or less linearly (Fig. 6). In the sample 17E542 (also haplogroup orph2) a karyotype n=44 was also observed at the MI stage, along with n=43, most likely due to intraindividual chromosome fragmentation or a single-chromosome disjunction. This sample was found outside the contact zone nectaring on the flowers along a road and could be a dispersive specimen.

Distribution of species within P. aroaniensis complex in the Balkans
A distribution map based on current literature and predictions of the anomalous blues within the P. aroaniensis species complex in the Balkan peninsula is shown in Fig. 7a and a detailed map of the specimens found in Lurë region in Fig. 7b While several populations from Central and Northern Greece, as well as from other countries of the Balkan Peninsula were formerly identified as P. aroaniensis, based on Vishnevskaya et al. (2016) P. aroaniensis is only found in Southern Greece (Peloponnese). The sister taxon P. timfristos, of which it was then separated is currently only known from Mt. Timfristos and Mt. Parnassos in Central Greece. However, there is still a missing gap existing between P. lurae sp. nov and P. timfristos in Central to Northern Greece and it is not impossible that the latter can be found further northwards in suitable karst biotopes. For P. orphicus, the taxon is generally found in the northern parts of the Balkans (confirmed in Northern Greece and Bulgaria, and in North Macedonia). Here, we described also new populations of P. orphicus in Albania, and the Lurë area is the westernmost distribution of the species in the Balkans. Yet, between the Albanian populations and the former mentioned there still exist an intermediate gap of at least 500 km with insufficient distribution knowledge. Finally, in the light of the data obtained in this paper, the (possible) occurrence of P. aroaniensis s.l. in Kosovo, Northern Greece, North Macedonia needs to be further investigated as they could harbour unknown P. orphicus populations, and probably also populations of P. lurae sp. nov because biotopes with ophiolites exist there.

Morphometrics of male wings
The upperside wing colour is one of the main characteristic features of the anomalous blue butterflies. Vishnevskaya et al. (2016) used some external traits of the wing underside for differentiating between forms (types) of Balkan specimens under the "brown" complex, which we here reanalyse in view of Albanian taxa, here described for the first time, and identified under cryptic taxa P. ripartii, P. orphicus and P. lurae: 1. "Polyommatus ripartii type": hindwing underside with well-developed white streak, spots are small or medium-sized, marginal marking is reduced. According to Vishnevskaya et al. (2016) this type is found in different species including P. admetus, P. timfristos, P. orphicus, and P. ripartii pelopi (plates of male specimens shown in Vishnevskaya et al. (2016), which we confirm for the latter two species analysed. This type was never found in P. lurae specimens (data not shown).
2. "Polyommatus aroaniensis type": the white streak on the hindwing underside demon strates different level of reduction. This type is found in P. aroaniensis s.s., P. lurae sp. nov (Fig. 8m). P. timfristos, P. orphicus orphicus/eleniae and Albanian P. orphicus (Fig. 8h). It is also found in the population of P. ripartii from the Crimea (Vila et al. 2010) and Croatia (Lovrenčić et al. 2016), but according Vishnevskaya et al. (2016) very rare in the Balkan peninsula while we also found some ripartii specimens with great reduction of the white stripe (example in Fig. 8o).
3. Polyommatus orphicus type: forewing underside with clear white postdiscal streak between discal spot and submarginal marking, white streak on hindwing under side is prominent, often with an additional small white streak. According to Vishnevskaya et al. (2016) this type is common in P. orphicus orphicus while mentioning that its most characteristic feature (the white postdiscal streak between discal spot and submarginal marking on the forewing underside) can be found in other species, e.g. P. aroaniensis. Only about one third of the Albanian specimens of orphicus showed the additional streak (data not shown), while it was never present in specimens of P. lurae.
4. Polyommatus lurae sp. nov type: forewing underside with no white postdiscal streak between discal spot and submarginal marking, white streak on hindwing under side is minimal and mostly completely lacking or invisible (Fig. 8b). This type is also found in a minor amount of Albanian P. orphicus (Fig. 8h). Based on pictures of the type series, also some P. aroaniensis specimens from Greece harbour this trait (Brown 1976).
We also analysed light reflection of male wings based on standardized colour measurements (RGB and HSV values) for 27 specimens. We focused on the species P. orphicus and P. lurae, and also included the few potential hybrid specimens (based on the atypical combination of mitochondrial haplogroup and karyotype results and always collected at the contact zone in the Lurë region. NMDS plots (Fig. 9) showed that wing reflection measurements matched significantly the phylogenetic clades aroa3 and orph2 (Df= 2, F=4.11, P=0.030). Specimens identified as P. orphicus generally showed a measurable reflectance on the postdiscal band of forewing (and hindwing) (Fig. 8h), while this trait is absent in P. lurae sp. nov (Fig. 8c). Interestingly, the two potential hybrids analysed showed an intermediate position in wing colour reflectance, falling in the overlapping area of the two species, as visualized by blue dots in Fig. 9. The intermediate reflectance is also noticeable on the putative hybrid specimen depicted in Fig. 8f.  Figure 7. a distribution map based on literature and predictions of the anomalous blues within the P. aroaniensis species complex in the Balkan peninsula; colours correspond to the species in the legend; box is indicating the Lurë region in Albania b detailed map of the Lurë region with observations of P. lurae sp. nov (black dots) on dark ophiolites, P. orphicus (brown dots) and putative hybrids (blue dots) in the contact zone (blue dots) and a dispersive specimen outside suitable biotope (paler blue dot).
Next to this, a link with the soil substrate was tested and statistical analysis revealed that both species P. lurae and P. orphicus could significantly be linked with their locations harbouring typical soil substrates, i.e. dark ophiolitic versus light karts soils, respectively (Df= 3, F=4.39, P=0.014).

Taxonomy
The results showed a consensus between morphometrics, mitochondrial DNA and karyotype, in delineating three clades under the P. aroaniensis species complex. Two of them are generally accepted as species: P. aroaniensis and P. timfristos (Wiemers et al. 2018). Given that the genetic, morphological and karyotypic differentiation of the third clade is comparable to that between the other two species, we describe the Albanian population as a new species belonging to the P. aroaniensis s.l. species complex. Description. Typical dark ground colour of both veins and intervein space of dorsal wing sides. A character that appears useful for separation of P. orphicus and P. lurae sp. nov is the brighter yellow-greenish reflection of the former which is generally lacking in the newly described taxon. However, worn individuals of the two taxa may be indistinguishable externally and also from P. ripartii, which is found sympatrically in all locations studied. While Misja (2005) reports P. admetus from the same Lurë region, we never found P. admetus in sympatry with the new taxon in all Lurë locations surveyed. While the latter observation may be based on a wrong identification of the newly described taxon, also lacking white stripes on the hindwing underside, P. admetus is easily separated from P. lurae, especially because of the strongly marked underside in P. admetus, with a double row of small dots in the submarginal zone of underside wings, which has never been observed nor reported in literature in the taxa of the P. aroaniensis species complex, including P. lurae.
COI barcode sequence of the holotype. 657 base pairs: AACATTATATTT- Description. Males. (Fig. 8c, l, m). Forewing length 15.8-17.9 mm. Upperside: ground colour completely dark chocolate brown. Discoidal, submarginal and antemarginal markings absent on both fore-and hindwings. Veins poorly contrasting. Forewings with a developed sex brand and dark scale tuft. Fringe grayish brown. Underside: ground colour yellow-brown with ochreous to reddish coffee-milk tint. Minimal greenish blue basal suffusion. One basal black spot is present only on hindwings. Discoidal black spot is present on the forewings, but can be slightly seen on the hindwings (absent or vestigial). Postdiscal black ocelli most prominent on forewings; when present encircled by a whitish border. Postdiscal black ocelli on the hindwing small and sometimes lacking. Submarginal and antemarginal mark ing is absent on the forewings, and absent or vestigial on the hindwings. White streak on hindwings generally absent or very faint. Only rarely, the white streak is vestigial; no single specimen was observed with an additional short streak between postdiscal and submarginal areas of the wing, straight under the main white streak. Fringe brown, slightly darker than the underside ground colour.
Male genitalia. The valva of the male genitalia of P. lurae sp. nov is depicted in Fig.  10. Male valves have a structure typical for other species of the subge nus Agrodiaetus (Coutsis (1986), Coutsis, pers. comm.). According to Kolev (2005) who studied the morphometry of the male genitalia of P. orphicus no overlap with P. ripartii was observed. As male genitalia within the P. aroaniensis species group do not significantly differ from each other, those from P. lurae may follow the same trend, but no additional analyses nor measurements have been performed.
Females. Forewing length 15.8-17.5 mm. Upperside: ground colour as in males, but lighter dark brown and without sex brand and scaletuft. Fringe greyish brown. Underside: ground colour and general design as in males but fringes lighter-coloured. Greenish blue basal suffusion almost invisible. White streak on hindwing underside mostly absent (Fig. 7j, k). If present, it demonstrates a variable level of reduction.
Life history. Polyommatus lurae inhabits xerothermic and xeromontane ophiolitic habitats. While in some of the localities the soil can be mixed with a minor degree of a calcareous component, P. lurae was never found at pure calcareous biotopes. Indeed, at such localities only P. orphicus was found, together with P. ripartii, which is in agreement with the original description of these species (Kolev, 2005). The vegetation of the type locality is sparse and dominated by low-growing grasses and flowering plants identified as Artemisia alba Turra and Satureja montana Linnaeus. Besides, other xerophilous species were observed, including scattered Juniperus bushes and low Pinus nigra trees (Fig. 1d). In all known localities P. lurae is syntopic with P. ripartii, a species widespread in the Balkans, although especially abundant in calcareous habitats (pers. obs. L. Parmentier).
Distribution and biotope. The three known localities of P. lurae (including the type locality) are situated in the Lurë region, in the vicinity of the National Park (Parku Kombëtar Lurë-Mali i Dejës), North of the village Cidhën, along a North-Southern orientated mountain ridge and gorge at altitudes between 950 and 1.600 m (Fig. 9a). The habitats are all situated within ophiolitic soil substrates (in some localities these substrates are slightly intermixed with a minor amount of lighter karst substrate), which are not rare in some parts of Albania. In these typical ophiolitic soil substrates the presumed host plants of the genus Onobrychis were observed (Fig. 1e, f ). However, there are as yet no observations regarding the first stages of this taxon and the larval host plant is unconfirmed.
The aforementioned ophiolitic substrates can be found in a discontinuous range from Southern Albania (Provinces Korcë, Qukës) up to the Northern part of the country (provinces Dibër, Kukës). Within Europe these rather rare substrates are present mostly Figure 9. NMDS plot representing the morphometric analysis of dorsal male wings reflectance measurements of P. orphicus (orange dots) and P. lurae sp. nov (black dots). Stress = 0.02; specimens indicated in blue are potential P. lurae × P. orphicus hybrids collected in the contact zone and showed intermediate wing reflectance. Different symbols represent collection sites of three studied habitat with different substrate types (ophiolitic, karst, mixed). Ellipses represent 95% confidence intervals of specimen groups 'orphicus', 'lurae' and 'hybrid'. in Albania, while neighbouring countries of North Macedonia and Kosovo contain them to a minor degree. Thus, it is not impossible that the species is also present in other ophiolitic habitats where the presumed host plant is growing. Collection material from another locality in Voskopojë (Korçë prov.), situated more South, also harbouring typical dark ophiolitic soils was studied. In this locality, a single specimen (RVcoll14B767) genetically attributable to P. lurae was found by Sylvain Cuvelier and Morten Mølgaard, but it is not included in the type series because of the lack of karyological data and morphometrics. Additional specimens from this locality could not be found even after thorough explorations in 2018 and 2022, while only P. ripartii could be confirmed.
Differential diagnosis. From nominotypical P. orphicus the new taxon is generally distinguished by the strong reduction of a white postdiscal streak on the forewing underside, a darker colour of the upperside and underside wing, lack of wing reflectance, and less contrasting veins on the upperside. Its karyotype is different by at least one fixed chromosome fission (n=41-42) and its COI barcode. From P. aroaniensis, which is the most similar taxon externally, fresh individuals of the new taxon are distinguished by the constant presence of a typical dark ground colour of both veins and intervein space of dorsal wing sides and a generally darker colour of the upperside and underside wing (while in aroaniensis a warm reddish brown colour is typical). However, worn individuals may be indistinguishable externally, while they still can be identified by karyotype (n=48) and by the COI barcode. In the case of P. lurae, its dark habitus is linked to its typical environment with dark ophiolites, while the taxa P. orphicus and P. aroaniensis are generally found in biotopes with paler karst soil substrate. From the sympatric and syntopic P. ripartii, the new taxon is more easily distinguished by the absence of a white postdiscal streak on the forewing underside and, on average, a more reduced appearance of postdiscal spots, and on the upperside the veins are less pronounced and of a similar tone than the paler ground colour. This may be useful for discriminating even slightly worn individuals of the two taxa, while worn individuals are mostly indistinguishable externally. Yet, its karyotype (n=90) and COI barcode are strongly different. P. admetus has not been observed on the same biotopes and thus the new taxon could be separated geographically. Besides, P. admetus has a very distinct appearance by especially its strongly marked underside (with a double row of small dots on the marginal to submarginal zone of the underside hindwings, a trait that is lacking in the aforementioned species. Etymology. Derivatio nominis. The adjective lurae has two meanings: "ascribed to Lurë" and "surviving attacks of congeners". First, the species name is deducted from the Albanian "Lurë region, where the type locality lies, and referring to the old village Lurë e Vjetër situated in central-Eastern Albania (Dibër province). The name alludes to the fascinating history of the old Lurë village: during the Ottoman war, the village was asked 300 women by the enemies. Armed men, disguised with the duvak, the traditional red bridal veil, were sent instead on horseback to the Ottoman camp. As a result, the Ottomans were taken by surprise and the Lura tribe eventually won the battle. Also, this second meaning seems adequate for the taxon lurae: this species likely experienced periods of close contact with congener species more largely distributed in the Balkans, as is the case at present, but nevertheless has been able to avoid complete admixture and still survives in its unique ophiolitic biotope.

Colour morphometrics, a new method for identification of cryptic Agrodiaetus taxa
The use of standardized light reflectance measurements to discriminate between species is a recent method used for identification (Bálint et al. 2010;Bálint et al. 2012;Wasik et al. 2014). Other morphological traits such as underside markings and prominence of the white stripe are useful, but they are not discriminative enough to unambiguously distinguish between the taxa here studied. Most of P. lurae specimens have a dark ground colour in the underside of the wings, and no or a very faint white stripe. However, one specimen identified as lurae by mitochondrial DNA, karyotype and morphometrics (colour reflectance of dorsal wing colour was typical) showed a more pronounced white stripe. Here we only used RGB an HSV values in the analysis, but more sophisticated measurements to generate full reflectance spectra and SEM graphs may be more powerful to discriminate between these taxa. Such analyses could also shed light on the physical structures that generate the typical dark colour of P. lurae, compared to the greenish reflectance in P. orphicus specimens, as has been demonstrated in other species of Polyommatus (Bálint et al. 2012). Besides, morphometrics on preimaginal stages could also be potentially interesting. Almost no information is available on this aspect, while recent findings showed that differences in larval morphology and in larval host plant preferences may be key in resolving the taxonomy of cryptic species (Hernández-Roldán et al. 2016;Hinojosa et al. 2022).

Karyotyping and difference with related taxa
The karyotype of P. orphicus was studied previously (Kolev 2005;Vishnevskaya et al. 2016) from localities in Northern Greece and Bulgaria. Two different haploid chromosome numbers (n=41 and n=42) were found to be present in these populations. The variation in chromosome numbers in these populations was explained by polymorphism for one chromosome fusion/fission. This polymorphism resulted in three types of MI karyotype: n=41 (homozygous for chromosomal fusion/fission, one pair of fused chromosomes, 2n=82), n=42 (homozygous for chromosomal fusion/fission, two pairs of unfused chromosomes; 2n=84) and n=41 (heterozygous for chromosomal fusion/fission, 40 bivalents and one trivalent; 2n=83) (Vishnevskaya et al. 2016).
The P. orphicus karyotype found by us in Albania (n=42) fits into the previously described variability. At the same time, it can be assumed that in the Albanian population there is a tendency to fixation of the chromosome number n=42, although the studied data are still insufficient to consider this proved.
The taxon we describe as P. lurae sp. nov also exhibits intrapopulation variability in chromosome numbers (n=42, n=43; estimated diploid numbers are 2n=84, 2n=85, 2n=86) due to polymorphism for one chromosome fusion/fission, but most likely in another chromosome pair. Thus, despite chromosome polymorphism in each of the taxa P. orphicus and P. lurae, they have, most likely, a fixed difference in one chromosome pair (Fig. 11).
In the contact zone in Albania, both mitochondrial haplogroups orph2 and aroa3 occur together. It can be assumed that they arose as a result of hybridization, which is confirmed by the intermediate nature of the colour of the wings.
In the case of hybridization, if contacting taxa have postzygotic reproductive isolation, then hybrid individuals should represent only F1 hybrids (further hybridization is impossible due to sterility). If the hybrids are fertile, then a mixture of hybrids of different generations and the results of backcrosses should be observed.
The reconstruction of karyotypes of pure forms of P. orphicus and P. lurae sp. nov and their putative hybrids is shown in Fig. 11. As follows from this scheme, F1 hybrids should all have the same number of elements visible in the first metaphase of meiosis (n=41), despite the fact that they may include 1 to 2 complex multivalents. Such a karyotype was not observed in the putative hybrid zone. From this we conclude that, likely, there is no complete reproductive isolation between P. orphicus and P. lurae, and the observed karyotypes n=43 are the result of repeated hybridization and backcrosses. Another hypothesis to explain the pattern we observe would be that P. orphicus lineage orph2 in Albania displays a karyotype n=42-43 and the specimens with n=43 are not admixed, although the fact that they were only found in the contact zone and that the two specimens measured displayed intermediate morphology favour the hybridisation hypothesis.

Taxonomic position and difference from sister taxa
The data obtained demonstrate that P. orphicus and P. lurae represent two distinct phylogenetic lineages with a parapatric distribution. Indeed, both P. orphicus and P. lurae formed a highly supported monophyletic lineage based on three phylogenetic analyses (BI of COI barcode, ML of COI+ITS2 and BI of COI+ITS2) (Fig. 2, Suppl. material 2). These two lineages are also substantially differentiated with respect to morphology (different wing reflectance), and karyotype (difference in one chromosome pair). Therefore, they can be considered species from the viewpoint of the phylogenetic species concept. These two lineages (=phylogenetic species) overlap in a small contact zone in Albania, and the combination of mtDNA, karyotype and morphological data suggest that they may hybridize and no complete barrier to reproduction exist.
Theoretically, the main lineages in the P. orphicus, P. timfristos, P. aroaniensis and P. lurae subcomplex could also be interpreted as infraspecific taxa, if the polytypic species concept is applied (Vishnevskaya et al. 2016). In this case, these taxa would be subspecies under the entire P. aroaniensis species complex. In Albania we showed that there is a contact zone between P. orphicus and P. lurae where unusual combinations of mitochondrial and karyotype, as well as intermediate morphotypes, exist. However, none of the aforementioned taxa appear to be fully sympatric in distribution and, taken together, they form a highly supported monophyletic lineage based on analysis of COI sequences (Fig. 2) and the concatenated COI+ITS2 sequences (Suppl. material 2). Under this scenario, this subspecies-complex would be considered a diverse array of allopatric populations, each of which possesses unique genetic attributes (karyotypes and molecular markers) and is distributed in a particular area within the Balkan peninsular. While one can argue that differences in chromosome numbers in the subgenus Agrodiaetus do not necessarily result in complete reproductive isolation and, at least in some particular cases, do not prevent interspecific hybridization and genetic introgression (Lukhtanov et al. 2015b), this does not necessarily mean that chromosomal rearrangements are irrelevant to the formation of genetic barriers between populations (Vishnevskaya et al. 2016). Figure 11. Scheme showing variation in number of chromosomes (lines) and visible elements (=bivalents+multivalents) in MI meiosis in P. orphicus, P. lurae and their putative F1 hybrids. a P. orphicus, homozygous for chromosomal fusion/fission, one pair of fused chromosomes, 41 visible elements b P. orphicus, heterozygous for chromosomal fusion/fission, 40 bivalents and one trivalent; 41 visible elements c P. orphicus, n=42 (homozygous for chromosomal fusion/fission, two pairs of unfused chromosomes; 42 visible elements d P. lurae, homozygous for chromosomal fusion/fission, one pair of fused chromosomes, 42 visible elements e P. lurae, heterozygous for chromosomal fusion/fission, 41 bivalents and one trivalent; 42 visible elements f P. lurae, homozygous for chromosomal fusion/fission, two pairs of unfused chromosomes; 43 visible elements g-j different variants of F1 hybrids. These variants include tri-and quadrivalents; however, the number of visible elements in MI remains 41. Chromosome changes have been shown to be important for speciation in Polyommatina butterflies (Lukhtanov et al. 2005;Kandul et al. 2007;Talavera et al. 2013a;Vishnevskaya et al. 2016) and even a weak reduction in fertility in heterozygotes for multiple chromosomal rearrangements can result in selection against them and in the formation of a boundary between chromosomally diverged homozygous populations. More detailed studies investigating lab-controlled crosses between sister taxa and the fertility of their progeny would be interesting to shed light on this topic, as has been achieved in wood white (Leptidea) butterflies (Dincă et al. 2013). Recent taxonomical publications have treated P. orphicus, P. aroaniensis and P. timfristos as species-level taxa (Eckweiler and Bozano 2016;Wiemers et al. 2018) and our study is following this rationale also for P. lurae sp. nov.
Regardless of its taxonomic status as a species or subspecies, P. lurae represents a unique entity within the genus Polyommatus that deserves additional study. A better understanding of its evolutionary history and its relationship with its unique biotope and related taxa may be helpful in understanding mechanisms of chromosomal diversification within the subgenus Agrodiaetus, and may further elucidate the biogeography of the south Balkan and Aegean regions.

Conservation of the species and habitat
The Lurë region has become a National Park (Parku Kombëtar Lurë-Mali i Dejës) since 1966 to protect its ecosystems and biodiversity. Since 2018 by encompassing the entire section of Kunora e Lurës, its name has changed to Parku Kombëtar Lurë-Mali i Dejës, spanning an expanded area of 202.42 km 2 . Despite its conservation status the area suffered massive deforestation from illegal logging and forest fires that severely affected ecosystems and it is estimated that as much as 50% of the original Lura National Park has been destroyed (Rama 2018). Moreover, it is not fully covering important biotopes such as some of the P. lurae biotopes.
Next to this, the first author noticed that sheep overgrazing is also affecting the ecosystems. As Onobrychis plants are very palatable to sheep, heavy grazing limits the growth and expansion of Onobrychis, sometimes leading to the extinction of the plant (Lafranchis et al. 2007). While traditional grazing by sheep is beneficial, and can help in keeping open clearings, uncontrolled and overgrazing can have a devastating impact on butterflies, and other insects such as bees, which has been increasingly reported in different parts of Europe (Kruess and Tscharntke 2002;Potts et al. 2009;Verbrugge et al. 2022) and such an evolution in Albania would be dramatic for its biodiversity, especially for ecosystems harbouring unique species diversity.
The future of various endemic species of Polyommatus in Europe is strongly dependent on keeping open dry clearings at montane-subalpine levels where its foodplant is growing; This is the case for P. orphicus and P. aroaniensis in Greece but even so for P. lurae in Albania.
As a distinct taxonomic entity occupying a very restricted area linked to a unique biotope in Albania the newly described species should be considered a candidate on the list of protected species in Albania and the whole of Europe by adding to the European In summary, the Lurë region harbours unique endemic flora and fauna, in addition to being home for the species here described, which is currently only found very restricted and locally. Therefore, the preservation of this habitat needs being ensured. This encompasses also control of human activities as illegal logging, burning and uncontrolled grazing by livestock, all major factors that have been identified contributing to butterfly decline in Europe (van Swaay and Warren 2006). As Albania is setting up programs to be member of be the European Union, installing adequate protection legislation for its biodiversity heritage (e.g. the EU Habitats Directive 92/43/EEC) will be needed.