Western Palaearctic Ectoedemia (Zimmermannia) Hering and Ectoedemia Busck s. str. (Lepidoptera, Nepticulidae): five new species and new data on distribution, hostplants and recognition

Th e nine western Palaearctic species of the subgenus Zimmermannia Hering, 1940 and 48 species in the subgenus Ectoedemia Busck, 1907 of the genus Ectoedemia are reviewed. One species in the subgenus Zimmermannia and four species in the subgenus Ectoedemia are described as new: Ectoedemia (Zimmermannia) vivesi A. Laštůvka, Z. Laštůvka & Van Nieukerken sp. n. from southern Spain and Cyprus with unknown host plant, Ectoedemia (E.) hendrikseni A. Laštůvka, Z. Laštůvka & Van Nieukerken sp. n. from southern France on Quercus suber, E. (E.) heckfordi Van Nieukerken, A. Laštůvka & Z. Laštůvka sp. n. from southern England on Quercus petraea and Q. robur, E. (E.) phaeolepis Van Nieukerken, A. Laštůvka & Z. Laštůvka sp. n. from Spain and Portugal probably on Quercus ilex and Q. rotundifolia and E. (E.) coscoja Van Nieukerken, A. Laštůvka & Z. Laštůvka sp. n. from Spain on Quercus coccifera. Th e following species are redescribed: Ectoedemia (Zimmermannia) hispanica Van Nieukerken 1985, Ectoedemia (Zimmermannia) reichli Z. & A. Laštůvka 1998, Ectoedemia (E.) algeriensis van Nieukerken 1985, E. (E.) pseudoilicis Z. & A. Laštůvka 1998 and E. (E.) alnifoliae van Nieukerken 1985. Ectoedemia albiformae Puplesis & ZooKeys 32: 1–82 (2010) doi: 10.3897/zookeys.32.282 www.pensoftonline.net/zookeys Copyright E.J. van Nieukerken, A. Lašto uvka, Z. Lašto uvka. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. MONOGRAPH Launched to accelerate biodiversity research A peer-reviewed open-access journal

In the same period the knowledge of the family Nepticulidae has increased enormously: three major fauna books were published ; Laštůvka and Laštůvka 1997;Puplesis 1994) and a host of checklists and faunistic papers appeared, several particularly relevant for Nepticulidae (e.g. France: Van Nieukerken et al. 2006, Greece: Laštůvka andLaštůvka 1998, Iberian peninsula: Van Nieukerken et al. 2004a, Italy: Laštůvka and, Russia: Van Nieukerken et al. 2004b and Sweden: Bengtsson et al. 2008). Th e faunistic knowledge was summarised in two European checklists, the latest on the internet ( Van Nieukerken 1996;. In this paper we aim to update the taxonomic, faunistic and biological knowledge of these two subgenera. In the 1985 revision two species associated with Quercus were mentioned from Spain, but remained undescribed because of lack of material. Both species have now been found in numbers and one has been reared. A very similar but diff erent one also suddenly appeared in southern England (Edmunds 2009). Another new Quercus feeding species was discovered in southern France and a new Zimmermannia was discovered in Spain and Cyprus. Additional material made it possible to describe the females of Ectoedemia (Zimmermannia) hispanica and E. (Z.) reichli and the males of Ectoedemia (E.) alnifoliae. Rearing and collecting large numbers of E. (E.) algeriensis from France (Van Nieukerken et al. 2006) showed that the supposed male in the 1985 treatment was correctly associated. Studies on molecular markers in the genus aided us in decisions about species status, and will be published elsewhere.
We concentrate on the Western Palaearctic region, here restricted to the land mass west of the biogeographic meaningful "Turgai straits", now more or less following the 64-65 east meridian, from North to South along the rivers Ob, Tobol, Turgay, Aral See, Karakum desert and border between Iran and Afghanistan/Pakistan. Th is means that one species treated in 1985 is excluded as East Palaearctic: Ectoedemia (Zimmer-mannia) nuristanica Van Nieukerken, 1985; further the four species described from Tadzhikistan since 1985 are excluded (Puplesis 1988a;. However, the Kopet Dag range in Turkmenistan and all of Iran are included: they have an essentially European-Mediterranean fauna. Th e fauna associated with oaks apparently is much richer than was understood in 1985, and especially in the Mediterranean region identifi cation can be diffi cult. We recognize now 26 species in all, seven in the suberis group, and 19 in the subbimaculella group, although the specifi c status of a few in the latter is uncertain. To aid in the identifi cation, we provide new keys for the Quercus feeding species as well as one for the subgenus Zimmermannia. Together with the recent revision of the oak mining species of Stigmella ( Van Nieukerken and Johansson 2003), all western Palaearctic Nepticulidae mining Quercus can now be safely identifi ed in the adult stage. As in Stigmella, identifi cation of mines alone is only possible for some of the species, but DNA barcoding of larvae is already becoming a good alternative to rearing adults ( Van Nieukerken & Doorenweerd in preparation).

Methods
Th is paper should be seen as an update to the earlier revision ( Van Nieukerken 1985); because that is now also available online we try to avoid unnecessary repetition. Th us we only give descriptions for new species and redescriptions for those discovered after 1985, or for which much more material has become available since then. Diagnoses are given for species related to the newly described species since 1985, and where we have additional data compared to 1985.
After the species heading, full references, nomenclatorial details and synonymies are only given for taxonomic decisions published after Van Nieukerken (1985).
A section "Biology" is given for all species, including all the available hostplant information. A section "Distribution" summarises new records and only in some cases (re)describe the complete distributional area. For the material section see below.
Remarks. In the Iberian Peninsula and S. France the species can be confused with the externally similar E. hispanica. Th e hairpencil and costal emargination in the male are diagnostic, as are the diff erences in number of setae on the female terminal segments.

Ectoedemia (Zimmermannia) liebwerdella Zimmermann
Biology. Hostplants: reared from Fagus sylvatica L. Adults have been collected in forests without Fagus, and only Quercus present, and also far outside the distribution area of Fagus, so oaks are the likely alternative host. Larvae with a one or two year life cycle. Adults fl y from 7 June to 22 August.

Ectoedemia (Zimmermannia) longicaudella Klimesch
Biology. Hostplants: reared from Quercus robur L. Adults have been collected in forests with other species of Quercus as well. Larvae with a one or two year life cycle. Adults fl y from 6 June to 30 August, not as late as atrifrontella.
Biology. Hostplants: not reared, most likely feeding on evergreen Quercus. Adults fl y from 8 June to 23 August.
Remarks. Since the original description was based on just two worn males, the species is here redescribed. Th e material includes also that given earlier for France, Spain and Portugal (Van Nieukerken et al. 2004a; Diagnosis. Externally similar to Ectoedemia hispanica, but thorax greyish brown and not pale and hindwing with a pale grey-brown hair pencil (white in hispanica). Male genitalia very diff erent from all other Zimmermannia species by presence of one pair of carinae (ventral) only and simple triangular valvae, more reminiscent of Ectoedemia s. str. Female as yet unknown.
Female. Unknown. Male genitalia (Figs 4,(11)(12)(13). Capsule length ca. 350 μm. Vinculum distinctly concave anteriorly. Tegumen broad, blunt, with simple long setae. Gnathos large, lateral processes short and broad, central process long and acute. Valva length ca 225 μm, more or less triangular, with long distal process, distinctly curved towards middle, sublateral processes about one third transtilla length. Aedeagus ca 355 μm long, with only  apomorphy. Overall it resembles Zimmermannia species better in the combination of colour pattern, hair pencil, gnathos shape, valva and aedeagus shape than Ectoedemia s. str. We expect that this is also a barkminer, possibly of Quercus, but other hosts are also possible. Th e distributional disjunction is remarkable. Etymology. Named after Dr. Antonio Vives Moreno (Madrid), in gratitude for his long lasting support of our investigations in Spain; vivesi is a noun in genitive case. Van Nieukerken Biology. Hostplants: not reared, most likely feeding on evergreen Quercus. Adults fl y from 15 June to 22 September.

Remarks.
In the female genitalia of two specimens (EJvN3106, Tahir Geçidi and EJvN3121, Kirki), some small spines were noted in the ductus bursae, also the setae were not pectinate as in the original specimens. Since not many female genitalia have been thoroughly compared we are not certain whether this is due to variation, or that these represent another unknown species or may be the unknown female of E. vivesi.

Ectoedemia (Zimmermannia) amani Svensson
Remarks. Th e knowledge of the distribution of this species has been increased enormously. In 1990 it was known from a number of records from Sweden, Norway, Denmark, Austria and Macedonia. Now it is known from almost all European countries except the Iberian Peninsula. Since this species has a trans-palaearctic distribution, a range expansion seems a possibility. However, statistical data (see discussion) do not support an increasing proportion of this species in the total record of Nepticulidae, so that increased collection eff ort currently provides the best explanation.
Material. Azerbaijan. 1♂ diff er from other Zimmermannia by the shape of the valva, short gnathos and confi guration of carinae. Female genitalia resemble most those of E. liguricella, but have fewer convolutions in the ductus spermathecae. Redescription. Male. Forewing length 2.7-3.8 mm, wingspan 6.8-7.2 mm. Head: frontal tuft yellowish white to ferruginous, collar paler, whitish. Antennae long, with 38-44 segments. Th orax and forewings brown, slightly irrorate with yellowish brown, often an indistinct pale tornal spot. Cilia line indistinct, terminal cilia yellowish white. Hindwing with a very indistinct greyish hairpencil, not surrounded by androconial scales; humeral lobe not pronounced. Abdominal tufts grey white.
Male genitalia (Figs 8-10). Capsule length ca. 380 μm. Tegumen broadly rounded, forming indistinct pseuduncus. Gnathos with short, rather broad central element. Valva length ca. 270-275 μm, broadly triangular, in middle suddenly narrowing strongly into curved distal process. Sublateral processes long, almost as long as transverse bar of transtilla. Aedeagus length ca. 380-395 μm long; ventral carinae long, parallel and wide apart; a pair of outwards curved strong lateral carinae, and a pair of similar outwards curved dorsal carinae (in slides often hidden behind lateral pair). Vesica with one straight long and pointed cornutus, and some very small additional ones.
Biology. Hostplant unknown. Adults collected from 26 May to 18 June in Greece and 28 June to 7 August elsewhere.
Distribution. Described from the male holotype from Greece (Preveza) and later recorded from the Czech Republic (Šumpich et al. 2007). Here we report new records from (northern) Italy, Slovakia (see also Tokár et al. in press), Croatia: Krk and Switzerland.
Remarks. Since the original description was based on a single male, and published in German, the species is here redescribed. Th e species was regularly collected in forests without Quercus, but we have no positive indication what tree or shrub the hostplant could be.

Subgenus Ectoedemia
Diagnosis. Moderately sized Nepticulidae, usually with a distinct cilia line and pale cilia, and pale (dull white or metallic) fascia or separate spots; scaling rather coarse. Male abdominal tip usually with conspicuous anal tufts. In Europe other Ectoedemia subgenera that resemble these species are Etainia Beirne, 1945, which all have a fascia in addition to a dorsal and costal spot and Fomoria Beirne, 1945, which usually have a dorsal spot only (also in Ectoedemia intimella, but that has a pale antenna), and when a costal spot is present as well, both spots are far apart. Th e few species without colour pattern may be confused with Bohemannia pulverosella (Stainton, 1849) or Trifurcula, see generic diagnosis.

Key to males of Fagaceae feeding species
Th e keys of the previous revision (Van Nieukerken 1985) can still be used to identify most Ectoedemia (s.s.) species, since most additions are in the Fagaceae feeders, we provide here a new key to males for that group. Th is key works best for freshly collected or reared males; worn specimens may be diffi cult to identify without dissection. For checking hairpencils the forewing needs to be slightly lifted.
Since females are much harder to identify by external characters, we do not provide a key for those. When no males are present, often the genitalia need to be dissected. Hindwing upperside and forewing underside with patch of brown androconial scales. Costal spot usually much smaller than dorsal spot (Fig. 112) Puplesis, 1994 andE. similigena Puplesis, 1994. Th ey are very similar to E. turbidella, and these species will be diagnosed below. Since only few worn specimens of the new species were available, we only provide a partial redescription. Also males of E. argyropeza have been described since 1985 (Bond and Van Nieukerken 1987).

Ectoedemia (Ectoedemia) intimella (Zeller)
Biology. Hostplants: Salix caprea L., S. cinerea L., S. pentandra L., S. fragilis L., S. phylicifolia L., S. dasyclados Wimm. (Ellis 2007). In Finland a population of E. intimella was discovered feeding on Populus balsamifera L. Th ese mines resemble those of E. hannoverella, but the mine usually enters the leaf blade further away from the petiole and the larva doesn't show the characteristic plates of E. hannoverella . Univoltine, larvae in September to November, adults from May to July. Distribution. Norway: Aarvik et al. 1997;Ireland: Shackleton 1977; Latvia: Šulcs and Šulcs 1989; Lithuania: Ivinskis et al. 1985;Slovakia: Laštůvka and Laštůvka 1990b;Spain: Laštůvka and Laštůvka 2008. For new French records: Van Nieukerken et al. 2006, here we record it new from southern France. Also found in far eastern Russia: Sakhalin (Puplesis 1994). Here for the fi rst time recorded from Japan, Hokkaido, not far from Sakhalin.

Remarks.
Ectoedemia hannoverella seems to be expanding its range, and has done so in the last 20 years at least westwards to Britain and northwards to Sweden. carinae; ventral pair at extreme posterior tip, basally connected, pointed; dorsolateral pair more anteriorly placed, longer than ventral ones, strongly curved in lateral view, dorsally connected, slightly asymmetrical.
Biology. Hostplant unknown, a Populus is likely. Adults found in May. Distribution. Ectoedemia similigena is only known from its type locality, Jalta botanical garden, on the Crimea (Ukraine). Th e species has not been found again, and we consider it possible that it is actually a species from somewhere in Central or Eastern Asia, introduced with plants.
Remarks. E. similigena is very similar to E. turbidella (see diagnosis), but on the basis of the diff erences seen we conclude tentatively that it is a separate species.
Biology. Hostplant unknown, a Populus is likely. Adults found in April and May. Distribution. South Turkmenistan and N. Iran. Remarks. Th e females from Iran, tentatively identifi ed as Ectoedemia turbidella by Van Nieukerken (1985)
Remarks. In Ireland a few males of this parthenogenetic species were discovered and described (Bond and Van Nieukerken 1987). (Klimesch) Biology. Hostplants: Ulmus spp. Univoltine, larvae September to October, adults fl y from April to June.

Th e Ectoedemia suberis group
In southwestern Europe a group of species forms a complex around Ectoedemia suberis. Only E. caradjai and E. aegilopidella do not belong to this complex. Apart from E. suberis and E. andalusiae, we recognise three more new species in this complex.
Th ese species have very similar external features and can be best separated on the male hairpencil colour and that of associated special scales and the hairiness of the female abdominal tip. Th e male and female genitalia show only some subtle diff erences, and also the leafmines are inseparable.
Th e decision to name these forms as full species was supported by molecular analyses both of the genes CO1 and Elongation Factor, which range from 5.07% to 8.23 % in CO1 and 2.5-2.9% in Elongation Factor, comparable to the diff erences found in the populella group and larger than between species in the albifasciella complex. Th ese data will be published elsewhere ( Van Nieukerken & Doorenweerd, personal communication). Table 1 summarizes the diff erences in this complex and the rather similar E. caradjai. (Klimesch) Fig. 44 Biology. Hostplants: Quercus ithaburensis Decne subsp. macrolepis (Kotschy). Larvae found in September, adults reared in April, in the fi eld collected in August, suggesting a second generation.

Ectoedemia (Ectoedemia) aegilopidella
Distribution. Only known from Greece: Rhodos, to be expected in Turkey. Old leafmines, found in mainland Greece (Lakonia), may be this species, but since E. heringi also occurs on the same host and may have rather similar mines, we cannot record it defi nitely from mainland Greece.

Ectoedemia (Ectoedemia) caradjai (Groschke) Figs 43, 53
Diagnosis. Ectoedemia caradjai can be separated from all other species in the suberis group by the white basal patch on the forewings, and in the male by the absence of androconial scales surrounding the hairpencil on the hindwing. In the male genitalia the truncate gnathos is diagnostic, the valvae resemble those of E. phaeolepis, and somewhat E. heckfordi (Fig. 53).
Biology. Hostplants: Quercus pubescens Willd., Q. petraea (Matt.) Liebl., Q. pyrenaica Willd., Q. frainetto Ten., Q. infectoria Olivier, Q. coccifera L. (Van Nieukerken 1985; Van Nieukerken et al. 2004a) Larvae from July to October. Adults from April to September. Since in Croatia from July larvae adults still emerged in July and from August larvae in September, we assume that there are at least three generations (at least partly) in southern Europe.
Distribution. Spain

Ectoedemia (Ectoedemia) suberis (Stainton) Figs 33, 34, 48
Diagnosis. Male Ectoedemia suberis is separated from the four following species by the completely white androconial fi eld on the hindwing, surrounding a white hair-pencil, and scales not reaching the dorsal margin of the hindwing (Figs 33, 34). E. caradjai has a basal white streak on the forewing and lacks the androconial fi eld; other similar Ectoedemia (such as haraldi) are usually smaller, have fewer antennal segments and have no hairpencil. Females are distinguished from E. andalusiae, heckfordi and phaeolepis by the hairy abdominal tip, and from hendrikseni by the larger number of antennal segments. Male genitalia diff er from E. andalusiae and hendrikseni by the very slight basal bump, which is much more distinct in the other species, whereas this bump is virtually absent in E. phaeolepis and caradjai (Fig. 48). See also E. phaeolepis. Aedeagus at 360-400 μm distinctly longer than in E. phaeolepis and andalusiae (less than 350μm). In female genitalia hard to distinguish from E. andalusiae or E. phaeolepis, apart from slight statistical diff erences in numbers of setae on anal papillae and number of convolutions in ductus spermathecae (6 in andalusiae, 5.5 in heckfordi, 4.5 in phaeolepis).
Biology. Hostplants: evergreen oaks Quercus ilex L., Q. rotundifolia Lam., Q. suber L. and Q. coccifera L. and possibly on semi-evergreen oak Q. faginea Lam. (vacated mines only) ( Van Nieukerken 1985;Van Nieukerken et al. 2004a;. Mine not separable from the similar mines on evergreen oaks of E. andalusiae, phaeolepis and hendrikseni, but on Q. coccifera, E. andalusiae is more often found than suberis. Univoltine, with larvae from January (Algarve, Andalusia) to late April (Spanish mountains, France) and adults fl ying from July to October.
Distribution. Earlier records from Tunisia, based on leafmines only, are confi rmed by the record of reared adults below. Given the fact that the closely related E. hendrikseni, E. andalusiae and E. phaeolepis cannot be separated by their mines, the previous records from Algeria on the basis of their leafmines ( Van Nieukerken 1985) should be reconsidered and regarded as doubtful, although in the light of distribution data suberis is still the most likely candidate.
Male genitalia (Figs 45,49). Vinculum distinctly concave anteriorly. Tegumen with a bulbous, rounded pseuduncus, with several long setae. Gnathos broad, its processes and connecting bar of equal breadth, lateral processes rounded at their ends, only slightly shorter than central element. Valva broad, with an angular inner lobe on its base and with only slightly elongate distal process, sublateral processes about ¼ transtilla length. Aedeagus with distinct cathrema, with numerous very small cornuti in vesica, carinae simple, slightly curved, not pointed.
Female genitalia (Figs 54, 58, 62). T7 on either side with a patch of ca 35 long setae along margin with T8; T8 on disc with ca 40 long setae on either side; T8 with rounded corners, slightly indented posteriorly. Papillae anales broadly rounded, with many short setae, approximately 80-90 on each. Apophyses short and thick. Vestibulum with distinct vaginal sclerite and a spiculate pouch with small spines arranged as pectinations. Corpus bursae ca. 790μm long, almost globular, covered with minute pectinations; signa dissimilar, resp. ca 415 and 610 μm long (n=1), approximately 2.1-2.9 as long as wide. Ductus spermathecae with about 5.5 narrow sclerotized convolutions.
Biology. Hostplants: only reared from Quercus coccifera L. ( Van Nieukerken 1985;Van Nieukerken et al. 2004  Leafmines (Fig. 65). Probably indistinguishable from the mine of E. suberis; a narrow, partly contorted gallery in the fi rst half changes suddenly in a large blotch mine with two lateral bands of frass.
Larva. Green. Life history. Larvae collected in March, several weeks earlier than in E. suberis, most adults on the wing in May and June; an individual record in October may represent a (partial) second generation or a late emerging specimen. In this case the species fl ies together with E. suberis.
Distribution. Known from a small area in the Provence region in Southern France: the Estérel massif in the Alpes Maritimes and Var.
Etymology. Dedicated to the memory of Henning Hendriksen (Copenhagen) who collected this species for the fi rst time; hendrikseni is a noun in the genitive. Th e ardent collector and specialist of microlepidoptera Hendriksen died in February 2009 at the age of 80. Van Nieukerken Figs 37,38,50,57,61 Diagnosis. Male Ectoedemia andalusiae is separated from the very similar E. suberis by the ochreous hairpencil, from E. hendrikseni by the fi eld of white scales that does not reach the dorsal margin of the hindwing (Figs 37, 38). Male genitalia most similar to suberis, but capsule shorter and with blunt and wide tegumen (Fig. 50). Female genitalia resemble those of E. heckfordi and E. phaeolepis, but have more convolutions in the ductus spermathecae (6 against 4.5-5.5) (Figs 57, 61).

Ectoedemia (Ectoedemia) andalusiae
Biology. Hostplants: only reared from Quercus coccifera L. ( Van Nieukerken 1985;Van Nieukerken et al. 2004). Mine not separable from that of E. suberis or phaeolepis. Possibly bivoltine, with larvae from January to March and in June, and adults fl ying from April to July and September-October. From larvae collected in the middle of June in Algarve we reared adults in following days (end of June, beginning July), so that more than two generations are possible in Portugal. Th e second generation may fl y together with E. suberis.
Distribution are externally not separable. Male genitalia diff er from suberis, andalusiae and hendrikseni by the very slight basal bump, which is much more distinct in the other species, whereas this bump is virtually absent in E. phaeolepis and caradjai. Th is and the next species diff er from E. caradjai by the pointed gnathos (truncate in caradjai). Aedeagus at 360-400 μm distinctly longer than in E. phaeolepis and andalusiae (less than 350μm). In female genitalia hard to distinguish from E. andalusiae or E. phaeolepis, apart from slight statistical diff erences in numbers of setae on anal papillae and number of convolutions in ductus spermathecae (6 in andalusiae, 5.5 in heckfordi, 4.5 in phaeolepis).
Biology. Hostplant. Reared from Quercus petraea (Matt.) Liebl.; in 2006 vacated mines were also found on Q. robur L.. Most mines were found on saplings, seedlings or low growth of Q. petraea and in the shade.
Leafmines (Figs 68, 69, 122). Egg on leaf underside, against leaf margin or a vein. Mine starts as a short, highly contorted gallery, occupying a small space only, fi lled with black frass, later broken; mine suddenly enlarged into an elongate blotch, with frass concentrated in two lines along the edges. Mine similar to that of other species in this group.
Larva. Distinctly green. Life history. Larvae have been found from 2 August to early September and adults emerged indoors in April and May, certainly univoltine.
Remarks. Th e mines of this species were discovered by Bob Heckford in a small forest in Devon. In the following years he and Phil Sterling found the mines in two more localities, all close to the fi rst one. Th e mines were either on saplings, seedlings or on low branches of the trees, always in the shade. Th ese mines were fi rst reported in Britain as "Ectoedemia caradjai" (Edmunds and Langmaid 2004).

Ectoedemia phaeolepis
Diagnosis. Very similar to E. heckfordi, but the dark brown androconial scales on the hindwing reach nearly to the middle of the wing and to its hind margin, hair pencil fuscous, see further E. heckfordi. (Figs 41, 42). Forewing length 2.4-2.8 mm, wingspan 5.4-6.0 mm. Head: frontal tuft dark ochreous, with several brownish scales, collar ochreous; scape yellowish white; antenna dark brown-black, with 39-41 segments. Th orax nearly black with several brownish scales in its hindpart. Forewing covered with dense black scales, the light ground coloration partly visible only near apex; fascia nearly white, with a sharp basal and more blurred distal edge; cilia-line more or less distinct; cilia Female. Forewing length 2.5-2.7 mm, wingspan 5.5-6.0 mm. Antenna with 28-31 segments. Th e coloration as in male, hindwing without androconial scales.
Biology. Hostplant. Adults have always been collected near Quercus ilex L. or Q. rotundifolia, repeatedly accompanied by vacated mines on these evergreen oaks; these are thus the most likely hostplants. Leafmines (Fig. 66). Described from vacated mines found in association with adults. Egg position not observed. Mine starts as a narrow sinuous gallery which changes later into a large blotch with two lateral bands of frass, similar to those of other species in the suberis complex).
Larva. Unknown. Life history. Adults were collected from late May to early August. Distribution. Iberian Peninsula: Portugal and Spain. Etymology. Th e name phaeolepis: is derived from Greek phaios, dusky, brown and lepis, scale: with brown scales, referring to the androconial scales on the hindwing; it should be treated as a noun in apposition.
Remark. Th e specimen from Portugal, Serra de Monte Figo was earlier misidentifi ed as E. caradjai . (Bedell) Figs 70,76,85,119 Diagnosis. Ectoedemia quinquella is very similar to E. algeriensis and E. coscoja, but has a dark thorax, at least the anterior part (Fig. 70). E. coscoja also diff ers in having a pale head. Male genitalia (Fig. 76) similar to those of E. coscoja, valva tip less pronounced; female genitalia ( Fig. 85) (Căpuşe and Kovacs 1987) needs to be confi rmed, the record from Croatia (Laštůvka and Laštůvka 1997) was a mistake. Th e records from Italy by Jäckh ( Van Nieukerken 1985) based on illustrations only, are here confi rmed by the study of the original specimens (see below). Th e species is expanding northwards in Belgium and the Netherlands: Alders 2006;Huisman et al. 2001;Van Nieukerken 2006; Remarks. Th e character to distinguish E. quinquella from E. algeriensis, the partly darker thorax (with white tegulae and posterior tip of thorax) is problematical in the Italian specimens which have an almost completely white thorax. Still, most specimens retain at least one row of black scales along the anterior margin of the thorax and the dark scaling of the forewings is much darker (more black) than in algeriensis. In the 1985 description the fi gures for aedeagus length were wrong, they should read: 250-270 μm.  3♂, 2♀, Savona, Andora, Conna, S. Sebastiano, 500 m, 4.VII.1969, li, E. Jäckh (USNM); 1♂, 1♀, Savona, Testico, 470 m, 5.VII.1969, E. Jäckh (USNM). Van Nieukerken Figs 71,79,86,88 Diagnosis. Resembles E. quinquella and E. coscoja, which both have a similar pattern of three spots, but the spots are usually less distinct in algeriensis. E. coscoja is separated by the pale head, and E. quinquella by the partially black thorax (anterior part), which is completely white in algeriensis, and the yellow hairpencil in quinquella, versus the white one in algeriensis. E. algeriensis has also many more antennal segments than the other two species.

Ectoedemia (Ectoedemia) algeriensis
Redescription. Male (Fig. 71). Forewing length 2.4-2.9 mm, wingspan 5.3-5.7 mm. Head: frontal tuft almost black, collar slightly paler; scape white; antenna brown, with 48-54 segments. Th orax and tegulae completely white. Forewing brown, with pattern of two to three indistinct white spots, one at 1/3 at costa, one at dorsal margin slightly beyond middle and the third at 2/3 in disc, usually not touching costa, often reduced or almost absent; sometimes a small indistinct basal spot present; cilia-line distinct; cilia silvery white; underside light brown-grey. Hindwing and cilia dark grey; a white hairpencil, surrounded by yellow scales. Abdomen grey, with white anal tufts.
Leafmine. Egg on leaf upperside. Mine a highly contorted gallery, often following veins partially, fi lled with black frass. Not separable from other gallery mines of E. haraldi, ilicis or heringella.
Larva. According to original description the larva is green. However, while collecting the French material, EJvN did not note any green colour and larvae were tentatively identifi ed as ilicis/haraldi. Since the latter have normally whitish larvae, the original description is probably incorrect, or refers to the colour as appearing inside the mine. No ventral plates have been observed.
Life history. Probably univoltine, adults on the wing in June and July, larvae found in February-March in France, April in Algeria (high mountains).
Distribution. France, Mediterranean area: Van Nieukerken et al. 2006. Otherwise only known from the Atlas mountains in Algeria and Morocco. Van     Diagnosis. Forewing with three white spots as in E. quinquella and E. algeriensis, but separated from both by pale frontal tuft, which is black in the other two. Th e thorax is completely white as in algeriensis, the spots are almost touching the wing margins, whereas the distal spot in quinquella is usually separate. Th e male genitalia are very similar to those of algeriensis, but diff er by the shorter valvae with a more pointed apex. Th e female genitalia have very characteristic terminalia with very large setose anal papillae with many setae.

Ectoedemia coscoja
Description. Male (Fig. 72). Forewing length 2.0-2.2 mm, wingspan 4.3-5.1 mm. Head: frontal tuft yellowish ochreous to ferruginous, collar paler; scape white; antenna brown, with 32-35 segments. Th orax completely dull white. Forewing dark brown, with indistinct pale basal streak and three distinct white spots, one at 1/3 near costa, usually separated from costa by some brown scales, one at dorsal margin in middle and the third at 2/3 near costa, also often separate; cilia-line distinct; cilia silvery white; underside light brown-grey. Hindwing and cilia dark grey; an short ochreous hairpencil; underside ochreous grey. Abdomen grey, with white anal tufts.
Biology. Hostplant. Quercus coccifera L. Leafmines (Fig. 67). Egg deposited on leaf upperside. Mine a sinuate or contorted gallery, often following veins, slightly and continuously enlarging in its second part, fi lled with brown to black frass. Not separable from other gallery mines that occur sympatrically: E. haraldi and ilicis.
Larva. Whitish, no chain of ventral plates. Life history. Larvae collected in January and February. Probably univoltine, adults on wings in June and July.
Distribution. Spain: Catalunya, Aragon, Andalusia, most records not far from the coast.
Etymology. Th e name coscoja is the Spanish vernacular word for Quercus coccifera, referring to the hostplant and distribution. To be pronounced in the Spanish way as "koskokha". A noun in apposition.
Remarks. One female paratype from Marbella (1984) was treated previously under the remarks of E. algeriensis (Van Nieukerken 1985: 45) as a specimen with uncertain identity. (Klimesch) Biology. Hostplant: Quercus cerris. Univoltine, larvae very late in the season, mid October to late November, adults fl y in June.
Distribution. Here recorded new for Morocco, Greece: Crete and Turkey. Apparently the most widespread Ectoedemia species feeding on evergreen oaks.
Distribution. Only in Southwestern Europe and Northwest Africa (Van Nieukerken 1985; Van Nieukerken et al. 2004a; in ilicis. Female E. pseudoilicis can be easily recognised by the protruding ovipositor. Male genitalia similar to those of E. ilicis and heringella, but ilicis has the gnathos undivided, albeit with a serrate margin and the valvae of pseudoilicis are more elongate, tip less curved. E. heringella easily separated by the androconial scales on the hindwing. Redescription. Male. Forewing length 1.8-2.5 mm, wingspan 3.8-5.4 mm. Head: frontal tuft yellow to ferruginous , collar paler; scape white, sometimes with a few brown scales; antenna brown, with (32) 35-41 segments. Th orax and forewing dark brown with single dorsal white spot, slightly beyond middle, of varying size; cilialine distinct; cilia silvery white; underside pale brown. Hindwing and cilia dark grey; costal bristles present, no hairpencil. Abdomen grey, with brown anal tufts.
Biology. Hostplant: Quercus coccifera L., here confi rmed by reared adults. Leafmine. Egg on leaf upperside. Mine a more or less contorted gallery, almost fi lled with black frass. In the Greek localities almost always occurring together with E. haraldi; it has been impossible to separate the mines from these mixed samples.
Univoltine, larvae collected in February, adults fl ying in April to July. Distribution. Widespread in mainland Greece, also newly recorded here from Greece: Crete and Turkey. Apparently vicariant with the western Mediterranean Ectoedemia ilicis.
Univoltine, larvae from November to April, adults May (April?) to July. Distribution. Widespread in Eastern Mediterranean, from Cyprus westwards to Corsica and southeastern France (Alpes Maritimes), recently recorded as invasive insect from Great Britain: Langmaid and Young 2003;Prichard 2005;Vickery 2004;Palmer 2008;Sims 2008. Records for France: Van Nieukerken et al. 2006Greece: Laštůvka and Laštůvka 1998. Here new for Sardinia and tentatively recorded for Tunisia on the basis of two females.
Remarks. Ectoedemia heringella suddenly appeared in the beginning of this century in the parks of London, where it soon reached almost pest status with sometimes all leaves of the planted Quercus ilex covered in mines (Fig. 116). Th e fi rst adults were collected in the Wildlife Garden of the Natural History Museum (Langmaid and Young 2003;Prichard 2005;Vickery 2004), from where EJvN identifi ed the unrecognised specimens in 2000. It is likely that the species was introduced to Britain, possibly with plant material, the nearest natural occurrence being the Alpes Maritimes in France.
Recently it has been spreading through southern England. EJvN found similar vacated mines in Paris: Jardin des Plantes, but it is not certain that they belong to this species.
Male genitalia (Figs 81, 103). Capsule length 230-310 μm. Vinculum slightly concave anteriorly. Tegumen distinctly produced in a hemispheric pseuduncus, with simple long setae. Gnathos with central element short, with blunt, truncate tip, not separated in two parts; lateral processes long. Valva length 175-220 μm, broad with distal process pointed and strongly curved inwards, outer margin slightly bulged, inner margin in middle distinctly convex, with long setae on dorsal surface and on caudal part, sublateral processes more than one third transtilla length. Aedeagus 250-300 μm long, with distinct cathrema, numerous very small cornuti in vesica, carinae simple, elongate.
Leafmines (Fig. 118). Egg usually on leaf upperside, on underside in Cyprus specimen. Leafmine a much contorted gallery, usually confi ned to small part of the leaf; mine following short straight courses and often turns suddenly ca. 90 degrees or more, often doubling back, in this way using up the majority of a small portion of the leaf; last part of the mine a wide gallery, wider than larva, resembling a blotch; frass black and almost completely fi lling the narrow gallery part, a bit more dispersed in fi nal part.
In Turkey similar to more or less sympatric E. haraldi, but E. alnifoliae mines tend to be more contorted and angulated than those of haraldi. Larva. White with greenish or yellowish tinge, brown head and prothoracic shield, distinct row of ventral ganglia. In earlier instars row of ventral plates absent.
Life history. Larvae found in early March, adults found in second half of April and May, in indoor breedings emerging from March. Most likely univoltine.
Distribution. Cyprus, new records for Turkey (south coast and east, Hakkari) and Greece: Samos. Apparently widespread in eastern Mediterranean region.
Remarks. Th e species was originally described on the basis of a single female from Cyprus. Th e fi nding of males in Cyprus and the reared series from Turkey has made it possible to redescribe it fully here.
Material  (Caradja) Fig. 120 Trifurcula rufi frontella Caradja 1920: 161. Nepticula nigrosparsella Klimesch 1940 [synonymised by . Ectoedemia (Ectoedemia) rufi frontella;  Remarks. Mines and larvae may be confused with those of E. quinquella (Fig. 119) and pubescivora, which both also have black plates in the early instars, but rufi frontella has characteristic broken frass in early mine, the larva has a strong greenish tinge (Fig. 120); the mine of E. pubescivora usually runs along veins (Fig. 121) Fig. 123 Biology. Hostplants: mainly evergreen Rubus species: mainly R. ulmifolius Schott. and R. sanctus Schreb. (sometimes regarded as synonyms). Larvae found from autumn throughout winter, September to April, again in July; adults in April to July. In the northern part of its range certainly univoltine, possibly with more generations in parts of the mediterranean.   (Sokolov et al. 1980). Leafmines may be confused with those of Stigmella inopinata A. & Z. Laštůvka (Laštůvka and Laštůvka 1990a), they only diff er in the more jagged early mine of spiraeae versus a straight one in S. inopinata. Univoltine, larvae from July to October, they appear in some years (e.g. 1990) in July after a cold spell, in other years not before September or October (Laštůvka and Laštůvka 1991). Adults from May to July. Distribution. In Europe only known from a very small area near the border of Hungary and Slovakia (see also Laštůvka and Laštůvka 1991), widespread in East Palaearctic. A new record from western Siberia is given here (Fig. 108).

Ectoedemia (Ectoedemia) erythrogenella (Joannis)
Remarks. Puplesis (1994) synonymised the Siberian E. jacutica with E. agrimoniae. However, study of one of the specimens shows that it is in fact identical to E. spiraeae, which is also clear from the original genitalia drawing (absence of carinal processes, shape gnathos, valva). Th is also fi ts the distribution, E. spiraeae being known from Japan and China as well [unpublished data EJvN]  at least two generations in Croatia, where the larvae of the second generation leave the mine before pupation. Larvae: June-July, August-October. Mines can be found much later in the season, with cocoons inside. Adults in April to June and again July-August.

Ectoedemia (Ectoedemia) hexapetalae (Szőcs)
Biology. Hostplant: Filipendula vulgaris Moench. Th e record in France is from an area where this host is unknown (Dupont 1990), so that it probably feeds there on another host, possibly another rosaceous herb. Larvae in June, July and August-September, adults in May and again late June to July.
Distribution. France: Van Nieukerken et al. 2006. Otherwise still only known from eastern Austria and central Hungary.

Ectoedemia (Ectoedemia) sp. n.
Ectoedemia sp. n.? Bengtsson et al. 2008: 273. Th is species, related to E. hexapetalae, was found in Norway. Additional fi ndings in France prove its separate identity, it will be described elsewhere.

Ectoedemia (Ectoedemia) rubivora (Wocke)
Biology. Hostplants: Rubus species, particularly on R. caesius L., R. saxatilis L., R. chamaemorus L. and R. arcticus L., but also on forms of the fruticosus L. complex. In the Netherlands we found several times vacated mines resembling those of rubivora on Agrimonia eupatoria in areas where E. rubivora is abundant on Rubus. Th ese mines are very diff erent from those of E. agrimoniae, which is not known from the Netherlands. Univoltine, larvae from late August until Octorber, adults in June and July.

Ectoedemia (Ectoedemia) spinosella (Joannis) Figs 109, 110
Ectoedemia (Ectoedemia) albiformae Puplesis & Diškus, 2003: 186 . In most of Europe P. spinosa is the main host, in Greece the almonds P. dulcis and P. webbii. On most other hosts confusion with mines of E. mahalebella is possible. Univoltine in much of its area, in Greece bivoltine. Larvae in June (Greece) and again July to November, depending on latitude and season. Adults fl y in May-July. Distribution. Russia: Van Nieukerken et al. 2004b; Belgium: Van Nieukerken 2006;Switzerland: Sauter 1983. Also recorded from Turkmenistan: Kopet Dag . A former record from Lithuania (dot on map, Van Nieukerken 1985) was a mistake. Th e record from France (26): Les Prés ( Van Nieukerken et al. 2006) wa misidentifi cation for Ectoedemia mahalebella.
Remarks: E. albiformae was described from Turkmenistan, Western Kopet Dag. Th e authors compared it with E. petrosa Pupesis, 1988, but (after studying a male paratype, see Figs 109, 110) we cannot separate it from E. spinosella in any detail, apart from the white colour of the androconial scales that surround the brown hairpencil, which are brown in normal spinosella. We also once found this form in Greece, and since E. spinosella also occurs in the western Kopet Dag , we assume albiformae is most likely a colour aberration of spinosella, and hence synonymise it here. A female paratype of albiformae turned out to be an unknown Stigmella species.
Material . Van Nieukerken et al. 2006 for records in France and Italy.

Discussion
State of knowledge. In 1985 ( Van Nieukerken 1985), 50 species were recognised for the West Palaearctic, eight in Zimmermannia and 42 in Ectoedemia s. str., including two unnamed species. One of the Zimmermannia's is now regarded as east Palaearctic, so that 49 (7 + 42) remain. With the new descriptions since and in this paper, in 20 years the number has grown to 57 (9 + 48), an increase of 17% (eight species). Although a considerable part of Europe and West Asia have been better investigated in the last decades, there are still many gaps in our knowledge which may lead to new discoveries and a better knowledge of biology and distribution. Whereas we have many data from the Iberian Peninsula by light collecting in summer and mine collecting in winter, the "normal" autumn fauna of the deciduous forests remains virtually unstudied. One autumn visit could reveal several new records. Much less known are the potentially rich areas of Turkey, the Caucasus and northern Iran and other parts of the Middle East, which probably will harbour several undescribed Ectoedemia species; also northern Africa is still poorly investigated. In addition, detailed molecular studies can reveal hidden species and complex cases of evolution.
Range expansion. Some species seem to have been collected much more frequently in recent years, giving the impression of range expansion. Particularly for southern European species it is logical to attribute this largely to an enormous increase of collecting eff ort: very little collecting has taken place there before the seventies. Especially bark-mining species in the subgenus Zimmermannia, that are almost only known from light collected specimens, have been collected frequently in recent years: this is apparent from the numerous new records of E. amani and E. reichli. However, there is little reason to attribute this to range expansion. Table 2 shows the increase of records of Zimmermannia in the database of material studied by Van Nieukerken (including records contributed by the co-authors), the total being 483 (one record is defi ned as the catching of any number of specimens of a single species on one date or date range on a single locality). Th e proportion of E. amani records varies between 6 and 8.5% from the seventies to the present decade. Since a few records are available from earlier decades, these data do not show a signifi cant change of the proportion of this species in the Zimmermannia records, thus the enormous increase of collecting eff ort is most likely the best explanation for the numerous new records recently.
Th e fi nding of the new species E. heckfordi in the south of England is very surprising, and could indicate a real range expansion of a species hitherto hidden in Western France, where there has hardly been collecting activity. On the other hand, such a species might also have been overlooked in the few small forests where it has been found to date.
Th e range expansion of E. heringella in Britain since the late nineties is obvious, but almost certainly aided by man. In the Netherlands we have seen a range expansion of E. quinquella, which appeared in the eighties in the far south, in this century it was found more to the north and since 2007 also in the coastal dunes. Th is expansion is likely to have been the result of the changing weather or climate: E. quinquella has a Mediterranean-atlantic distribution type, indicating that it is probably susceptible to cold winters. Another species that shows range expansion is E. hannoverella which turned up in Sweden in the eighties and in Britain in 2002. Th e range expansion of poplar feeders (E. hannoverella and turbidella) may be related to extensive planting of poplars now all over Europe.  Spain) and Andreas Werno (Landsweiler-Reden, Germany). Sjaak Koster (Losser, Netherlands) is acknowledged for carefully making many preparations of genitalia. Kees van den Berg (Naturalis) assisted with rearings and in many other ways. Camiel Doorenweerd and Frank Stokvis (both Naturalis) provided many molecular data that supported many decisions and identifi cations in this treatment. Camiel and Kees are also thanked for their company and work during joint collecting trips. Ian Th irwell (Portsmouth, Great Britain) kindly allowed us to reproduce his photograph of a larva of E. heckfordi. Robert Hoare (Auckland, New Zealand) kindly commented on the manuscript. Th omas Simonsen and an anonymous referee are acknowledged for their constructive remarks. Many data in France and Italy were obtained during the EU funded EDIT WP 7 project "All Taxa Biodiversity Inventories in the Mercantour/Alpi Marittime natural parks". Th is study was partly supported by research grant No. MSM6215648905 of the Ministry of Education, Youth and Sports of the Czech Republic to the junior authors.