The previously unknown male genitalia of Phtheochroa unionana (Kennel, 1900) are described and illustrated. The species is dimorphic: one form is white with very faint yellow scales in the fascial areas and the other is white with distinct orange fasciae. The everted vesicae of the males do not show interspecific variation but are remarkably different from those of a closely related species. Apparently, the morphology of the everted vesica is a useful tool for species recognition in this genus. The conspecificity of the two forms of P. unionana is further corroborated by evidence from COI barcodes.

The genus Phtheochroa Stephens, 1829, comprises 107 species worldwide (Gilligan et al. 2014) and 53 species with a Palaearctic distribution (Razowski 2009). No synapomorphies are known for the genus, but some groups of species demonstrate clear morphological affinities (Razowski 1991). According to Razowski (1991), P. unionana (Kennel, 1900) belongs to a group of ca. 13 species defined by a simple valva without a free termination of the sacculus. In this group several species externally show similarity with P. unionana. Their genitalia are considerably simplified in comparison to other Phtheochroa spp. and species recognition sometimes is difficult. Further studies of the genitalia morphology combined with molecular data probably will reveal other problems and unknown facts for this group.

During an expedition to Armenia in 2014 a single male of a pure white Phtheochroa was collected. The genitalia of the specimen did not fit any known species, which, combined with the forewing colour, convinced us that this was a male P. unionana. Study of additional material also collected from Armenia revealed other P. unionana specimens. Unexpectedly, the genitalia of an undetermined Phtheochroa from the same area with orange fasciae were nearly identical to those of P. unionana, indicating conspecificity of the two forms, a hypothesis supported by subsequent DNA barcoding.

Material examined: Phtheochroa unionana: Armenia: 4 ♂♂ (2 white, 2 fasciate), Tavush region, Dilijan, N40°45’, E44°51’, 1340–1450 m, 12–14.vii.2011, leg. O. Karsholt, coll. ZMUC; 1 ♂ (fasciate form), Kotayk region, Tsaghkadzor, N40°32’, E44°32’, 1870–2350 m alt., 9.–11.vii.2011, leg. O. Karsholt, coll. ZMUC; 1 ♂ (white form), Lori region, Lermontovo vill., N40°44’55”, E44°39’40”, 1860 m alt., 29.vii.2014, at light, leg. B. Zlatkov & D. Chobanov, coll. BFUS. Georgia: 1 ♀ (fasciate form), Lesser Caucasus, Kvemo Kartli region, Sakire, N41°14’13”, E44°17’02”, 1260 m alt., 28.vii.2014, leg. B. Zlatkov & D. Chobanov, coll. BFUS. Phtheochroa procerana (Lederer, 1863): Bulgaria: 1 ♂, Bulgaria, Black Sea coast, Balchik–Kavarna, N43°24’21”, E28°12’28”, 100 m alt., 21.vi.2007, leg. B. Zlatkov & S. Beshkov, coll. BFUS; 1 ♂, Bulgaria, Veliko Tarnovo region, Emen Gorge, N43°08’20”, E25°21’41”, 150 m alt., 16.vii.2011, leg. B. Zlatkov & O. Sivilov, coll. BFUS.

The moths were collected at a “light tower” with a 160 W MBFT lamp and blacklight fluorescent tube, and traps with blacklight tubes. The genitalia were dissected following Robinson (1976) with the exception of the phalli; they were processed following Zlatkov (2011). The description of the cornuti generally follows Anzaldo et al. (2014). The phalli with everted vesicae were submerged in Euparal essence and attached to a needle with a diameter of 0.15 mm inserted through the entering excavation of the ductus ejaculatorius into the phallus. A compound microscope with attached camera lucida was used for the line drawings.

DNA barcode sequences of the mitochondrial COI gene (cytochrome c oxidase 1) were obtained from three specimens of P. unionana and an other three of P. procerana. DNA samples from dried legs were prepared according to prescribed standards using a standard high-throughput protocol (deWaard et al. 2008). Samples were processed at the Canadian Centre for DNA Barcoding (CCDB, Biodiversity Institute of Ontario, University of Guelph) to obtain DNA barcodes (Ratnasingham and Hebert 2007). DNA sequencing resulted in barcode sequence of 658 bp and a sequence of 604 bp for P. unionana and two sequences of 623 bp and 627 bp for P. procerana; sequencing of a third specimen of each species failed. Details of successfully sequenced voucher specimens including complete voucher data and images can be accessed in the Barcode of Life Data Systems (Ratnasingham and Hebert 2007). Degrees of intra- and interspecific variation in the DNA barcode fragments were calculated using Kimura 2 parameter (K2P) model of nucleotide substitution using analytical tools in BOLD systems v3.0 (BOLD 2015). A neighbour-joining tree of DNA barcode data of European taxa was constructed using MEGA 6 (Tamura et al. 2013) under the K2P model for nucleotide substitutions.

P. unionana was described from two male specimens from the Caucasus (without details of the locality), both with white forewings with barely discernible yellow fasciae (Kennel 1900). The specimens were later lost and the male genitalia remained unstudied (Razowski 1970). Female specimens collected subsequently were assigned to this taxon based entirely on the appearance of the forewing, but no further males were reported prior to the present study. Thus for the first time we are able to describe the genitalia morphology of the male. Razowski (1970, 2009) placed P. unionana near P. procerana based on the external appearance and female genitalia. With regards to the male genitalia, this position seems correct. The everted vesicae of both species demonstrate similarity though there are certain differences in the details: the same components are present in both species but their position and shape is different. It should be emphasised that the shape of the vesica, even the relative position of the diverticula and cornuti, are relatively constant at specific level though some small differences can be detected (Fig. 3a–d). A remarkable character is present in both species: a sclerotized semi-cylindrical plate around the gonopore. Though its function is not known, it suspends eversion of the ductus ejaculatorius. The peculiar ventral “diverticulum” with a conical shape and uneven surface does not correspond to any structure of the tortricid vesica known to us. It may be even non-eversible in the living moth. The female genitalia agree well with the illustrations by Razowski (1970, 2009) but the latter probably were drawn from the dorsal side instead of the ventral. This assumption is supported by the genitalia of the related species: the lateral sclerite connecting the corpus bursae with the ductus bursae is located at the left side in P. procerana (specimens studied by us), P. durbonana and P. purissima (judging from illustrations in literature).

Though a special case, comparison of the vesicae of two related species proves the taxonomical significance of this structure in the Cochylini, at least in Phtheochroa. Detailed comparison is achievable only after complete inflation of the vesicae, then numerous characters became visible and can be used for morphological analysis. The taxonomic significance of this character was tested by comparison with the closely related species P. procerana from which fresh material was available (Fig. 1c). As well as the morphology, the barcode divergence of 7.4% between P. unionana and P. procerana also clearly supports two separate species.

The conspecifity of two strikingly different forms in P. unionana is less supported by molecular data because the intraspecific distance between the forms is considerable at 2.38% but this may be due to geographic variation as one sequenced specimen originated from Armenia and the other from Georgia. Such divergence rates have been attributed either to intraspecific variation or interspecific divergence, varying from case to case (Huemer et al. 2014 and references within). However, the full conformity of genitalia morphology in both forms and their co-occurrence support a single species hypothesis. Male genitalia structures including the everted vesicae of the fasciate and white syntopic specimens appeared identical. Furthermore, the genitalia of a female specimen with fasciate forewings appeared identical to the available illustrations of the female genitalia of P. unionana. As a result we conclude that P. unionana has two forms that differ by forewing pattern. Extreme variation in the wing pattern is common in the subfamily Tortricinae, especially in the genus Acleris, and in many Cochylini, e.g. Cochylimorpha and Aethes.

We are most grateful to Paul Hebert and his team at the Canadian Centre for DNA Barcoding (Guelph, Canada) whose sequencing work was enabled by funding from the Government of Canada to Genome Canada through the Ontario Genomics Institute. We are also grateful to the Ontario Ministry of Research and Innovation and to NSERC for their support of the BOLD informatics platform. Special thanks go to Ole Karsholt (Copenhagen, Denmark), who provided most of the P. unionana specimens. Robert J. Heckford (Plymouth, U.K.) improved an earlier draft of the paper linguistically and provided valuable comments. Thanks to Joaquín Baixeras (Valencia, Spain) and John Brown (Washington DC, USA) for critical review of the manuscript.