The Eastern Palaearctic Cosmopterix feminella Sinev, 1988, introduced in Italy: taxonomy, biology and a new synonymy (Lepidoptera, Cosmopterigidae)

Cosmopterix feminella Sinev, 1988, previously known from the East Palearctic, Primorskiy Territory in Russia and Japan has been collected at light in Europe. In northern Italy 58 females were collected in two localities in the province Asti, two in Alessandria, three in Udine, and in one locality in Pordenone. Cosmopterix feminella is most likely parthenogenetic as only females are known. The caterpillars are leafminers on grasses. The species is redescribed and illustrated. DNA barcodes are provided and compared with other European species. Cosmopterix feminae Kuroko, 2015 is synonymised with C. feminella. The species was probably accidentally introduced into Italy.


Introduction
The genus Cosmopterix Hübner, 1825 (Gelechioidea: Cosmopterigidae) comprises small, sometimes very small, moths with a striking and often beautiful forewing pattern. The majority of the species can easily be recognized by the broad yellow to orange fascia on the forewing. This fascia is often bounded by tubercular golden or silver metallic fasciae or spots. The wings are very narrow and acutely pointed. The larvae are leafminers, usually oligophagous or monophagous and have been found in the following plant families: predominantly on Cyperaceae, Fabaceae and Poaceae, fewer species on Asteraceae, Cannabaceae, Convolvulaceae, Dipterocarpaceae, Hydrangeaceae, Lamiaceae, Theaceae and Urticaceae (Robinson et al. 2010;Kuroko 2015;Koster 2016). The genus occurs on all continents except Antarctica. The majority of the species have been found in tropical areas, while in temperate climates the number of species declines rapidly towards the poles. Europe (including Macaronesia) being well explored, has only 13 species recorded (Koster and Sinev 2003;Huemer and Koster 2006). Sinev (1997) mentioned 40 species in his review of the genus from the whole Palaearctic region, of which 12 species were also known from Europe.
Recently the number of species of the Palaearctic region further increased through the revision of the fauna of Japan by Kuroko (2015), who listed 49 species for this country, of which 25 were described as new. On the basis of these data, Sinev (2002) and Koster (2016), the total number of species worldwide is currently 344.
In Italy in June 2015 eight specimens of an unknown Cosmopterix-species were collected in a light trap in Cascine Bet, near Mombarone (Asti, Valmanera biotope) in the region Piemonte by O. Maioglio. An additional specimen of this unknown species was collected in a light trap by C. Cabella in September 2016 in Pertuso Le Strette, Cantalupo, also in the region Piemonte, roughly about 90 km east of the first locality. This material was brought to the attention of Giorgio Baldizzone who sent the material to the first author. Almost simultaneously, between August 2015 and September 2016, Helmut Deutsch collected six females in the region Friuli Venezia Giulia in three localities in the province Udine at light. It was tentatively identified as C. crassicervicella Chrétien, 1896 and sequenced in the framework of the barcoding campaign "Lepidoptera of the Alps". More material from the same and more localities became available in 2017 and 2018.
The collected specimens resembled Cosmopterix attenuatella (Walker, 1864) and C. crassicervicella, but lacked the orange brown or orange yellow markings on the dorsal side of the abdomen. After the examination of the female genitalia the species could be determined with certainty as C. feminella Sinev, 1988. DNA barcodes showed that the species from Piemonte and Friuli Venezia Giulia were conspecific.
Here we report the species new for Italy and Europe, and redescribe it in the format of the Microlepidoptera of Europe book series (Koster and Sinev 2003) and provide DNA barcodes.

Material and methods
Material. The material from the region Piemonte has been collected with Actinic light and a light trap. Similarly specimens from Friuli Venezia Giulia were all collected at light. Material is listed in Table 1, and more detailed in Suppl. material 1.
Morphology. Genitalia were dissected following the methodology presented by Brown (1997) and Robinson (1976). Chlorazol black was used to stain the unsclerotized parts of the female genitalia.
The morphological terminology follows Koster and Sinev (2003) and Koster (2016). In Cosmopterix the sterigma, formed from lamella antevaginalis and lamella postvaginalis, in the female genitalia is important for identification. It is often poorly sclerotized and can easily be distorted by too much pressure on the coverslip of the genitalia slide. Therefore the sterigma has been depicted here in both ventral and lateral position.
Illustrations. The photograph of the adult was taken with a Zeiss AxioCam digital camera attached to a motorized Zeiss SteREO Discovery.V12, using Carl Zeiss AxioVision software. The female genitalia were depicted in line drawings in ventral position, and the drawing of the enlarged sterigma are given separately; these drawings have been reduced to 70% of the original drawing size. Drawings of the genitalia were made with a compound microscope using the camera lucida method. For this purpose a strong light source (slide projector) was used for the illumination on the mirror of the microscope. A prism was placed on top of the microscope eyepiece to bend the projection 90° and project the subject on drawing paper. All outlines were drawn by pencil and later set in Indian ink. Scale bars alongside the drawing show the size of the genitalia in 0.1 mm.
The map (Fig. 5) was prepared with ArcGIS software.
DNA extraction and sequencing. DNA barcodes of two specimens of C. feminella (RMNH. INS.15509 and RMNH.INS.15510) were derived from extracts taken from abdomens following the procedures outlined by van Nieukerken et al. (2012a) and Doorenweerd et al. (2015), the abdomen and genitalia were taken from the lysis and prepared by JCK in the normal way. Legs of four other specimens were sent for analysis to the CCDB (Canadian Centre for DNA Barcoding, University of Guelph, Canada) where DNA extraction, PCR amplification, and sequencing were performed following standard high-throughput protocols (deWaard et al. 2008). The resulting COI sequences are combined with those from a representative set of other European Cosmopterix species from our own projects, or being publicly available, in the public BOLD dataset "Cosmopterix feminella in Europe [DS-COSMFEM]" [doi: 10.5883/DS-COSMFEM] and GenBank (Suppl. material 1). Unfortunately we have no barcode of the closely related C. attenuatella available from European samples, in BOLD there are only two incomplete barcodes from Galapagos and North America respectively, not yet in the public domain. The Neighbour Joining tree was prepared with the tree building tools on BOLD (Ratnasingham and Hebert 2007) and edited with Adobe Illustrator.

Diagnosis.
In Europe C. feminella resembles C. crassicervicella and C. attenuatella. The latter species occurs in Europe only in the Macaronesian Archipelago. It differs from both by the narrower forewings, which are ten times as long as wide compared to eight times in C. feminella and C. crassicervicella. In the forewing C. feminella can be distinguished from both other species by the absence of a white costal line in the basal area and by the absence of the apical protrusion of the orange-yellow fascia, and in the abdomen by the uniform ochreous-brown dorsal coloration. In C. attenuatella the abdomen is dorsally more or less spotted orange-brown and in C. crassicervicella the abdominal segments two to six are dorsally orange-yellow. The female genitalia of C. feminella can be recognised by the large central hump on the posterior edge of sternite VII, by the distally hood-shaped sterigma and by the small but prominent crescent-shaped signa.
Redescription. Female (Fig. 1). Forewing length 3.9-4.1 mm. Head: frons shining grey with greenish and reddish reflections; vertex and neck tufts shining dark brown with reddish gloss, medially and laterally narrowly lined white; collar shining dark brown with reddish gloss; labial palpus, first segment very short, grey, second segment four-fifths length of third, both segments dark brown, dorsally and ventrally lined white. Antenna: scape dorsally shining brown with reddish gloss and white anterior line, ventrally shining white; flagellum shining dark brown with short white line from base, changing in annulated white line to one-half, followed towards apex by four to five dark brown, one white, one dark brown, one white, four dark brown and three to four white segments at apex, sometimes apical segment grey to dark brown. Thorax and tegulae shining dark brown with reddish gloss, thorax with white medial line, often widening posteriorly, tegulae narrowly lined white inwardly. Legs: shining dark brown with reddish gloss; foreleg with white line on tibia and tarsal segments one to three and five; tibia of midleg with white oblique basal and medial lines and white apical ring, tarsal segments one and two with greyish white apical rings, tarsal segment five white; tibia of hindleg as midleg, tarsal segments one to four with narrow white apical rings, segment five dorsally white; spurs shining white, ventrally more grey. Forewing shining dark brown with reddish gloss; four white lines in basal area, subcostal from base to one-fourth and bending inwardly in distal half, short medial from one-fifth to just beyond subcostal, subdorsal underneath medial, about as long as medial or reaching slightly beyond it, very narrow dorsal from base to one-fifth, sometimes hardly visible; bright orange-yellow transverse fascia beyond mid length, narrowing in dorsal half; bordered at inner edge by broad tubercular golden metallic fascia, narrowing towards costa and with small subcostal patch of black scales on outside; bordered on outer edge by two tubercular golden metallic spots at costa and dorsum, dorsal about twice as large and nearer to base, costal and dorsal spots inwardly edged by few blackish scales; rather broad greyish white costal streak, inwardly edged dark grey; longitudinal silvery white spot in middle of apical area and shining white line in apical cilia; cilia dark brown around apex, paler towards dorsum. Hindwing shining brownish grey with reddish and greenish gloss; cilia dark greyish brown. Underside: forewing shining greyish brown, white costal and apical line distinctly visible; hindwing shining grey. Abdomen dorsally shining ochreous-brown with reddish golden gloss, laterally shining grey, ventrally shining grey, segments broadly banded white posteriorly; anal tuft dorsally ochreous-brown, ventrally blackish brown.
Male genitalia. Male unknown. Female genitalia (Fig. 2a-c). Posterior edge of sternite VII deeply concave with very large convex hump in centre. Sterigma elongated, distally hood-shaped, ostium bursae rounded, basally V-shaped with rounded labiate extension. Ductus bursae narrow, slightly shorter to about as long as corpus bursae. Ductus seminalis from upper part of corpus bursae. Corpus bursae elongate, two small, crescent-shaped signa. Biology. The biology has been described by Kuroko (2015) (as C. feminae), from which we cite the following. Host plants: Digitaria ciliaris (Retz.) Koeler and D. violascens Link (Poaceae). Larva (last instar): head blackish brown, body cylindrical, pale yellow, prothoracic shield and anal plate black, prothoracic legs blackish brown. The larva mines the leaves and makes an irregular blotch mine from near the base towards the apex of the leaf. The frass is ejected from a hole at the beginning of the mine. The larva changes leaves to make more than one mine. Pupation takes place outside the mine in a spindle-shaped whitish brown cocoon in a folded space on a leaf. In Japan (Kyushu: Mt. Hikosan) up to four generations per year occur. The larva of the last generation hibernates full grown and pupates in the middle of May the following year. The adults are on the wing in late May, in late July, from mid-August to mid-September and in late October. The European specimens of C. feminella were collected in the months June, August, September and October.
DNA barcodes. We obtained completely identical DNA barcodes from six specimens of C. feminella from Italy, all belonging to the new Barcode Index Number BOLD:ADG7284, and with a distance of 5.8% to the nearest neighbour, an unidentified Cosmopterix species from Madagascar with BIN BOLD: ACT2622 (specimen BIOUG18998-F03). This specimen was collected with a Malaise trap and therefore its external morphology is poorly preserved (Lopez-Vaamonde et al. 2018). Unfortunately no barcodes are yet known from the Asian populations. The distance to barcodes of European species is larger, the smallest being 6.1% to C. scribaiella (Zeller, 1850) (Fig. 6). Remarks. Cosmopterix feminella has been described on the basis of four females, caught between 25 July and 15 August in Primorskiy Kray, in the south-east of the Russian Far East.
In his study on the genus Cosmopterix of Japan, Kuroko (2015) described a new species, C. feminae, based on reared females only, and compared it with C. feminella from Russia, of which  also only females were known. His conclusion was that the Japanese species shows sufficient differences to describe it as a new species, based on the following features: the apical segment of the antenna is black in C. feminae; but white in C. feminella; in the forewing the subcostal line starts from the base of the wing in C. feminae, which is not the case in C. feminella; in the female genitalia the sterigma (lamella antevaginalis) lacks the tongue-shaped extension at the top as seen in C. feminella; the ductus bursae is about half the length of the corpus bursae in C. feminae instead of nearly the same length in C. feminella. Cosmopterix feminae is here synonymized with C. feminella for the following reasons: In the European series from Italy the apical segment of the antenna is white in 12 specimens, but it is dark grey in two specimens and black in one specimen, thus this character is apparently not constant. According to the original description of C. feminella the subcostal line also starts from the base of the forewing. Sinev (1988) provided a drawing of the forewing in which the subcostal line does not reach the base of the wing. Actually it does start from the wing base, but the curve of the wing margin hides this when seen from above, as in the figure. Probably this has convinced Kuroko that this line starts beyond the base of the wing. As Sinev (1988) mentioned, the tongue-shaped extension at the top of the sterigma (lamella antevaginalis), belongs to the hood-shape form of the sterigma. When the sterigma is embedded on the genitalia slide in ventral position, the so-called tongue-shaped extension can be caused by pressure on the cover slip. Kuroko (2018) described the length of the ductus bursae as about half the length of the corpus bursae, however, in his drawing (Plate 21, fig. H) the length of the ductus bursae is only slightly shorter than the length of the corpus bursae and equal in length to the length of the ductus bursae compared to the corpus bursae in the drawing by Sinev (1988: fig. 2, below in centre). The above differences are thus considered insufficient to regard C. feminae as a separate species.

Discussion
The type series of both C. feminella from Russia and C. feminae from Japan, including reared material, consist of females only. On this basis Kuroko (2015) suggested the possibility of parthenogenesis. Also all specimens found in Italy are females. All this taken together suggests that it is very likely that we are dealing here with a parthenogenetic species, although confirmation from e.g. rearing experiments is required for more certainty.
The finding of so many specimens of a species otherwise only known from the eastern Palearctic, thousands of kilometres away, is probably best explained by accidental introduction. How this happened is difficult to trace, but it is likely aided by human transport, either imported with plant material from Asia, introduced unintentionally as a contaminant with foreign grass seed, or even as adults inside some means of transport. One piece of grass with a pupa might have been sufficient, as for introduction of a parthenogenetic species just one specimen could be enough. Globalization has led to an increasing number of introduced alien insects from many parts of the world (Roques et al. 2016). From all alien Lepidoptera known from Europe around 2009, most species originated from Asia (Lopez-Vaamonde et al. 2010), but for leafminers this number is relatively low, with best known example Phyllonorycter issikii (Kumata, 1963) on Tilia species, whereas most alien species originate from North America (Kirichenko et al. 2018). In several cases, Italy was the first country where these leafmining aliens were recorded, as in Parectopa robiniella Clemens, 1863, Antispila oinophylla van Nieukerken & Wagner, 2012 and Coptodisca lucifluella (Clemens, 1860) (Hellrigl 2006;Nieukerken et al. 2012b;Bernardo et al. 2015). Not all introduced species are able to establish populations, as was probably the case in the single record of the Eastern Asian fern feeder, found in Austria: Cuprina fuscella Sinev, 1988 (Stathmopodidae) (Wieser 2004). Possibly, Cosmopterix feminella is not the first alien Cosmopterix species in Europe, the widespread C. pulchrimella Chambers, 1875, feeding on Parietaria species (Urticaceae) is suspected to have been introduced into Europe from North America (Koster 2010). It was first found in Croatia in the 1930s (Koster 2010), and it is still spreading northward (Parsons and Sterling 2004; Ellis 2019).
For establishing populations, it is important that the species finds suitable hostplants. Both cited hostplants are known as introduced plants in Europe. The distribution of Digitaria ciliaris has been discussed in detail by Wilhalm (2010). The species has been described from China, but is nowadays widespread in Europe, especially in the south, but becomes scarcer towards the north. The species is not easy to separate from the native European D. sanguinalis (L.) Scop. Digitaria ciliaris can be found in northern Europe in harbour areas, dumps, waste places, gardens, roadsides, etc. In southern Europe it is also accompanying grass in lawns and plantations. Digitaria violascens is native in tropical Asia and probably also in tropical America from where the type was described (Verloove 2008). In Europe it has been found in France and Italy in different habitats. It is possibly introduced unintentionally as a contaminant with foreign grass seed. However, it is quite possible that C. feminella also feeds on other grass species, and the genus Digitaria has three native species in south-western Europe (Verloove 2008), giving ample opportunities for the moth to establish populations. A focussed search for leafmines is the next step, and it remains important to follow the further establishment and invasion of this species in Europe. As long as the species is attacking only the currently known food plants, which have also been introduced into Europe, it will probably not have any environmental or economic impact.