The Dilaridae of the Balkan Peninsula and of Anatolia (Insecta, Neuropterida, Neuroptera)

Basing upon all available information on type material of Dilar turcicus Hagen, 1858, Dilar syriacus Navás, 1909, and Dilar lineolatus Navás, 1909, together with a large number of dilarid specimens, the pleasing lacewings of Anatolia and Southeast Europe are revised. The current taxonomic concept of D. turcicus is conirmed, and a lectotype is designated. Dilar turcicus is widely distributed in the southeast of Europe (being the only representative of Dilaridae in this region), in Anatolia, and, most probably, in the Caucasus region. Dilar syriacus and D. lineolatus remain nomina dubia. Dilar syriacus might occur in Anatolia, while D. lineolatus is a species occurring in western Central Asia. Two new species, Dilar anatolicus sp. n. and Dilar fuscus sp. n. are described from Anatolia. Wings and genital segments of the three species occurring in Anatolia are illustrated, and a map documenting the known distribution of these species is provided.


Introduction
The Dilaridae is a small family of the order Neuroptera comprising about 100 described valid species worldwide.Most species -all assigned to the genus Dilar Rambur, 1838 -have been recorded in the Northern Hemisphere, and particularly in Central, East and Southeast Asia (Aspöck et al. 2001, Oswald and Schiff 2001, New 2003, Zhang et al. 2014a, b, c, 2015), but a few species representing other genera, i.e.Nallachius Navás, 1909, Neonallachius Nakahara, 1963, and Berothella Banks, 1934, have been found in North and South America as well as in South and Southeast Asia, respectively (New 1989, 2003, Oswald 1998), and one species has been described from South Africa (Minter 1986).
In the Western Palaearctic most species of Dilar have been found in the southwest of Europe in the Iberian Peninsula (Aspöck et al. 2001, Monserrat and Triviño 2013, Monserrat 2014).Only one species -Dilar turcicus Hagen, 1858 -has so far been known to occur in the Balkan Peninsula (Aspöck et al. 1980(Aspöck et al. , 2001)).This species has also been recorded from various parts of Anatolia.Until now no proven records of other species of Dilaridae in Turkey have been published.
For many years we have, however, been aware of the existence of at least two further Dilar species which had been collected by two of us (H.A., U.A.) in Anatolia.However, the fact that the type material of D. turcicus had never been examined and the possibility of the occurrence of other described but not identiiable species (nomina dubia) prevented us from describing our unidentiied species.
In the course of recent studies on Dilaridae of Asia, we have again examined the situation and have tried to clarify all open questions as far as possible, so that we can now present a summarising overview on the Dilaridae of Anatolia and of Southeast Europe.Genitalic preparations were made by clearing the apex of the abdomen in a cold, saturated KOH solution for 3-4 h.After rinsing the KOH with acetic acid and water, the apex of the abdomen was transferred to glycerine for further dissection and examination.Habitus photos were taken by using Nikon D60 and D800 digital camera with Nikon MI-CRO NIKKOR 105 mm lens, and the genitalic igures of the new species were made by hand drawing under Leica M165C stereo microscope.The genitalia of the lectotype of D. turcicus were photographed with a Leica DFC camera attached to a Leica MZ16 binocular microscope and processed with the help of Leica Application Suite.They were then stacked with ZereneStacker 64-bit and processed with Adobe Photoshop Elements 8.The terminology of the genitalia generally follows U. Aspöck and H. Aspöck (2008).

Material and methods
Distribution map: Localities were taken from original literature (Suppl.material 1) and, together with the specimens examined by us, listed continuously with a number for each locality in MS Excel 2010 (Suppl.material 2); geographic coordinates introduced by us are in italics.

Note.
A description of the genus has been given on several occasions in recent publications (Zhang et al. 2014a(Zhang et al. , b, c, 2015)).Remarks.After the original description by Hagen (1858) based upon two specimens, one from Turkey and one from Armenia, the species was redescribed, characterised and/or illustrated on several occasions (e.g.Navás 1909, 1914, Aspöck et al. 1980), but none of the two syntypes has ever been examined.When we realized that Anatolia harbours at least one further species of Dilar which could not be differentiated eidonomically from D. turcicus, an examination of the types became necessary.Both syntypes still exist.The syntype from Turkey ("Türkei", Smyrna [=Izmir] written on the original collecting label) is preserved in MFN in Berlin.Unfortunately, this specimen consists only of fragments of the wings (Fig. 8), which do not allow a deinite identiication.Therefore, we tried to examine the other syntype, which is from Armenia ("Armenien", without further details on the locality).This syntype is deposited in the MCZ 1 of the Harvard University in Cambridge and was kindly sent to us for examination.The specimen is a male in a fairly good condition (Figs 1-7), although it lacks the left fore-and hindwings.In particular, the abdomen is present so that we could study the genital segments.They agree perfectly with the genital segments of that species which has been regarded as D. turcicus Hagen until now (see Fig. 439 in Aspöck et al. 1980).We herewith designate the syntype from "Armenia" as the lectotype of Dilar turcicus Hagen, 1858.
D. turcicus shows a considerable variation in size and coloration of wings.Lengths of forewings vary from 7.5 to 13.5 mm.A reliable identiication seems -at least presently -only possible on the basis of the male genital segments.
Distribution.As far as we know the southeast of Europe does not harbour any other species besides D. turcicus.We have examined more than 200 specimens from various localities on the Balkan Peninsula and have not 1 Hermann August Hagen (1817Hagen ( -1893) ) moved from Germany to Cambridge, Mass., USA in 1867, where he became the irst professor of entomology in the USA.He took (at least parts of) his collection with him to the Museum of Comparative Zoology, where it is still preserved.found any hint for the occurrence of a second Dilar species.Thus, it is a reasonable assumption that all published records of D. turcicus from localities situated in Europe really refer to this species.We have therefore included all published European records of D. turcicus in the map (Fig. 31).All records outside Europe (mainland) are based upon specimens examined by us.The documented distribution (Suppl.materials 1 and 2) comprises: Croatia, Bosnia-Herzegovina, Montenegro, Kosovo, Macedonia, Albania, Greece including several islands (Corfu, Chios, Skopelos, Samothraki, Thasos, Lesbos), Bulgaria, Ukraine, Russia, Turkey (Aspöck et al. 1980(Aspöck et al. , 2001;;Devetak 1991Devetak , 1992a, b;, b;Devetak et al. 2015;Ghilarov 1962;Pongrácz 1913;Popov 1964Popov , 1993Popov , 2001Popov , 2002Popov , 2004Popov , 2007;;Zakharenko 1982Zakharenko , 1988;;Zakharenko and Krivokhatsky 1993; specimens examined in the course of this study are listed in Suppl.material 2).
The exact type locality of D. turcicus remains unknown.In the middle of the 19 th century the name Armenia was used also for regions in the east of Anatolia.Whether D. turcicus occurs in Armenia of present time is not sure.However, Ghilarov (1962) described the larva of D. turcicus from the Caucasus region, not far away from the proven record of D. turcicus in the Crimean peninsula.Thus, one can hardly doubt that D. turcicus occurs in present day Armenia.There are also several published records of D. turcicus in Anato-lia (Kiyak andÖzdikmen 1993, Canbulat andKiyak 2005, O.B. Kovanci andB. Kovanci 2015, Canbulat 2007).However, these records could not be checked by us and have therefore been included in the distribution map as Dilar sp.
Dilar corcyraeus has already been synonymised 35 years ago (Aspöck et al. 1980).It was described on the basis of a single female, which is deposited in the NHMW, Vienna.The type locality is the Greek island of Corfu, from where also D. turcicus has been recorded (Aspöck et al. 1980).Navás (1909) himself had already considered that it might be a variety of D. turcicus.Here we provide a photograph of the type of D. corcyraeus (Fig. 9).
Biology.The biology of D. turcicus is poorly known, with only a few scattered observation records (Ghilarov 1962;Aspöck et al. 1980;New 1986).Adults are usually found on shady places with rich low vegetation, bushes and trees, frequently near stone walls with numerous crevices where the larvae may pupate after their development in the soil.The number of larval instars is unknown, but up to 12 moults have been observed under experimental conditions (New 1986).However, there is no indication that the number surmounts the three instars known from Neuroptera under natural conditions.Duration of development from egg to adult is probably at least one year.Adults are mainly active after sunset and are attracted by artiicial light.Dilar turcicus has a considerably   large vertical distribution, with records ranging from few meters up to 2100 m above sea level.
Remarks.This species has been described based on a single male specimen without abdomen deposited in the MNHN, Paris.According to the description of Navás (1909), the specimen resembles D. turcicus.As the type lacks the abdomen, it cannot be identiied reliably.Therefore, Monserrat (1988) has declared Dilar syriacus as a nomen dubium, which was conirmed by Legrand and Lachaise (1994).The type was collected most probably near Beirut.Perhaps it will be possible to clarify the species as soon as one can collect there and ind out which species occur in the surroundings of Beirut.
Remarks.This species was described on the basis of a specimen from "Turkmenien (Tekke)".Turkmenien is the old German word for that part of western Central Asia which is largely identical with the present Turkmenistan.
In his description of the species Navás (1909) translated "Turkmenien" as "Turcomania".Later "Turcomania" was erroneously interpreted as Turkey by Oswald (1998)."Tekke" is a frequent geographic name also in Anatolia, and thus Oswald The coloration of the wings and of the wing venation is rather dark and characteristic so that it may be possible to clarify the species.However, at present D. lineolatus has to be regarded as a nomen dubium, because due to the lack of the genitalia a clariication of the taxonomic status of this species is not possible.Diagnosis.This species is characterized by the forewings with numerous pale brown spots and by the male gonocoxite complex 9, 10 and 11 with short, feebly sclerotized ninth gonocoxite and elongate, blade-like tenth gonocoxite.
Description.Male.Length of forewing 11.0-12.2mm, of hindwing 9.5-10.3mm.Head pale yellowish brown, with pale yellow setose tubercles.Compound eyes blackish brown.Antenna with ca.29 segments, pale yellow, with scape and pedicel pale brown, lagellum unipectinate on most lagellomeres, medial branches much longer than those on both ends, longest branch nearly 3.0 times as long as the corresponding lagellomere, but branch of 1st lagellomere short and dentate, distal eight lagellomeres simple.
Prothorax pale yellowish brown, pronotum dark brown with several hairy yellowish tubercles; meso-and metathorax yellow, each notum laterally with a pair of broad, brown markings and a pair of pale brown narrow stripes; mesonotum medially with a pair of additional brown markings near anterior margin.Legs pale yellow, with femora blackish brown at tip.Wings hyaline, slightly yellowish brown, with numerous pale brown spots.Forewing ~2.5 times as long as wide, densely spotted, with markings slightly darkened toward anal region; a few markings on posteroproximal area connected with each other, forming short transverse stripes; an immaculate area present distal to median nygma; two nygmata present on proximal portion, both slightly smaller than median nygma.Hindwing ~2.1 times as long as wide, slightly paler than forewing, with dark markings largely reduced; one nygma present at middle.Veins pale yellowish brown; Rs with four main branches on both fore-and hindwings.
Abdomen pale yellowish brown, pregenital segments dorsally brown.Ninth tergite in dorsal view with nearly anterior margin and a deeply V-shaped posterior incision, leaving a narrow median portion and a pair of subtriangular hemitergites, which are obtuse and slightly incurved distally; in lateral view broad, with straight ventral margin and arcuate posterior margin.Ninth sternite obviously shorter than ninth tergite, with nearly truncate posterior margin.Ectoproct in dorsal view with a pair of digitiform projections, posteroventrally with a pair of subsemicircular and lattened lobes, a pair of biid unguiform projections and a pair of short, feebly sclerotized, digitiform projections.Ninth gonocoxite short, shield-like, mostly membranous and transparent, with sclerotized lateral margin and with a strongly narrowed, obtuse apex; tenth gonocoxite slenderly elongate, blade-like, nearly 2.5 times as long as ninth gonocoxite, straightly directed posteriorly with median portion feebly curved; fused eleventh gonocoxites (= gonarcus) beamshaped, laterally connectedto base of ninth gonocoxites.Hypandrium internum subtriangular, with lateral margins slightly arcuate.
Seventh sternite in lateral view subtrapezoidal.Eighth abdominal segment without subgenitale.Basal part of bursa copulatrix sac-like, nearly rhomboid in ventral view, proximally membranous and strongly rugose, distally sclerotized into a pair of lateral sclerites and a median ridge.Ectoproct rather small, ovoid.Distribution.Turkey: Western, central, and southern parts of Anatolia (Fig. 31); supposedly reaching Syria or Lebanon to the south.
Etymology.Adjective, masculine, nominative, singular; an attribute to the genus name.From lat. anatolicus 3 = referring or belonging to Anatolia, the Asian part of Turkey.
Remarks.Dilar anatolicus sp.n. can be distinguished from all other western Palaearctic species of Dilar based on the slenderly elongate male ninth gonocoxites and the short, largely membranous male tenth gonocoxites.In appearance, D. anatolicus sp.n. looks similar to its co-existing Turkish species D. turcicus, but it can be easily distinguished from the latter by the male ninth tergite without median projection and by the different shape of sclerotization of the female bursa copulatrix.Interestingly, D. anatolicus sp.n. appears to be closely related to D. sinicus Nakahara, 1957, which is distributed in northern China (Zhang et al. 2014a), by having the similar male gonocoxites 9, 10 and 11 with short, shieldlike ninth gonocoxite and slenderly elongate tenth gonocoxite.However, in D. anatolicus sp.n. the male ninth gonocoxite is largely membranous and transparent with obtuse apex, while in D. sinicus the male ninth gonocoxite is entirely sclerotized with acutely pointed apex.Possibly D. anatolicus sp.n. occurs sympatrically with D. syriacus (see under D. syriacus).Diagnosis.This species is characterized by the entirely brown wings without distinct markings (at least in the female).
Head brown, with three pale yellow setose tubercles; vertex medially with a blackish brown vittae.Compound eyes blackish brown.Antenna with ca. 25 segments, pale brown.
Prothorax blackish brown, pronotum with pale yellow anterior margin and several yellowish tubercles clothed with brownish hairs; mesothorax and metathorax blackish brown, dorsally yellow at middle.Legs yellowish brown, with each segment blackish brown at tip.Wings entirely brown, without distinct markings.Forewing ~2.5 times as long as wide, with distal half slightly paler; proximally with two or three nygmata, medially with one nygma, which is slightly larger than proximal ones.Hindwing slightly paler than forewing; one nygma present at middle.Veins brown; Rs with six main branches on both fore-and hindwings.Distribution.The only reliable record of D. fuscus sp.n. is from the type locality in western Anatolia (Fig. 31).
Etymology.Adjective, masculine, nominative, singular; an attribute to the genus name.From lat. fuscus 3 = dark.The name refers to the dark coloration of the wings.Remarks.Despite the unknown male, D. fuscus sp.n. is a spectacular new species and can be easily distinguished from all other species of Dilar based on the entirely brown wings without distinct markings.The female genitalia of D. fuscus sp.n. are principally similar to the co-existing species D. turcicus based on the basal part of bursa copulatrix with lateral portions slightly sclerotized.It can, of course, not be excluded that the hitherto unknown male has a different coloration of the wing membrane or even a patterned wing membrane.Such kind of sexual dimorphism has been known from other Dilaridae (Zhang et al. 2015).

Discussion
Three species of the genus Dilar, i.e.D. turcicus, D. anatolicus sp.n. and D. fuscus sp.n., are now known from the Balkan Peninsula and Anatolia, with clear speciic identities.In the southeast of Europe only D. turcicus occurs, while in Anatolia all these three species are present.Dilar fuscus sp.n. is known in the female only, which can easily be identiied by the unusual dark coloration of the wings (see Figs 26-27).The other two species -D.turcicus and D. anatolicus sp.n. -can easily be differentiated on the basis of morphological characters of the male genitalia, usually even in dried specimens.Dilar turcicus is equipped with a conspicuous median dorsal processus on male ninth tergite, whereas D. anatolicus sp.n. lacks this processus.A reliable differentiation of these two species on the basis of coloration and wing marking patterns isat least presently -not possible.
D. turcicus is a Pontomediterranean faunal element with a distribution comprising large parts of Southeast Europe, probably all major parts of Anatolia, and most probably parts of the Caucasus region.To the best of our knowledge D. turcicus is the only representative of the family Dilaridae in the Balkan Peninsula and in the southern parts of East Europe.The northernmost records of this species in Europe are in Croatia, Bulgaria, Ukraine, and Russian north of Caucasus.Most probably the species had survived during the last glacial period in various refugial areas in the southeast of Europe and in Anatolia respectively.
In Anatolia D. anatolicus sp.n. and D. fuscus sp.n. have been known to us for long, but with respect to the unsolved questions concerning the described species they remained undescribed until now.After examination, re-examination and re-consideration of the three pertinent species -Dilar turcicus Hagen, Dilar syriacus Navás, and Dilar lineolatus Navás respectively -the situation has largely been clariied (concerning D. syriacus and D. lineolatus as far as possible at present).Dilar syriacus (type locality: probably surroundings of Beirut) remains a nomen dubium, it cannot be excluded that it will turn out to be identical either with D. turcicus (then it would be a junior synonym) or with D. anatolicus sp.n. (then it would become a senior synonym).However, it may also be a species different from both.
(1998)  decided to choose a locality named "Tekke" situated in Turkey as the type locality of D. lineolatus.He even gave the geographic coordinates: 40°09'N/29°41'E, which is certainly wrong.Possibly,Oswald's (1998) interpretation gave rise to the subsequent records in Anatolia(Canbulat and Kiyak 2005).We do not think that D. lineolatus really occurs in Turkey.The type locality of D. lineolatus should be somewhere far more in the east.The holotype of D. lineolatus (Figs 11-13) is in poor condition.It is a female, but it lacks the abdomen.Only one forewing and the hindwings are present.We (H.A. & U.A.) examined this specimen already in 1967 (H.Aspöck and U. Aspöck 1968) and our opinion on this species of that time can now be conirmed: D. lineolatus is a valid and distinct species and cannot be assigned to any of the species recorded from Western and/or Central Asia.