Tischeria caucasica is very similar to T. ekebladella (Fig. 1A–C). However, as stated by Klasiński et al. (2020), males can easily be distinguished from T. ekebladella and T. ekebladoides Puplesis & Diskus, 2003 by the unique, very long and distally bifurcate appendages of the juxta (Fig. 2A‒H).

In the dissected males of T. caucasica from Iran, there is some variation in the shape and length of the horn-like appendages of the juxta. Based on the results of this study, the apex of the longer pair of juxtal appendages is not always bifurcated. In some of the examined males, both appendages have bifurcated tips (Fig. 2A, F); while in some others the shorter one has simple and longer one has bifurcated tip (Fig. 2B, D, E). This has also been illustrated by Klasiński et al. (2020: 15, figs 15, 21–23), although without any explanation of the variation. Moreover, as we revealed in this study, the longest appendage of the juxta if bifurcated apically, the two branches are either assymetrical or symmetrical (Fig. 2A, B, D, ­F).

According to Klasiński et al. (2020), the two very long horn-like processes of juxta are bent basally; however, as we discovered in this study, the shorter one is always bent at its distal two-thirds, but the longer process is usually curved internally (Fig. 2A–F). Additionally, the anterior tip of the triangular ventral plate of vinculum varies from narrow (Fig. 2A) to relatively wide (Fig. 2B, D).

In the genitalia of examined females, as described by Klasiński et al. (2020), the antrum has a weakly sclerotized anterior margin with a unique membranous part, and the shortest prela has a wide base (Fig. 3A–D); however, the corpus bursae is not always heavily folded longitudinally (Fig. 3A).

No external difference between T. caucasica and its closely related species was reported by Klasiński et al. (2020). Based on the results of this study, T. caucasica is close to both T. ekebladella and T. ekebladoides in wing pattern and size. However, there are a few differences between T. caucasia and T. ekebladella (Fig. 1A–C, E–G): 1) T. caucasica has a partly less pointed hindwing compared with T. ekebladella; 2) Fringes of the hindwing in T. caucasica are slightly paler than in T. ekebladella; 3) Ventro-apical fringes of forewing in T. ekebladella are slightly darker than in T. caucasica and 4) On the underside of the forewing, the costal area at the apex is slightly paler in T. caucasica compared with that of T. ekebladella.

Tischeria caucasica differs from T. ekebladoides in the following external features (Fig. 1A, B, D): 1) T. caucasica has a wider and less pointed hindwing compared with T. ekebladoides; 2) The ground colour of the body and forewing in T. caucasica are darker than in T. ekebladoides; 3) In T. ekebladoides the frontal tuft and collar area have scattered brown scales and 4) The forewing of T. ekebladoides is slightly narrower than that of T. caucasica.

Based on the results of this study, the female of Iranian T. caucasica population has a slightly longer forewing than the male (Fig. 1A, B), with no significant difference between the two generations. Forewing length in the first generation was 3.5–4.6 mm (x = 4.11 mm ± 0.26, n = 44) in the male and 3.6–4.9 mm (x = 4.37 mm ± 0.26, n = 49) in the female. In the second generation it was 4.0–4.4 mm (x = 4.12 mm ± 0.16, n = 8) in the male, and 4.0–4.7 mm (x = 4.31 mm ± 0.25, n = 7) in the female.

Leafmines of T. caucasica cannot be separated from those of T. ekebladella. Moreover, the larvae of these two species are indistinguishable externally.

A Neighbor Joining tree is provided for barcodes of T. caucasica and T. ekebladella, together with barcodes of the other European species T. dodonaea Stainton, 1858 and T. decidua Wocke, 1876 (Fig. 11). Each species has its own BIN, and is recognizable by its DNA barcode, except the pair T. caucasica and T. ekebladella. Unfortunately no barcodes are available for T. ekebladoides.

DNA barcodes of two Iranian specimens appeared to be identical to that of a Dutch specimen of Tischeria ekebladella and fall within one BINBOLD:AAF8247, that includes also all other available T. ekebladella barcodes. The barcode of a Georgian larva from leafmines examined by EvN (RMNH.INS.31425) groups with the Georgian adult of T. caucasica (GBRD286), together as sister to all remaining T. ekebladella and somewhat different from the Iranian T. caucasica. Within BINBOLD:AAF8247 the average distance is 0.38% and the maximum distance is 1.93% (n = 54). The nearest neighbor is BOLD:ACU6278, representing the East Palearctic Tischeria siorkionla Kozlov, 1986, with a distance of 8.01%.

Last instar larva (Fig. 4A‒I) pale yellowish green, with pale light-brown head, pronotal and anal plates slightly darker than body. Head flattened with almost 5–6 stemmata arranged linearly, two posterior ones with a relatively larger distance to the remaining ones (Fig. 4C). Hairs and bristles distributed as in T. ekebladella (for comparisons with that species refer to Grandi 1929, 1931, 1933, which are virtual identical descriptions).

Labrum (Fig. 4G) rectangular, and trilobed distally, with the two lateral lobes protruded considerably beyond median one, and a pair of fronto-lateral bristles; lateral and sub-posterior elements and medial ones represented by very small hairs, barely visible at very strong magnification. Mandible (Fig. 4F, H) sub-triangular, slightly longer than wide, tridentate apically with blunt tips; oral margin of mandible at proximal half with some setiform processes bent backwards and of different lengths (only visible under very large magnification). Maxillae (Fig. 4I) as in T. ekebladella, with large laminar stems provided with two very minute placoid structures, its detailed structure hardly visible under available magnification. Maxillary palpi (Fig. 4I) composed of three articles, decreasing in width towards the apex. Labium (Fig. 4I) nearly as that of T. ekebladella, but bi-articulated labial palpi were hardly visible under the available maximum magnification.

Antennae (Fig. 4H) three-segmented, segment 1 sub-cylindrical, slightly less than twice as long as wide, segment 2 slightly longer than wide, slightly less than half length of segment 1; distally with a long bristle, bristles on antennal surface hardly distinguishable under available magnification.

Thorax and abdomen. Nearly as in T. ekebladella; however, first thoracic segment not wider than other two (Fig. 4A‒E). In the examined larvae the 2^nd and 3^rd thoracic segments mostly with same width and length (although in some specimens segment 2 is slightly wider than 3), segment 1 always the shortest, its width slightly less than two remaining thoracic segments or occasionally with same width (Fig. 4A‒E). Abdominal segment 1 normally shorter than segments 2–6 (Fig. 4A, B), but in some specimens segments 1 and 2 equal in length and shorter than segments 3–6 (Fig. 4D). Abdominal segments 3–6 and sometimes 2–6 are usually the longest. Abdominal segment 1 usually slightly narrower than 2–6 or 3–6 (Fig. 4A, B, D).

Thoracic segments each with pair of very slightly sclerotized circular to elliptical plates on both dorsal and ventral surfaces towards the lateral sides (Figs 4D, E, 5A, B, F). The plates of the 2^nd and 3^rd tergites and sternites have nearly the same size and shape, and if elliptical positioned transversely. The plates of the 1^st tergite are elongate and somehow obliquely arranged to the longitudinal axis of the body (Figs 4D, E, 5A); while those of the 1^st sternite are sometimes irregularly shaped (Fig. 5F). These plates usually have a less sclerotized centre and are hardly visible, but they can be observed after staining the abdomen (Figs 4D, E, 5A, B, F).

Thoracic segments with single long bristle laterally, and pair of short bristles dorso-laterally (Fig. 5B). Thoracic legs nearly pyramid-shaped or finger-shaped (Figs 4D, E, 5F), very short, bearing two microscopic internal hairs distally (Fig. 5F).

Abdominal tergites 1–8 with pair of long bristles on either side. Tergites 1 and 2 with two pairs of short bristles positioned rather medio-laterally, and nearly in a line (Fig. 5C). Tergites 3–6 with two pairs of short bristles, which are not in the same line and positioned more laterally, and two pairs of nearly short, robust asymmetrical bristles internal to the lateral pairs, implanted on a common base or with bases completely confluent (Fig. 5D, E). Tergites 7 and 8 each with a submedial hair. Tergite 8 with five pairs of long marginal hairs, and some serration towards the middle part. Four pairs of the prolegs on segments 3–6 atrophied, without any trace of crochets (Figs 4D, 5G), but with three hairs externally (Fig. 5G). Last proleg provided with pair of linear series of crochets, which in specimens examined each contain between 20–24 elements. They are not in equal numbers normally (n = 3) (Fig. 5H). For measurements of head capsule and body see Table 1.

Pupa (Fig. 6A‒E) pale ochre to ochreous-brown, with brown scales at the dorsal surface of abdominal segments. The sex of the pupa can be easily recognised by shape of last abdominal segment (Fig. 6D, E). In the female, the paired triangular processes at the end of last abdominal tergum are slightly larger and closer to each other than in the male (Fig. 6D). Papillae anales of female genitalia usually visible. Female pupa slightly larger than male. Length of male pupa 3.70–4.40 mm (x = 4.06 mm ± 0.17, n = 22); female pupa 3.50–5.1 mm (x = 4.28 mm ± 0.43, n = 25).

Host plants: In Iran Quercus castaneifolia, Q. infectoria, Q. libani, Q. macranthera, and Q. robur (Fagaceae). In Peykan Shahr the percentage of infection in Q. robur was very high (Fig. 7B‒I), higher than in Q. infectoria (Fig. 8A‒C), and in the latter species it was higher than in Q. libani (Fig. 8D, E). In Q. libani the infection rate was very low (one or two mines on each leaf) (Fig. 8D). Mines from Georgia (one larva barcoded RMNH.INS.31425) that most likely belong to this species, were found on Q. petraea subsp. polycarpa (Schur) Soó and Castanea sativa Mill.

Based on the results of this study, T. caucasica has two generations per year in Tehran province. The eggs are white-grey in colour, deposited on the upper side of the leaf near the mid-rib or on larger lateral ribs. The mines consist of milky white upper-surface primary flat blotches, sometimes with orange hues, and like T. ekebladella, with no traces of preceding corridors (Figs 7A‒K, 8A‒E). As they get older, they turn very pale brown (Figs 7A‒H, 8A‒C). The cocoon was made nearly at the centre of each blotch (the so called nidus) and pupation took place within the mine. No frass was observed inside the mine (Figs 7G, K, 8B, C).

Leafmines were observed in 2017 and 2018 from late May to the first half of December, when the blotches were very minute. The greatest number of the mines in the first generation appeared in the first half of June. Adults were observed in early to mid July, when the next generation would occur. In the first generation, many adults emerged and, as a result, the population density increased remarkably during the second generation. In November, during leaf fall, the last instar larvae were overwintering inside the blotches. Live larvae were observed, yet sedentary through a cut in the upper epidermis in winter at the height of the cold. They became active and pupated after the end of the cold weather period.

Four larval instars were estimated by measuring the width of the head capsule of each larval instar and applying Dyar’s law (Dyar 1890). Body length and head capsule width, min-max (mean ± SE) in all larval instars, are shown in Table 1.

In Tehran (Peykan Shahr) the activity period of T. caucasica larvae started from the end of May and continued in two generations until the beginning of December. In the first generation during 2018, the 1^st larval instar hatched late May (22^nd) and feeding on leaf tissue within the mine. The first and second larval instars formed a short linear mine towards the leaf edge. Third instar larvae appeared at the end of May (27^th) , the last instar nearly at the first half of June (10^th), and adults emerged from mid June (16^th) to the 1^st of July.

The number of mines per leaf varies; in some cases there are as many as 12 mines (Fig. 8C). The first larval instar of the second generation appeared at the end of August (30^th) and the last instar on late September. They overwintered as 4^th instar larvae and remained quiescent until pupation in late May to early June. According to the results of this study, T. caucasica had a severe outbreak between 2017 to 2019, despite spraying the trees and destroying the fallen leaves. In 2020, the infection was considerably lower, and there was no trace of infection in the garden until May 2021.

Two species of parasitic wasps, Brachymeria excarinata Gahan, 1925 (family Chalcididae) and Aprostocetus sp. (family Eulophidae) emerged from the mines.

Georgia (Klasiński et al. 2020) and Iran (Tehran: Peykan Shahr; East Azarbaijan: Ainalou; Gilan: Molumeh; Mazandaran: Neka) (Fig. 9).

Iran, Tehrān Prov., Tehrān, Peykān Shahr, 35°44'27"N, 51°10'50"E, 1317 m, 33 ♂♂ 26 ♀♀, emerged 6, 9, 16, 19, 21, 23, 24.vi.2018, 11 ♂♂ 23 ♀♀, emerged 3, 6, 7, 10, 13.vii.2019 (first generation); 7 ♂♂ 8 ♀♀, emerged 9.xii.2018 (second generation), S. Farahāni leg., all reared from Quercus robur (genitalia slides HA-2375, HA-2376, HA-2377, HA-2378, HA-2379, HA-2380, HMIM); Same locality, 1 ♂, 1 ♀, emerged 11.vi.2016, genitalia slides EvN5029 (♂), EvN5030 (♀), specimens barcoded, RMNH.INS.25029, RMNH.INS.25030 (RMNH); 2♂, 1♀, not pinned, same data, RMNH.

East Āzarbāijān Prov., Arasbārān protected area, Āsheghlou to Vāyeghān Rd., near Āinalou, 38°58'4.3"N, 46°42'27.6"E, 513 m, 17.ix.2008, J. Buszko leg., Quercus macranthera, 3 mines, 1 ♂ 2 ♀♀, emerged 16.–26.iii.2009, genitalia slide ♂ EvN5287, RMNH.INS.25287, RMNH, HMIM; Gilān Prov., Molumeh, 36°51'17.01"N, 49°55'48.60"E, 11.ix.2008, J. Buszko leg., Quercus macranthera, 3 mines, 1 ♂ 1 ♀, emerged 17.iii.2009, genitalia slide ♀ EvN5291, RMNH.INS.25291, RMNH, HMIM.

Māzandarān Prov., Nekā, 36°30'16.7"N, 53°23'27"E, 530 m, 30.ix.2007, J. Buszko leg., Quercus castaneifolia, 4 mines, 2 ♂♂ 2 ♀♀, emerged 20.–26.ii.2008, genitalia slides ♂ EvN5289, EvN5290, RMNH.INS.25289, RMNH.INS.25290, RMNH, HMIM.

Georgia, 1 ♂ Lesser Caucasus, Samtskhe-Javakheti, Borjomi, Kvabiskhevi, 8.vii.2019, Leo Vahatalo, Anssi Vahatalo leg., LepiLED, genitalia slide EvN5274, specimen barcoded, GBRD.286, Research Collection of Anssi & Leo Vahatalo.

Leafmines most likely belonging to T. caucasica: Adjara AR, Chakvistavi, 41°40'37"N, 41°52'23"E, 19.ix.2018, M.V. Kozlov & V. Zverev, Castanea sativa, 7 mines, RMNH.INS.48084–48085 (RMNH); Adjara AR, Chikuneti, 41°34'13"N, 41°51'46"E, 26.ix.2018, M.V. Kozlov & V. Zverev, Quercus petraea subsp. polycarpa, 17 mines, RMNH.INS.48093–RMNH.INS.48095; Adjara AR, Zeda Chkhutuneti, 41°28'57"N, 41°51'46"E, 23.ix.2018, M.V. Kozlov & V. Zverev, Castanea sativa, 1 larva RMNH.INS.31425 (barcoded), 5 mines, RMNH.INS.48128–48129 (RMNH).

Belgium, Namur, Nismes, Tienne Breumont, 50.07849°N, 4.5444°E, 200 m, larva 2.x.1999, E.J. van Nieukerken & T. Jin leg., dry hill with limestone grassland, Quercus robur, 1 ♂ emerged 3.iv.2000, RMNH.INS.23623 (RMNH), genitalia slide EvN3623; France, Var, St. Maximin-le-Ste. Beaume, Camping, 1 ♂, 13.vi.1987, K.N. Nieuwland leg., RMNH.INS.15431 (RMNH); Netherlands, Limburg, Sevenum, 1 ♀, 4.vi.1983, G.R. Langohr leg., RMNH.INS.15428 (RMNH); Netherlands, Limburg, Well, 1 ♂, 2.vi.1983, G.R. Langohr leg., RMNH.INS.15429 (RMNH); Netherlands, Noord-Holland, Oude Schuithuis, 1 ♂, 1.vii.1981, J.C. Koster leg., RMNH.INS.845174 (RMNH), slide only, genitalia slide JCK752; Netherlands, Overijssel, Denekamp, Kanaal Almelo-Nordhorn, 1 ♀, 10.viii.1984, J.C. Koster leg., RMNH.INS.845695 (RMNH) slide only, genitalia slide JCK1273; Netherlands, Overijssel, Diepenveen, Platvoet, W. Oord leg., 1 ♀, emerged 18.xii.1982, RMNH.INS.15430 (RMNH); Netherlands, Overijssel, Losser, De Tip, 1 ♀ , 14.vii.1981, J.C. Koster leg., RMNH.INS.15432 (RMNH), genitalia slide JCK0755, RMNH.INS.845177.

Europe: Albania, Austria, Belarus, Belgium, Croatia, Czechia, Denmark, Estonia, Finland, France (including Corsica), Germany, Great Britain, Greece (including Crete), Hungary, Ireland, Italy (including Sardinia), Latvia, Lithuania, Luxemburg, Netherlands, N. Macedonia, Norway, Poland, Portugal, Romania, Russia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Ukraine; Asia: Turkey (Diškus and Puplesis 2003a; Koçak and Kemal 2009; Karsholt and van Nieukerken 2017). Former records of Caucasus (Diškus and Puplesis 2003a) probably all refer to T. caucasica, this may be also (partly) true for the Turkish records.

Castanea mollissima Blume, C. sativa, Quercus cerris L., Q. dalechampii Ten., Q. faginea Lam., Q. x hispanica Lam., Q. frainetto Ten., Q. ithaburensis subsp. macrolepis (Kotschy) Hedge & Yalt., Q. macranthera, Q. macrocarpa Michx., Q. petraea (Matt.) Liebl., Q. pubescens Willd., Q. robur, Q. robur subsp. pedunculiflora (K.Koch) Menitsky, Q. rubra L., Q. serrata Murray (Diškus and Puplesis 2003a; Ellis 2021).

Spain, paratype, Hispania, Andalusia, Camino de Istan, ca 200 m, 1 ♂, 6.iv.1984, E. Traugott-Olsen leg., ZMUC; Same data, 1 ♀, 9.iv.1984, E. Traugott-Olsen leg., ZMUC, genitalia slides Diškus 032, Diškus 086.

Europe: Portugal, Spain; Northern Africa: Tunisia (Diškus and Puplesis 2003a).

Quercus canariensis Willd. (= mirbeckii Durieu), Q. suber L. (Diškus and Puplesis 2003a).

Iran, East Āzarbāijān Prov., Kaleybar, 16.ix.2008, J. Buszko leg., Quercus macranthera, 1 ♂, emerged 26.iii.2009, genitalia slide EvN5288, HMIM.

Europe: Austria, Belarus, Belgium, Czechia, Denmark, Estonia, Finland, France, Germany, Great Britain, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Netherlands, Norway, Poland, Portugal, Romania, Russia, Slovakia, Slovenia (new record, BOLD, see Fig. 11), Spain, Sweden, Switzerland, Ukraine; Asia: Turkey, Caucasus (Diškus and Puplesis 2003a; Gozmány 2012; Laštůvka and Laštůvka 2015; Karsholt and van Nieukerken 2017).

Castanea sativa, Quercus cerris, Q. faginea, Q. x hispanica, Q. ithaburensis subsp. macrolepis, Q. petraea, Q. pubescens, Q. robur and Q. rubra (Diškus and Puplesis 2003a; Ellis 2021).

Iran, Tehrān Prov., Tehrān, Lavāsān, 28.ix.2007, J. Buszko leg., Rosa, 2 mines, 1♂, 1♀, emerged 3.–8.iii.2008, genitalia Slide ♂ EvN5296, RMNH.INS.25296, RMNH, HMIM.

Europe: Austria, Belarus, Belgium, Croatia, Czechia, Denmark, Finland, France, Germany, Great Britain, Greece, Hungary, Italy (including Sardinia), Latvia, Lithuania, Malta, Moldova, Netherlands, N. Macedonia, Norway, Poland, Portugal, Romania, Russia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Ukraine; Asia: Turkey, Caucasus, Turkmenistan, far eastern Russia (Primorskiy Kray), South Korea, Japan (Honshu) (Diškus and Puplesis 2003a; Corley 2015; Aarvik et al. 2017; Karsholt and van Nieukerken 2017), and Iran (Nematollahi and Radjabi 2002).

Rosa spp. (including R. arvensis Huds, R. canina L., R. x centifolia L., R. x damascena Mill., R. gallica L., R. indica L. (= borboniana N.H.F. Desp.), R. luciae Franch. & Rochebr (= wichuraiana Crép.), R. multiflora Thunb., R. pendulina L., R. rubiginosa L., R. rugosa Thunb., R. sempervirens L., R. tomentosa Sm., R. virginiana Herm.) (Diškus and Puplesis 2003a; Kobayashi et al. 2016; Ellis 2021).

The species was earlier reported from Kashan on damask rose (Rosa damascena) (Nematollahi and Radjabi 2002). The short description provided is confusing as they compare the moth and leafmine to a Stigmella species (family Nepticulidae) and also give a size of the moth that fits a Stigmella better; we therefore doubt that this species was correctly identified. Our current record confirms the occurrence of this species in Iran. Recently Coptotriche angusticollella was reported from damask rose in Turkey (Demirözer et al. 2018).

Iran, Māzandarān Prov., Nekā, 30.ix.2007, J. Buszko leg., Prunus cerasifera, 1 mine, 1 ♂ emerged 23.ii.2008, EvN5295, RMNH.INS.25295, RMNH; Tehrān Prov., Tehrān, Lavāsān, 28.ix.2007, J. Buszko leg., Prunus cerasifera, 2 mines, 1♂, 1♀, emerged 3.–5.iii.2008, HMIM.

Europe: Austria, Belgium, Czechia, France (including Corsica), Germany, Great Britain, Greece, Hungary, Italy, Poland, Romania, Russia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Ukraine; Asia: Caucasus, Turkmenistan (Diškus and Puplesis 2003a; Laštůvka and Laštůvka 2015; Karsholt and van Nieukerken 2017).

Prunus cerasifera Ehrh., P. cerasus L., P. domestica L., Prunus domestica subsp. insititia (L.) Bonnier & Layens, P. microcarpa C. A. Mey., P. persica (L.) Stokes, and P. spinosa L. (Diškus and Puplesis 2003a; Ellis 2021).

Iran, Māzandarān Prov., Vāz, 3.x.2007, J. Buszko leg., Rubus, mines, rearing failed; Tehrān Prov.: Tehrān, Lavāsān, 28.ix.2007, J. Buszko leg., Rubus, mines, rearing failed.

Europe: Albania, Austria, Belarus, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Finland, France (including Corsica), Germany, Great Britain, Greece (including Crete), Hungary, Ireland, Italy (including Sardinia and Sicily), Luxemburg, Malta, Netherlands, N. Macedonia, Poland, Portugal, Romania, Russia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Ukraine; North Africa: Morocco; Asia: Turkey, Turkmenistan (Diškus and Puplesis 2003a; Koçak and Kemal 2009; Zerafa 2009; Karsholt and van Nieukerken 2017), Iran (Amsel 1959, 1961; Mirzayans and Kalali 1970).

Rubus caesius L., R. canescens DC., R. fruticosus L., R. grabowskii Weihe, R. idaeus L., R. laciniatus Willd., R. macrophyllus Weihe & Nees, R. nemorosus Hayne, R. pedunculosus D. Don, R. plicatus Weihe et Nees, R. saxatilis L., R. tomentosus Borkh., R. ulmifolius Schott (Diškus and Puplesis 2003a; Ellis 2021).

Already Christoph (1873) reported this species from Golestan, from a specimen taken around oaks; its is therefore possible that he mistook a Tischeria species. Later Mirzayans and Kalali (1970) recorded it from leafmines, and also our new records are based on the characteristic leafmines, as rearing failed, but confirmation by adults would still be required as several species feed on Rubus.