New findings of Flagellisargus J Zhang , 2012 ( Diptera , Brachycera , Archisargidae ) , with discussion of the placements of some controversial taxa

A new species of a new subgenus and a similar known species referred to the genus Flagellisargus J Zhang, 2012 are described and illustrated based on a male and a female impression fossils of these flies: Flagellisargus (Changbingisargus) parvus subgen. et sp. n. and Flagellisargus (Flagellisargus) cf. sinicus J Zhang, 2012. The latter taxon is the first record of a female Flagellisargus. Recently taken out of Archisargoidea, this study concludes that Flagellisargus should be an archisargid genus based on the known (male) and new (female) impression fossils. The placement of Daohugosargus J Zhang, 2012b is reassessed. It demonstrates close similarities in body structure and wing venation to archisargid flies, and can be retained as an archisargid genus. Archirhagio mostovskii J Zhang, 2015 is separated from Archirhagio zhangi K Zhang et al., 2009. Helempis Ren, 1998 could be, as a separate genus, placed in Archisarginae, Archisargidae.


Introduction
Archisargidae is an important, primitive, extinct family of the Lower Brachycera, Diptera.It is undoubtedly the largest early brachycerous group in the Mesozoic.To date, 55 species referred to 14 genera within two sub-Sharasargus fortis Zhang et al., 2008 Callovian-Oxfordian Daohuguo, China Daohugou Bed Sharasargus maculus Zhang, 2015 Callovian-Oxfordian Daohuguo, China Daohugou Bed Sharasargus oresbia (Ren, 1998) Early Cretaceous Huangbanjigou, China Yixian Formation Sharasargus ruptus Mostovski, 1996 Late Jurassic Shara-Teg, Mongolia Shara-Teg Bed Sharasargus spiniger Mostovski, 1996 Callovian-Oxfordian Karatau, Kazakhstan Karabastau Formation Sinallomyia ruderalis (Ren, 1998 Illustrations.Specimen descriptions, photomicrographs and line drawings were done without immersion with the exception of photographs of details of the antennae and tibial spurs.The line drawings were produced with the aid of a camera lucida and the digital photomicrographs were taken using a stereomicroscope.
Colour described here refers to that of the fossil, where patterning is preserved.
Wing venation terminology follows that of Wootton and Ennos (1989) and Shcherbakov et al. (1995).The cell traditionally named the anal cell is, in fact, considered to be the cubital cell herein.Diagnosis.Small-size archisargid flies (body excluding antenna and genitalia less than 5 mm long); antennal scape long; arista (or stylus) absent; fork of R4+5 shallow, distad of level of R2+3 end; R5 ending before wing tip; discal cell short and wide (nearly 2.3 times as long as wide).
Remarks.The subgenus Flagellisargus (Flagellisargus) stat.n. includes three known species: Flagellisargus (Flagellisargus) robustus J Zhang, 2012a, Flagellisargus (Flagellisargus) sinicus J Zhang, 2012a and Flagellisargus (Flagellisargus) venustus J Zhang, 2012a.Among them, the first and the third species are erected based on nearly complete male flies, the second one with head and abdomen missing.This known subgenus differs from Flagellisargus (Changbingisargus) subgen.n. in the following aspects: moderate-size archisargid flies (body excluding antenna and genitalia more than 9 mm long); antennal scape short; arista (or stylus) present; fork of R4+5 relatively deep, just at level of R2+3 end; R5 ending just at wing tip; discal cell narrow and long (nearly three times or more as long as wide).
Although the head and abdomen are missing, Flagellisargus (Flagellisargus) venustus demonstrates close similarities in wing venation to that of Flagellisargus (Flagellisargus) sinicus and Flagellisargus (Flagellisargus) robustus: fork of R4+5 relatively deep, just at level of R2+3 end; R5 ending at wing tip; discal cell narrow and long, nearly 3.5 times as long as wide (Fig. 3A-C).Thus, Flagellisargus (Flagellisargus) venustus should be retained in Flagellisargus (Flagellisargus) rather than be assigned to Flagellisargus (Changbingisargus) subgen.n.Diagnosis.Male archisargid flies 4.9 mm long (excluding antenna); antenna longer than head, scape more than one half of flagellum length; stem of Rs nearly as long as bR4+5; first fork of Rs slightly basad of level of M fork; crossvein r-m linking anterior margin of discal cell near to M fork; crossvein m-m long; section of mM3+4 short; male genitalia large, gonostylus subquadrate with apical denticle medially.
Abdomen with seven segments visible, nearly ovate-oblong, fourth widest, and nearly as wide as thorax, 1.8 times longer than head (excluding antenna) and thorax combined; genitalia rather large, subovate, longer but narrower than seventh abdominal segment, gonocoxite more or less oblong with its inner and outer margins slightly curved outwards, gonostylus subquadrate, wider than long, with a triangular apical denticle curved upwards, aedeagus invisible (Figs 1C, 2C).
Diagnosis.Moderate-size archisargid flies (body excluding antenna and genitalia more than 9 mm long); antennal scape short; arista (or stylus) present; fork of R4+5 relatively deep, just at level of R2+3 end; R5 ending just at wing tip; discal cell narrow and long (nearly three times or more as long as wide).
Abdomen with nine segments visible, nearly cylindrical, just a little narrower than thorax, 1.9 times longer than head (excluding antenna) and thorax combined; each of tergites with a wide, longitudinal, intermediate marking which is darkish brown; apex of abdomen with a scelerotized, needle-like ovipositor, and slightly longer than ninth segment (Fig. 4A).
Remarks.On the following characters, this fly could be assigned to Flagellisargus (Flagellisargus): body (excluding antenna and ovipositor) moderate-size (more than 9 mm long); antennal scape short (not elongated); arista (or stylus) well developed (about a quarter of flagellum length); fork of R4+5 just at level of R2+3 end; and R5 ending at wing tip.
Owing to having special characteristics (antennal flagellum with a darkish brown longitudinal furrow near to its outer margin and connecting base of arista and a tibial spur of hindleg well developed) this specimen shows close similarities in antennal and leg's structures to that of the known species Flagellisargus (Flagellisargus) sinicus.Unfortunately, its wing is incompletely preserved, and the discal cell, posterior branch of M, CuA, CuP and crossvein m-cu are rather ambiguous or invisible.For this reason, this impression fly could only be identified as Flagellisargus (Flagellisargus) cf.sinicus.

Discussion
Recently, Grimaldi and Barden (2016) proposed a single most-parsimonious tree indicating the relationships within the superfamily Archisargoidea.They considered that three genera possessing the plesiomorphic condition of unmodified (non aculeate) female terminalia are not basal to Archisargoidea: Daohugosargus J Zhang, 2012b, Orientosargus J Zhang, 2012b and Uranorhagio K Zhang, 2010.Meanwhile, "Flagellisargus has a plesiomorphic, non stylate type of antenna and may also lie outside the Archisargoidea sensu stricto, but this would need to be confirmed with female specimens (only males presently are known)" (Grimaldi and Barden 2016: 17).
However, this study argues that Flagellisargus has a well developed arista (or stylus) although it is short.This crucial character had been illustrated in the original generic diagnosis and specific descriptions (J Zhang 2012a: 879, 881, Figs 3, 7).Furthermore, the female Flagellisargus has been discovered, and described herein.Flagellisargus (Flagellisargus) cf.sinicus has a scelerotized, needle-like ovipositor (Fig 4A).It is clear that Flagellisargus should be an arichsargid genus even according to the alternative classification proposed by Grimaldi and Barden (2016).
As for Daohugosargus, this genus was proposed for Sharasargus eximius K Zhang et al., 2008, which is a monotypic genus based on an incomplete impression fly with terminal abdominal segments missing (K Zhang et al. 2008).Its sex is uncertain.It is difficult to see how this genus could be distinguished as a female fly, let alone with unmodified (non aculeate) female terminalia.Daohugosargus demonstrates plesiomorphic similarities in body structures (as preserved) and in wing venation to those uncontested archisargids, and differs only from them by the characteristic vein R2+3 which is short, S-shaped, and arising late from Rs.It would be unreasonable to move this genus out of the superfamily Archisargoidea based only on this difference.This study considers that Daohugosargus is related rather to Archisargidae, Archisargoidea than to any other superfamilial groups.
Furthermore, the conclusion is debatable whether genera having non-aculeate female terminalia lie outside of Archisargoidea.For example, there are two species, Archirhagio striatus J Zhang et H Zhang, 2003 and Archirhagio varius J Zhang, 2015, belonging to the archisargid genus Archihagio Rohdendorf, 1938 that need consideration.The former species has a highly sclerotized, aculeate ovipositor; while, the latter one possesses a blunt, enlarged, fleshy, hook-like ovipositor (Wang et al. 2017: Figs 4D, E, originally the "ovipositor" was labelled as a "hypogynial valve").However, Archirhagio varius cannot be excluded from Archihagio based on its species diagnosis although it has a non-aculeate female terminalia.Another example is the two species of Ovisargus Mostovski, 1996: O. gracilis Mostovski, 1996and O. singulus J Zhang, 2015.The former species has an aculeate ovipositor but the latter one has a podgy, conical (non aculeate) ovipositor.O. singulus should be assigned to Ovisargus based on the similarities in body structures and wing venation to that of O. gricilis regardless of the ovipositor.In addition, an aculeate ovipositor has evolved homoplastically in Diptera.It occurs in various groups, including a few Tipulidae, Phoridae, Pipunculidae, some Conopidae, Tephritoidea, Cryptochaetidae and Tachinidae (Pritchard 1983, Feener and Brown 1997, Skevington & Dang 2002, Stireman 2006, Grimaldi et al. 2011, Q Zhang et al. 2016).None of these groups (superfamilies or families) are distinguished based only on the specialized ovipositor.It is evident that the aculeate ovipositor is a convergent development in functional morphology, and does not reveal relationships between these taxa.
Using a geometric morphometric analysis, Wang et al. (2017) reviewed and revised the classification of Archirhagio.They redefined the diagnosis of Archirhagio zhangi K Zhang et al., 2009, and proposed Archirhagio mostovskii J Zhang, 2015 to be a junior synonym for Archirhagio zhangi based mainly on some similarities of wing venation and shape of abdominal segments.This study argues that both species were erected based on almost complete impression fossils of the male flies.As the placement is debatable, an overall, further comparative analysis in body structures and wing venation was necessary.Wang et al. (2017) ignored the sharp difference between both holotypes in some key taxonomic characteristics.Archirhagio mostovskii differs from Archirhagio zhangi in the following aspects: (1) male holoptic vs male dichoptic; (2) markings on abdominal tergites differ sharply; (3) size and shape of wing and wing venation differ distinctly; and (4) male genitalia differ distinctly.Thus, Archirhagio mostovskii can be separated from Archirhagio zhangi.Some detailed explanations are given as follows.In Diptera, the eyes of most families are holoptic (Cumming and Wood 2009); only a few families have a dichoptic male that is used in the family diagnosis in Lower Brachycera, e.g.Asilidae and Xylophigidae (Fisher 2009, Woodley 2009a).It is clear that the condition (male holoptic or dichoptic) is an important diagnosis for the identification of the Lower Brachycera.Both species, Archirhagio mostovskii and Archirhagio zhangi, are erected based on males, the former species having holoptic eyes with a very long midline (J Zhang 2015: Figs 2B, 4A); in contrast, the latter one has dichoptic eyes, which are widely separate (K Zhang et al. 2009: Fig. 2).It is impossible that the different compound eye types of the male mentioned above occur in the same species.On the basis of these crucial taxonomic characters, Archirhagio mostovskii should be separated from Archirhagio zhangi.
Secondly, the shape and arrangement of the abdominal markings frequently provide useful taxonomic characters for dividing various groups of the Lower Brachycera, at least at species level, and many such studies have been published (Jones and Anthony 1964, Smith 1989, Woodley 2009a,b, etc.).Archirhagio zhangi shows each of abdominal tergites I-VI with a patch at the posterolateral corner (a left patch in segments II-IV is also present but badly preserved to judge from the original photomicrograph -see K Zhang et al. 2009: Fig. 1A, Fig. 5B herein).In contrast, Archirhagio mostovskii has a wide, medially longitudinal stripe and a wide transverse band along the  K Zhang et al., 2009(after K Zhang et al., 2009, modified), C male terminalia (after K Zhang et al., 2009, modified).Scale bars 1mm.
hind margin on each of the abdominal tergites IV-VII, and the markings occupy almost the whole of tergites I-III (Fig. 5A herein).The sharply different markings on the abdominal tergites indicate that both male holotypes cannot be assigned to one and the same species.
Thirdly, Archirhagio mostovskii shows the wings are clearly shorter and wider than that of Archirhagio zhangi, (wing 12.1-13 mm long, 3.0-3.4mm wide, about three times as long as wide vs wing 17.5 mm long, 3.8 mm wide, 4.6 times as long as wide); the wing is about one half of body length in the former species vs about fourth-fifths in the latter one.It should be pointed out that the revised diagnosis of Archirhagio zhangi defining body length between 29 and 32 mm, wing length between 20 and 23 mm, is questionable because the holotype of Archirhagio zhangi (body 21 mm long, wing 17.5 mm long) and the holotype of Archirhagio mostovskii (body 22.2 mm long, wing 12.1-13 mm long) falls distinctly short of that size.This revised diagnosis is related neither to Archirhagio zhangi nor to Archirhagio mostovskii.Furthermore, in wing venation the character of cell r1 closed or nearly so is an important diagnosis for Archirhagio mostovskii, differing from Archirhagio zhangi, in which cell r1 is clearly open.This crucial character demonstrates close similarity to that of Calosargus Mostovski, 1997, another archisargid genus.Nevertheless, in Calosargus the cell r1 is closed before the anterior margin of the wing, which has a very short or relatively long petiole apically [e.g.Calosargus (Pterosargus) sinicus J Zhang, 2010 and Calosargus (Pterosargus) thanasymus Mostovski, 1997].This key character mainly differentiates Calosargus from Archirhagio.It is interesting that Archirhagio mostovskii is considered as a connecting link between Archirhagio and Calosargus.On balance, one should keep Archirhagio mostovskii as a separate species referred to Archirhagio but closely related to Calosargus.
Finally, the structural characteristics of male terminalia provide an unparalleled array of taxonomic characters in Diptera (McAlpine et al. 1981)."Male terminalia are a key morphological source of characters used to distinguish species in the vast majority of Diptera families and there are few modern taxonomic studies that do not include illustrations of male terminalia to aid in species diagnoses" (Sinclair et al. 2013).However, Wang et al. (2017) did not describe and illustrate the characteristics of male terminalia in the revised species diagnosis of Archirhagio zhangi, although they also commented that the original description of the male terminalia was incorrect.They only supplied two photomicrographs of an unnumbered specimen instead of the holotype male terminalia of Archirhagio zhangi (Wang et al. 2017: Figs 4B, C).Furthermore, they claimed that there are no significant modifications in male terminalia across the genus Archirhagio, consisting of the reduced ninth tergite, unsegmented gonocoxites, and pair of large parameres (Wang et al. 2017).Meanwhile, without providing any reference and citation, they declared that the terminology "aedeagus" used by J Zhang ( 2015) is incorrect, and should be instead of paired "parameres" (Wang et al. 2017).These deductions proposed by them clearly run counter to what many dipterists have concluded (McAlpine et al. 1981, Woodley 1989, Cumming and Wood 2009, Sinclair et al. 2013).This study argues that the kidney-shaped gonocoxite, bipectinate gonostylus and short and stout aedeagus demonstrate Archirhagio mostovskii as having distinctly different structures in the male terminalia from the specimens provided by Wang et al. (2017: Figs 4A,B,C).Unfortunately, there is neither description nor line drawing of the male terminalia of Archirhagio provided in their article; and thus a further comparison of male terminalia between Archirhagio mostovskii and Archirhagio zhangi is difficult herein.On the other hand, if those male terminalia investigated by them possess the same structures, then those specimens most likely belong to one and the same species that differs from Archirhagio mostovskii.It should be also noted that in Stratiomyomorpha + Muscomorpha sensu Woodley (1989), the aedeagus is indistinguishably fused to the parameral sheath to form the phallus (Cumming and Wood 2009).Currently, Archisargidae is assigned either within or near to the Stratiomyomorpha (Oberprielar and Yeates 2012) or Archisargoidea is probably an extinct sister group to the Muscomorpha (Grimaldi and Barden 2016).In any case, the paired parameres should be indistinguishable in the male terminalia of archisargids [e.g.Flagellisargus (Flagellisargus) sinicus -see J Zhang 2012a: Fig. 3] The so-called parameres of Arichrhagio, an undouble archisargid genus, identified by Wang et al. (2017) should be phallus "(aedeagus sensu authors concerning Stratiomyomorpha and Muscomorpha sensu Woodley, 1989)" (Cumming and Wood 2009).
Originally, the genus Helempis Ren, 1998 including two species: H. yixianensis and H. eucalla Ren, 1998 was placed in Protempididae (Ren 1998).The present author (J Zhang 2012b) thought that the two species could be united into one species, namely H. yixianensis, and Helempis, as a subgenus, could be transferred into Ovisargus referred to Archisardinae, Archisargidae.Through further contrastive studying, it could be reasonable to retain Helempis as a separate genus within the Archisarginae, Archisargidae.It differs from Ovisargus by the elongated discal cell and the deeper fork of R4+5, which is distinctly basad to R2+3 end.