Mangas Kovalev, 1986

Greenwalt, Dale E. & Blagoderov, Vladimir A., 2019, Review of the fossil record of Bolitophilidae, with description of new taxa and discussion of position of Mangas Kovalev (Diptera: Sciaroidea), Zootaxa 4567 (3), pp. 546-560 : 552-553

Mangas Kovalev, 1986

Type species. Mangas exilis Kovalev, 1986: 136 , by original designation.

Diagnosis. Large to medium size gracile flies. Male antennae at least as long body; long legs; tibial spur formula 1:2:2. Sc complete, sc-r in the middle of Sc. R 1 apex with pterostigma. Rs base curved, as long as section M 3; section M 2 at least twice the section M 3; M fork at least 3x the section M 3. Crossvein tb absent (M stem touching mcu) or replaced by a short fusion. Abdominal segments 1 and 2 shorter than segment 3; segments 5–7 of male more sclerotised that preceding ones. Male gonostyli simple.

Discussion. Kovalev (1986) established the new subfamily Mangasinae and described Mangas exilis from the Lower Cretaceous of West Mongolia. Kovalev (1986) figured the specimen with basal and posterior portions of the wing missing. Neither the origin of the posterior fork nor any portion of Sc is preserved. M. exilis seemingly lacks a pterostigma (but see discussion below) and, unlike all other species in the family Bolitophilidae , R 2+3. Kovalev (1986) suggested that the relatively long antennae of this specimen and the relative position of ta (r-m) and tb (basalmost portion of M 3+4) indicate a close relationship with Bolitophilidae . Chandler (2002) re-examined the specimen and stated that “ tb (bm-cu) and the basal part of CuA (is) identifiable by analogy to its venation”. Chandler (2002) concluded that M. exilis “cannot be demonstrated to be closely related to... a bolitophilid.” and was “allied to Heterotricha ”. Hippa and Vilkamaa (2005), in a cladistics analysis of both extant and extinct fossil Sciaroidea, and in which M. exilis was coded for 18 out of 77 characters analyzed in the study, concluded that M. exilis “belongs to the ‘keroplatid clade”, distinct from a clade that included Bolitophila cinerea , Mangas sp. (see below) and the extinct Archipleciomima obtusipennis Rohdendorf, 1962 . Blagoderov and Grimaldi (2004) suggested that M. exilis had a suite of characters including 1) a long, curved base of Rs situated distally, 2) r-m vertical, 3) section M2 long and longitudinal and 4) a short section M3 (M 1+2 stem) that were characteristic of Bolitophilidae . At least characters 3 and 4 from the list are apomorphies compared to other Mesozoic Sciaroidea.

A second specimen, from the Lower Cretaceous of the Transbaikal region, consists only of an isolated wing and was identified by Kovalev (1986) as Mangas sp. This specimen, like M. exilis , lacks R 2+3 although it does have a prominent pterostigma and its Sc is strongly sclerotized. The tb crossvein is absent, m-cu touches the M stem directly, a condition also found in various Bolitophilidae , for example, the extant species B. tenella Winnertz, 1863 , B. basicornis Mayer, 1951 , B. atlantica Fisher 1934 and B. dupla Garrett, 1925 ( Shaw 1962; Hutson et al. 1980). With the exception of the lack of R 2+3, this latter specimen keys to Bolitophilidae . Chandler (2002) re-examined the actual specimen and also recognised this specimen as a member of Bolitophilidae . Mangas sp., as “ Mangasinae gen. sp.” was identified as a sister group to Bolitophila by Hippa and Vilkamaa (2005) who retained Mangasinae as a subfamily of Bolitophilidae , defined by Kovalev (1986) by the apomorphy “absence of R 4 ” (R 2+3 here). They populated Mangasinae with Kovalev’s single undescribed specimen which they removed from the genus Mangas .

The reconsideration of the group here shows that conclusions of Chandler (2002) and Hippa and Vilkamaa (2005) are based on misinterpretations of fossil material, incorrect coding and literal interpretation of their cladogram. For example, a vein identified by Chandler as “ tb ” (2002: 136, fig. 92) is actually homologous to the section M2 of Kovalev (section of M between tb and r-m). In Hippa and Vilkamaa’s matrix (2005: 124–129, table 2), character 38 must be coded 1/1 instead of 1/0, character 45 – 0/? instead of 1/0, character 54 – 0/0 instead of 1/ 0 for Mangas sp./ Mangas exilis correspondingly; character 42 is based on incorrect interpretation of homology since section M2 of the M stem is present in almost all Bibionomorpha in one way or the other.

As Kovalev (1986) noted, the wing venation of Mangas is most similar to Bolitophila , with the exception of absence of R 2+3. This venation pattern is largely plesiomorphic ( Amorim & Rindal 2007; Blagoderov & Grimaldi 2004), shared by different basal Sciaroidea, such as Antefungivoridae , Keroplatidae Arachnocampinae and Heterotricha -group genera (sensu Chandler 2002). Mangas differs from Arachnocampa Edwards, 1924 in that the base of Rs is located near the middle of the wing, distad of m-cu (basally, proximad of m-cu in Arachnocampa ); sc- r is at the level of m-cu or slightly proximad (more basal in Arachnocampa ); CuA and A 1 converging (parallel or diverging in Arachnocampa ). It differs from Antefungivoridae and from the Heterotricha -group in its very short M3 section (compared both to section M2 and M fork) and the presence of a pterostigma; from the Heterotricha - group in the presence of a distinct M base (section M1). Antefungivoridae have vein R 5 curved posteriorly evenly, making cell r 1 tapered, while in Mangas R 5 is slightly sinusoid, and tip of cell r 1 being obtuse.

Remarks. Although the wing venation of Mangas exilis is preserved very poorly, its taxonomic position within Bolitophilidae can be established with certainty since its shares apomorphies, such as elongated flagellomeres, short section M3, cell r 1 wide and parallel-sided and obtuse at apex ( Kovalev 1986: Pl. XV, fig.6), and shortened abdominal segments 1 and 2 visible also in better preserved species.

The main arguments against inclusion of Mangas in Bolitophilidae are: 1) absence of pterostigma in M. exilis ; 2) poor preservation of the holotype of M. exilis preventing observation of Sc, and the bases of M and CuA. The absence of a pterostigma in the specimen of Mangas exilis may be an artifact of preservation. The specimen was preserved in rather coarse sandstones of the Gurvan-Erenii-Nuru locality, which did not allow preservation of delicate details, such as thin veins (e.g. bases of M). Interestingly, a specimen of another species from the same formation, Pachyrhyphus yakovlevi Kovalev (Anisopodidae) , described from much finer mudstones of the Boro- Nuru locality, contains a conspicuous pterostigma ( Kovalev 1986: fig. 98), while specimens of Pachyrhyphus from the Gurvan-Erenii-Nuru locality lack pterostigma ( Kovalev 1986: fig. 99). The wing venation of Mangas exilis cannot be reconstructed with certainty based on the holotype (the specimens was studied by VB in 1990–2005), however, the species shares at least two synapomorphies with Bolitophilidae : long antennae and a very short M stem (section M 3). None of fossil Diptera known from the Mesozoic with a similar wing venation pattern (Antefunguvoridae, Heterotricha -group genera, Bibionidae ) express these character states. Mangas exilis , based on the holotype specimen, cannot be distinguished with certainty from Arachnocampa (Keroplatidae) ; however, Mangas sp. ( Kovalev 1986: fig. 106, Chandler 2002: 136–137, fig. 91) does have characters of Bolitophilidae . Discovery of new material of Mangasinae and descriptions of new species confirms placement of Mangas within Bolitophilidae .

New material from Khosyurty (see below) corroborates Kovalev’s (1986) original diagnosis of the genus.