Review of the European Greenomyia Brunetti (Diptera, Mycetophilidae) with new descriptions of females

Abstract The females of the four continental Greenomyia Brunetti species in Europe are associated with the males, diagnosed and keyed, providing the first association and description of the females of Greenomyia baikalica Zaitzev, 1994 and Greenomyia stackelbergi Zaitzev, 1982. Colour photographs of their habitus and line drawings of their female terminalia are provided. Greenomyia mongolica Laštovka & Matile, 1974 is found to be a senior synonym of Greenomyia theresae Matile, 2002. syn. n. The diagnostic characters used to distinguish between Greenomyia and Neoclastobasis Ostroverkhova in keys did not hold up to a closer scrutiny and leave the status of Neoclastobasis as separate genus questionable.


Introduction
Th e genus Greenomyia Brunetti was erected to distinguish a single oriental species, Greenomyia nigricoxa Brunetti, 1912. Since then several species (mainly Holarctic) have been described and new combinations proposed. Laštovka and Matile (1974) and Chandler and Ribeiro (1995) characterised the genus and Matile (2002) provided a key to all 11 world species of Greenomyia. Species of Greenomyia are mostly dark coloured medium-sized fungus gnats with a typical wing venation similar to that of Leia Meigen, 1818 and allied genera where R 1 is notably shorter than the long and nearly horizontally aligned crossvein rm. Further diagnostic characters include the lateral ocelli being well separated from the eye margins and all veins reaching the wing margin (cf. Søli et al. 2000). Edwards (1925) introduced the tribe Leiini for Leia, Greenomyia and a number of other genera with an intermediate position between the subfamilies Sciophilinae and Mycetophilinae. Later the tribe has sometimes been given subfamily status (see review by Gammelmo 2004), but recent morphological and molecular studies have questioned its monophyly (e.g. Amorim and Rindal 2007;Rindal et al. 2009). Greenomyia appears most closely related to the genus Neoclastobasis Ostroverkhova, 1970, a genus that includes two eastern Palaearctic (N. sibirica Ostroverkhova, 1970 andN. kamijoi Sasakawa, 1964) and one European species (N. draskovitsae Matile, 1978). Th e genus Neoclastobasis is diagnosed by a prolonged apical palpal segment, the veins M 2 and CuA 1 terminating before the wing margin, and distinctive terminalia (Ostroverkhova 1970;Matile 1978;Zaitzev 1994;Søli et al. 2000).
While the males of most Greenomyia species are adequately illustrated and keyed (Zaitzev 1982(Zaitzev , 1994Laštovka and Matile 1974;Chandler and Ribeiro 1995;Matile 2002) the females of several species remain to be properly diagnosed and described. Th e current communication was initiated by the fi nding of three Greenomyia species from two localities 1 km apart in Vuollerim, Lule Lappmark in northern Sweden (Kjaerandsen et al. 2007). Th e material gave us the opportunity to associate and describe females of two species for the fi rst time. Th e shifted focus to females further revealed that the generic characters separating Neoclastobasis from Greenomyia do not hold, and highlights a need to re-evaluate their status as separate genera.

Material and methods
Material and collections from a wide range of Palaearctic sources were studied. Th e collecting methods, if known, are referred to in case by each specimen in the studied material section below. Th e following codens obtained from Evenhuis (2009) (Becker, 1908) [Endemic to the Canary Islands, not seen] 4 Last palpal segment elongated (Fig. 6). C terminating distinctly before apex of wing, making R 5 straight to slightly sinuate (Fig. 10). Cercus two-segmented but segments partly fused (Figs 14, 18)  Female. Th orax blackish, abdomen brown. Legs yellow, cx 3 with small dark markings basally. All trochanters yellow, with small black apicoventral spots. Tibiae with dense brown setae. Scape and pedicel dark yellow, fi rst fl agellomere basally yellowish, rest of fl agellum brown. Mouthparts pale yellow. Apical palpal segment 1.4-1.6 (n=2) times as long as penultimate segment. Wing with narrow preapical brownish band, gradually tapering towards hind margin. C terminating almost at apex of wing, R 5 distinctly arched (Fig. 9). Medial and cubital veins both reach wing margin, CuA 1 basally obsolete, brownish shade along posterior margin of CuA 2 . Terminalia brown, cercus one-segmented, yellow apically. Tergite IX larger than tergite VIII. Gonapophysis IX visible in lateral view, with wide pear-shaped medial incision apically. Sternite VIII deeply incised apicomedially and moderately emarginated basally.
Male. Coloration and other non-terminal characters including palpi similar to female. Apical palpal segment is 1.4 (n=1) times as long as penultimate segment.
Remarks. Zaitzev (1994) described this species from Siberian material (Buryat Republic). Subsequently only a few specimens have been recorded from Norway, Sweden, Finland and Russian Karelia (cf. Kjaerandsen et al. 2007;Fig. 29). Zaitzev (1994) fi gured the male terminalia whereas the female terminalia have not been fi gured earlier.
Using the key by Zaitzev (1994) the studied females run to G. baikalica and they are also morphologically conspecifi c with material of both sexes collected simultaneously in Russian Karelia (A. Polevoi, pers. comm.). Th e studied female specimens were collected in a boggy forest stand within a small (9 ha) protected remnant of semi-natural, mixed forest.   Female. Thorax brown to blackish. Abdomen entirely brown or first two segments slightly lighter. Legs pale to yellow, except cx 2 and cx 3 with dark markings basally and apically, all trochanters brown and f 3 brown, with lateral parts lighter to yellow. Tibiae with dense brown setae. Scape and pedicel dark yellow, flagellomeres brown. Mouthparts yellow. Apical palpal segment 5.0-5.4 (n=4) times as long as penultimate segment. Wing with broad preapical brownish band, reaching hind margin but gradually paler. C terminating distinctly before apex of wing, R 5 straight to slightly sinuate (Fig 10). M 2 and CuA 1 not reaching wing margin, CuA 1 basally obsolete or very weak. Terminalia brown; cercus yellow, two-segmented, apical segment small and partly fused with basal segment. Gonapophysis IX well sclerotized apically, visible in lateral view and with well developed narrow apical incision. Tergite VIII equal in size but slightly wider than tergite IX. Sternite VIII with medial incision apically and well emarginated basally.
Male. Coloration and other non-terminal characters including palpi similar to female. Apical palpal segment is 4.7-5.7 (n=4) times as long as penultimate segment.
Remarks. While studying the Swedish specimen from Vuollerim, it ran by the fi rst attempt using the key by Zaitzev (1994) to Neoclastobasis because of the extra long last palpal segment and M 2 and CuA 1 not reaching the wing margin. Th e colouration of the studied specimen is, however, diff erent and female terminalia lack strong spines on sternite VIII, being typical to all of the described Neoclastobasis species (Zaitzev 1982;JK and OK pers. obs 0.1 mm 0.1 mm 0.1 mm 0.1 mm covery of a Greenomyia female, with similar size and coloration as the male of G. borealis in the same Malaise trap sample from northern Greece (Kerkini Lake area) allowed a safe association of the sexes. Th e females from Sweden, Estonia and Kazakhstan were further found to be conspecifi c with the Greek material of both sexes. According to Chandler et al. (2006), a male specimen from Greek mainland (Vikos Aoos National Park) has mainly yellow coxae, while other European specimens of G. borealis have mainly dark coxae. Th is may represent an intraspecifi c variation, however, all specimens studied during the current investigation have coxae whitish yellow. Th e above-mentioned Estonian specimen represents the fi rst record of G. borealis from the country. Th e female specimen from Vuollerim was collected in the same garden as G. stackelbergi. G. borealis was previously known only with two 19th century fi ndings from southern Sweden. Diagnostic characters. Female. Th orax dark brown to blackish. Abdomen entirely blackish brown or fi rst three segments slightly paler laterally. cx 1 entirely yellow or darkened in basal half, cx 2 and cx 3 entirely dark brown to black. Fore trochanter yellow basally, brown apically. Mid and hind trochanters brown. f 1 and f 2 yellow, f 3 yellow with brown apical fi fth. Tibiae yellow, apically slightly darkened, with dense brown setae. Scape, pedicel and fl agellomeres brown. Mouthparts pale yellow. Apical palpal segment 1.8-2.2 (n=5) times as long as penultimate segment. Wing tip shaded on about apical third, with darkened area along fore margin. All veins reach wing margin, M 2 sometimes basally obsolete or very weak, A 1 ending close to, sometimes fused into base of CuA 2 . Terminalia brown. Cercus distinctly two-segmented, apical segment small, ovate. Gonapophysis IX membranous, widely protruding apically, not visible in lateral view. Tergite VIII larger than tergite IX. Sternite VIII apically with shallow medial incision, moderately emarginated basally.
Male. Coloration and other non-terminal characters including palpi similar to female. Apical palpal segment is 1.7-2.1 (n=5) times as long as penultimate segment.
Remarks. Th is species was originally described by Laštovka and Matile (1974) based on Mongolian material and subsequently widely recorded in Europe. Chandler (2005) did not include G. mongolica in the European list and assigned all records to G. theresae, a species described from northern Italy by Matile (2002). Careful comparisons of type material of both species at MNHN in Paris (independently undertaken by two of the authors, OK and JK; the holotype of G. mongolica deposited in the Hungarian Natural History Museum was not available for the study) did not indicate any substantial diff erences in their male terminalia. Th e minor diagnostic characters as indicated in the original description and illustrations by Matile (2002) are liable to diff erent angles of views only. Consequently we have come to the conclusion that G. theresae at present state of knowledge must be treated as a junior synonym of G. mongolica and that all published records in Europe should rather be associated with the latter. In addition to the studied type material, we also compared the terminalia of female specimens from the Russian Far East, Estonia and northern Greece without fi nding any reliable diff erences. Moreover, Papp (2000) confi rmed conspecifi city when he compared central European material from Hungary with the Mongolian type material. Male terminalia are fi gured by Laštovka and Matile (1974) and subsequently by Matile (2002), while female terminalia have previously been fi gured by Zaitzev (1982) and Kurina (1997). Our association of males and females are based on multiple simultaneous fi ndings in trap samples (see above) that agrees with previous descriptions of the female. In the Pre-Balkan mountain range in Bulgaria, the species has been collected in xerotermic oak forest (Bechev 2000). Th e species was quite common in samples taken in a bait trap, operated on the basis of a mixture of fermenting sugar and red wine, in southern Estonia (see also Kurina 2006). Th e above-mentioned specimens from Greece are the fi rst records from the country. Zaitzev, 1982 Figures 4, 8, 12, 16, 20, 24, 28  Diagnostic characters. Female.Th orax bi-coloured; mesonotum yellow with variably developed black thoracic stripes; pronotum and propleuron yellow, other pleural parts brown to blackish. Abdominal sternites I-IV entirely yellow or slightly brownish; tergites of fi rst four segments bi-coloured: basally yellow, apically brown (in a few occasions fi rst four tergites entirely brown). Legs all yellow except dark brown band on apical fourth of hind femur. Tibiae densely covered with brown setulae. Scape, pedicel, and 3-5 fl agellomeres yellow, rest of fl agellum light brown. Mouthparts yellow. Apical palpal segment 4.1-4.4 (n=5) times as long as penultimate segment. Wing hyaline with slight yellowish tinge, all veins reach wing margin, M 1 and CuA 2 basally obsolete or very weak. Terminalia brown, cercus one-segmented, apically yellow. Gonapophysis IX membranous, subsquare with shallow incision apically, not visible in lateral view. Tergite VIII wider than tergite IX. Sternite VIII medially with deep and narrow incision, lateral incisions more shallow.

Greenomyia stackelbergi
Male. Coloration and other non-terminal characters similar to female. Th e apical palpal segment is 4-5 (n=5) times as long as penultimate segment.
Remarks. Besides its peculiar distribution (see Fig. 29), G. stackelbergi is unique among the four studied species in having vivid yellowish colouration and hyaline wings. It was described from South Primorje in the Russian Far East (Zaitzev 1982) and has subsequently been recorded only from two semi-urban localities in the Nordic region: the single locality in Swedish Lapland (present material, Kjaerandsen et al. 2007) and from one locality in the capital of Norway, Oslo (Søli and Kjaerandsen 2008). Eight years of collecting (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009) with Malaise traps, yellow pan-traps and window traps near a compost in the garden of one of the authors (MK) yielded 153  Zaitzev 1982, 199433 Zaitzev 1982, 199438 Zaitzev 1982, 1994.  Laštovka and Matile 197437 Zaitzev 1982, 199440 Zaitzev 1982, 1994. G. stackelbergi (red squares): 1 Søli and Kjaerandsen 20084 Kjaerandsen et al. 200739 Zaitzev 1982, 1994 specimens, indicating rise and decline of a small population. None was collected in the fi rst and the last year, while four in 2003, one in 2004, 40 in 2005, 47 in 2006, 54 in 2007 and seven in 2008. Th e fl ight activity lasted almost the whole vegetation season, from the middle of June (in 2007) to the beginning of October (in 2004). A garden compost is the supposed microhabitat for this population of G. stackelbergi and its origin should be somewhere in the surroundings. A close potential natural habitat could be the Vuollerim ravine a few hundred meters away. Waste from picked forest fungi might be another possibility.

Discussion
Species descriptions of fungus gnats are largely based and depending on characters in the male terminalia. Females are often ignored in taxonomic reviews and only a few generic reviews cover all or the majority of associated females (e.g. Søli 1997;Martinsen and Søli 2000;Kjaerandsen 2006Kjaerandsen , 2009. Still, females usually have distinctive yet less pronounced characters in their terminalia. In the case of the few European Greenomyia species we found it fairly easy to safely associate the females based on body characters such as colouration patterns, wing shape and venation details shared between the sexes, and the associations were further strengthened by co-occurrence in multiple trap samples. Our study of Greenomyia revealed that the diagnostic characters used to distinguish Greenomyia and Neoclastobasis in keys (e.g. Søli et al. 2000) does not hold up to a closer scrutiny, especially when both sexes are considered. Both sexes of G. borealis have wings where M 2 and CuA 1 end slightly before the wing margin, and both sexes of G. borealis and G. stackelbergi have prolonged apical segment of their palps. Th ese characters are akin to those used to diagnose Neoclastobasis. Yet, although the three known species of Neoclastobasis are very similar to Greenomyia in general appearance they show distinctive features in their terminalia that separate them from Greenomyia. In Greenomyia the dorsal branch of the male gonostylus always has two distinct combs of blunt spines on an otherwise bare inner surface. In Neoclastobasis the entire inner surface is covered with short blunt setae and a single row of larger spines is situated basally. Females of Neoclastobasis have a few short and strong spines along the apical margin of sternite VIII, which are never found in Greenomyia. We think Greenomyia and Neoclastobasis may prove to be monophyletic sistertaxa, but pending genetic studies and a better defi nition of the entire Leiini clade we leave the question whether they deserve to retain their status as separate genera or could be joined into one. In the meantime separating the two genera must rest entirely on diff erences in their terminalia as described above.