MICHELSEN: of Sylvicola

VERNER MICHELSEN

Zoological Museum, University of Copenhagen

WOOD GNATS OF THE GENUS SYLVICOLA (DIPTERA, ): TAXONOMIC STATUS, FAMILY ASSIGNMENT, AND REVIEW OF NOMINAL SPECIES DESCRIBED BY J. C. FABRICIUS

Michelsen, V., 1999. Wood gnats of the genus Sylvicola (Diptera, Anisopodidae): taxonomic status, family assignment, and review of nominal species described by J. C. Fabricius. − Tijd- schrift voor Entomologie 142: 69-75. [ISSN 0040-7496]. Published 22 September 1999. Comments are given on the contents and relationships of Sylvicola Harris, 1780, an important genus of anisopodid gnats. Tonnoirina Amorim & Tozoni, 1994 is sunk as a new junior syn- onym. At the family-group level aspects of nomenclature, fossils and taxon/lineage definitions are dealt with, and the phylogenetic structure of extant wood gnats discussed. It is preferred to treat them all in a single family and to abandon other formal supra-generic classification. A composite maxillo-labial ‘mentum’ is a remarkable state of the adult mouthparts that seems to combine the Mycetobia- and Olbiogaster-groups. Finally, the identity of three nominal species of Sylvicola has been checked and lectotypes designated on basis of authentic type material. All three are from Europe and were described by J. C. Fabricius. Sylvicola subfuscatus Krivosheina & Menzel, 1998 is a new junior synonym of S. fuscatus (Fabricius, 1775) and S. fuscatoides sp. n. is proposed for ‘S. fuscatus Fabricius’ of recent authors. V. Michelsen, Zoological Museum, University of Copenhagen, Universitetsparken 15, DK- 2100, Copenhagen Ø, Denmark. E-mail: [email protected] Key words. − Diptera; Anisopodidae; Sylvicola; phylogeny; classification; nomenclature.

Wood gnats or Anisopodidae in the sense of Hen- ly, the identity of nominal species of Sylvicola de- nig (1973) and Peterson (1981), i.e. including Sylvi- scribed by J. C. Fabricius (and other 18th century au- cola Harris, Olbiogaster Osten Sacken, Mycetobia thors) is assessed. Meigen and related genera, are today a small family of Acronyms for depositories: ZIUL, Zoological Insti- nematocerous Neodiptera (see Michelsen 1996a) tute, University of Lund; ZMUC, Zoological Museum, with ca. 120 species described. However, a rich fossil University of Copenhagen. record suggests that the direct lineage leading to the modern family can be traced back to early Middle TAXONOMIC STATUS OF SYLVICOLA HARRIS Mesozoic (200+ Myr). The family is rather cos- mopolitan, occurring in arboreal habitats ranging Over time there has been consensus on gathering from tropical to cold-temperate climates. A few most species of wood gnats into a single genus based species of Sylvicola, the so-called ‘window gnats’, are on characters of the wing: vein R2+3 strongly sinuous very common in domestic situations. Anisopodid lar- and wing membrane with macrotrichia. By that diag- vae are terrestrial saprophages living in decaying or- nosis the genus is large (about 75 extant species ganic matter. Rotting wood and sap runs seem known) and distributed worldwide. Yet, the simple favourable to many species of anisopodids, but a question of what name to apply for it has been a most much wider range of breeding media, including ma- contentious issue. Meigen (1818) ignored his own nure and decaying vegetables, are utilised by the com- earlier (Meigen 1803, 1804) proposal of the name mon window gnats. Anisopus, and all subsequent 19th century authors The purpose of this paper is to discuss and clarify adopted the name Rhyphus Latreille, 1804. Upon the various topics that relate to the systematics of Sylvico- discovery of an obscure pamphlet issued by Meigen la. The nomenclature and synonymy of the genus is (1800), Hendel (1908, 1928) introduced the senior considered first, followed by some supra-generic is- name Phryne, a course followed notably by German sues (family nomenclature, status of fossil names, re- and French authors. However, a still older name was lationships and classification of extant genera). Final- introduced through Coquillett’s (1910) fixation of a

69 T  E,  142, 1999

type-species for Sylvicola. Harris (1780: 100, pl. 31) Genus Sylvicola Harris, 1780 proposed that generic name for nine newly described Sylvicola Harris, 1780: 100, index [4]. Type-species Sylvico- British species of the present families la brevis Harris, 1780 [= Tipula fenestralis Scopoli, 1763], (7), (1) and Anisopodidae (1). It was first by designation of Coquillett (1910: 610). given as ‘ ’, but appears in proper singular [Phryne Meigen, 1800: 16. Unavailable, suppressed by Sylvicolae I.C.Z.N. (1963: 339).] form in the legend to plate 31 (on p. [4] of the ‘In- Anisopus Meigen, 1803: 264. Type-species Anisopus fuscus dex’). Knab (1912) was not convinced about the va- Meigen, 1804 [= Tipula fuscata Fabricius, 1775], by des- lidity of either of the names Sylvicola and Phryne. In- ignation of Coquillett (1910: 507). stead, he adopted the name Anisopus on grounds that Rhyphus Latreille, 1804: 188. Type-species Tipula fenestralis is has priority over Rhyphus, a course followed notably Scopoli, 1763 [as ‘fenestrarum’], by monotypy. Tonnoiri- by British and American authors. Eventually, the se- na Amorim & Tozoni, 1994: 531. Type-species Anisopus festivus Edwards, 1928, by original designation. Syn. n. nior name Sylvicola gained general acceptance by the early 1960s, probably on the influence of Melville’s (1960) report leading to the suppression of all FAMILY ASSIGNMENT OF SYLVICOLA HARRIS: Meigen-1800 names in 1963. That report (pp. 31- A COMMENTARY ON THE SUPRA-GENERIC 32) clarified the nomenclatorial status of the oldest genus-group names in Anisopodidae. CLASSIFICATION OF WOOD GNATS Based on the small Nearctic fauna (five species only), a division of Sylvicola into two subgenera, Nomenclature Sylvicola s. str. and Anisopus was attempted by Pratt & The following list presents, in chronological order, Pratt (1980). However, as argued by Amorim & To- all the family-group names that have been proposed zoni (1994), such action probably leaves Sylvicola for extant and fossil (†) groups of anisopodid or al- s.str. paraphyletic in terms of Anisopus. These authors leged ‘anisopodiform’ gnats. Evenhuis (1994) should proposed instead a division of the former genus Sylvi- be consulted for full references to the literature on the cola into a Laurasian and a Gondwanan component. fossil names. Sylvicola was maintained for the Laurasian group of about 15 species, and a new genus, Tonnoirina was Rhyphidae Newman, 1834: 379, 387 [as Rhyphites]. Type- erected to accommodate the Gondwanan group (60 genus Rhyphus Latreille, 1804. Oriental, Australian, Afrotropical and Neotropical Mycetobiinae Winnertz, 1863: 657 [as Mycetobinae]. Type-genus Mycetobia Meigen, 1818. species). However, I find that classification equally †Protorhyphidae Handlirsch, 1906, Fossilen Insekten: 487. difficult to accept. The only evidence given in sup- Type-genus Protorhyphus Handlirsch, 1906 (Lower Juras- port of the monophyly of Tonnoirina is the dorsally sic). contiguous state of the male eyes. Considering the Anisopodidae Knab, 1912: 111 [as Anisopidae]. Type-genus frequent occurrence of extreme male holopticism in Anisopus Meigen, 1803. Replacement name for Rhyphi- dae. the basal groups of Neodiptera, this is not much of an [Phrynidae Hendel, 1928: 9, 63 [as Phryneidae]. Type- argument. Even among wood gnats, the contiguous genus Phryne Meigen, 1800. Unavailable replacement state of the male eyes is not confined to Tonnoirina. It name for Rhyphidae, type-genus suppressed for the pur- occurs at least in Sylvicola punctatus (Fabricius) and in poses of zoological nomenclature.] all Mesochria Enderlein. Finally, an examination of Olbiogastridae Hennig, 1948: 78 [as Olbiogasteridae]. material from Terra del Fuego, Argentina (in Type-genus Olbiogaster Osten-Sacken, 1886. †Oligophrynidae Rohdendorf, 1962, [Fundament. Pale- ZMUC) revealed that the ‘Gondwanan’ Anisopus ont.] 9: 332 [as Oligophryneidae]. Type-genus andinus Edwards, 1930 is not a species of Tonnoirina Oligophryne Rohdendorf, 1962 (Lower ). but a Sylvicola in the sense of Amorim & Tozoni: †Protolbiogastridae Rohdendorf, 1962, [Fundament. Pale- male eyes narrowly separated dorsally; female with a ont.] 9: 332. Type-genus Protolbiogaster Rohdendorf, single spermatheca. Edwards (1930: 115) was evi- 1962 (Lower Jurassic). dently right in suggesting that a close relationship ex- [Sylvicolidae Hennig, 1954: 290. Type-genus Sylvicola Har- ris, 1780. Unavailable, published as a synonym of ists between the ‘Gondwanan’ andinus and the Phryneidae.] ‘Laurasian’ fenestralis-species group. Sylvicolidae Alexander, 1962: 3, 7. Type-genus Sylvicola On the above evidence, I prefer to maintain Sylvi- Harris, 1780. Replacement name for Anisopodidae. cola in the accustomed broad sense. The formal syn- †Crosaphididae Evans, 1971, Mem. Queensl. Mus. 16: 146. onymy is as follows: Type-genus Crosaphis Evans, 1971 (Upper ). †Limnorhyphidae Hong, 1983, [Middle Jurass. Ins. N. Chi- na]: 132. Type-genus Limnorhyphus Hong, 1983 (Middle Jurassic). †Siberhyphidae Kovalev in Kalugina & Kovalev, 1985, [Dipt. Ins. Jurass. Sib.]: 123. Type-genus Siberhyphus Kovalev in Kalugina & Kovalev, 1985 (Middle Jurassic).

70