Introduction

The superfamily Archisargoidea comprises three families: Archisargidae, Kovalevisargidae and Eremochaetidae. The Archocyrtidae is probably the fourth family of archisargoids (Mostovski 1997). It is interesting that Archisargidae and Kovalevisargidae constitute one sister group; whereas Eremochaetidae and Archocyrtidae probably form another sister group. All these families synchronously appeared in the Callovian–Oxfordian from the Karabastau Formation at the Karatau-Mikhailovka locality in the Karatau Mountain Ridge, Chimkent Region, South Kazakhstan Province, Kazakhstan. To date, however, only the representatives of archisargids and eremochaetids are known to extend into the Early Cretaceous.

Lately, thousands of brachycerans have been discovered from the “Daohugou Formation” in the vicinity of Daohugou, Ningcheng, Chifeng, Inner Mongolia (JF Zhang and HC Zhang 2003; JF Zhang 2010a, 2010b, 2010c, 2011a, 2011b, 2012a, 2012b, in press; JF Zhang and Li 2012; KY Zhang et al. 2006, 2007a, 2007b, 2008a, 2008b, 2008c, 2008d, 2009, 2010a, 2010b) and the Yixian Formation in the vicinity of Huangbanjigou, Shangyuan, Beipiao, Liaoning, China (Ren et al. 1995; Ren 1998; Huang and Lin 2006). It is interesting that the members of almost all the archisargid and kovalevisargid genera recorded from the Karabastau Formation were also recovered from the “Daohugou Formation”. Meanwhile, on the basis of review of brachycerans from the Yixian Formation, it is clear that some relics of archisargid genera and species did also occur in the Early Cretaceous (see Discussion below).

A new family, Orientisargidae fam. n., composed of a new genus and species, Orientisargus illecebrosus gen et sp. n., is described here. Another new genus, Daohugosargus gen. n., is proposed for the known species Sharasargus eximius KY Zhang et al., 2008. Daohugosargus eximius (KY Zhang et al., 2008), comb. n. has a characteristic wing venation, which differs sharply from all the known representatives of archisargids, and may be temporarily assigned to the subfamily Uranorhagioninae (stat. n.) within Archisargidae. The systematic position for Uranorhagionidae is reassessed. It is a junior synonym for Archisargidae, and may be degraded as a subfamily within Archisargidae. Mostovskisarginae is a junior synonym for Uranorhagionidae. Meanwhile, Mostovskisargus JF Zhang, 2010 and Strenorhagio KY Zhang et al., 2010 can be synonymized with Uranorhagio KY Zhang et al., 2010. Two species, Strenorhagio deviatus KY Zhang et al., 2010 and Strenorhagio grimaldi KY Zhang et al., 2010, can be united into one species: Uranorhagio deviatus (KY Zhang et al., 2010), comb. n. Mostovskisargus portentosus JF Zhang, 2010, Mostovskisargus signatus JF Zhang, 2010 and Strenorhagio conjugovenius KY Zhang et al., 2010 are synonyms for Uranorhagio asymmetricus (KY Zhang et al., 2010), comb. n. The species of Mesosolva and related taxa recently described from the “Daohugou Formation” are reassessed: Brevisolva KY Zhang et al., 2010 is a junior synonym for Mesosolva Hong, 1983. A new specific name, Mesosolva zhangae nom. n., is proposed for the Brevisolva daohugouensis KY Zhang et al., 2010. Mesosolva jurassica KY Zhang et al., 2010 should be synonymized under Mesosolva sinensis KY Zhang et al., 2010. A new generic name, Sinallomyia nom. n., is proposed instead of Allomyia Ren, 1998 which is a junior homonym for Allomyia Banks, 1916 (a genus of Trichoptera). The Early Cretaceous Helempis eucalla Ren, 1998, Helempis yixianensis Ren, 1998, Pauromyia oresbia Ren, 1998 and Sinallomyia ruderalis (Ren, 1998) from the Yixian Formation previously placed, respectively, in the Tabanidae, Rhagionidae and Protempididae should be transferred to the Archisarginae of Archisargidae.

Material and methods

Specimen descriptions, photographs, and drawings were obtained without the application of glycerol to the surface of the specimens. The specimens collected by the author in field were examined under a stereomicroscope (Wild Heerbrugg) and illustrated with the aid of a drawing tube attached to it, re-adjusted using image-editing software (Adobe Photoshop CS). The digital photographs were taken using stereomicroscope (AXioCamHR3).

Wing venation terminology here follows 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. The specimens are deposited in the Nanjing Institute of Geology and Palaeontology (NIGP), Chinese Academy of Sciences.

Systematics Superfamily Archisargoidea Rohdendorf, 1962 Discussion

KY Zhang et al. (2010a) erected a new family, Uranorhagionidae KY Zhang, Yang et Ren, 2010 including five new species referred to two new genera based on several specimens from the “Daohugou Formation” in the vicinity of Daohugou, Inner Mongolia of China (not the true Jiulongshan Formation). They assigned Uranorhagionidae to the superfamily Tabanoidea, and considered Uranorhagionidae exhibiting a mixture of distinct characteristics of two families, the Rhagionemestriidae (Nemestrinoidea) and the Rhagionidae (Tabanoidea), but failed to discuss the relationship of Uranorhagionidae and Archisargidae (Archisargoidea).

Comparing Uranorhagionidae with the Archisargidae, however, almost all the characteristics derived from body structures and wing venation in the former family are very closely similar to the latter family. The major difference of significantly bent R2+3 also demonstrates close resemblance to that of Daohugosargus eximius, an undoubted representative of archisargids although its systematic position at generic level is debatable (see Figure 4). It should be noted that the position of r-m in Uranorhagionidae is unstable: in some species slightly distad to, or just at, in other species slightly basad to, Rs fork. Nevertheless, this character also exists in Archisargidae (most representatives versus Daohugosargus eximius). As far as the petiolate M1+2 behind d end is concerned, there are some members of archisargids with M1+2 fork just at d end [see Figure 5, Sharasargus oresbius (Ren, 1998), comb. n., originally Pauromyia oresbia Ren, 1998], which does belong rather to Archisargidae than to Rhagionidae (detailed discussion, see below); and is more or less similar to that of uranorhagionids. The distal position of the M1+2 fork with respect to the d cell distal end may be a unique feature of uranorhagionids, which has the taxonomic significance only at generic, at most subfamilial, rank. Thus, Uranorhagionidae is a junior synonym for Archisargidae, and could be degraded as a subfamily referred to Archisargidae.

KY Zhang et al. (2010a) described five new species respectively assigned to two new genera: Uranorhagio daohugouensis, Strenorhagio deviatus, Strenorhagio grimaldi, Strenorhagio asymmetricus and Strenorhagio conjugovenius. However, these impressions demonstrate close similarities in body structures and wing venation each other, and then could be united into three species referred to a single genus: Uranorhagio daohugouensis, Uranorhagio deviatus (KY Zhang et al., 2010) and Uranorhagio asymmetricus (KY Zhang et al., 2010). The genus Strenorhagio KY Zhang et al., 2010 could be synonymized with Uranorhagio KY Zhang et al., 2010. The Strenorhagio grimaldi and Strenorhagio conjugovenius are synonyms for Uranorhagio deviatus and Uranorhagio asymmetricus, respectively. It should be noted that minor differences in the wing venation should be attributed to individual variation, which can be quite substantial in the extinct archisargids. For example, the holotype of Uranorhagio asymmetricus (CNU-DIB-NN2007020) (KY Zhang et al. 2010, p. 569, fig. 10) shows the r-m clearly basad to Rs fork in its left wing, but just at Rs fork in its right wing. That is to say the position of r-m being inconstant not only in different individuals but even in a single specimen.

The author (JF Zhang, 2010a) described two new species referred to a new genus within a new subfamily based on two specimens from the “Daohugou Formation” in the vicinity of Daohugou, Inner Mongolia of China, and assigned them to Archisargidae — Mostovskisarginae: Mostovskisargus: Mostovskisargus portentosus and Mostovskisargus signatus. However, these taxa were published somewhat later (publication date: 16 March, 2010) than those described by KY Zhang et al. (26 February, 2010). Thus, Mostovskisarginae, Mostovskisargus, Mostovskisargus portentosus and Mostovskisargus signatus are junior synonyms of Uranorhagioninae, Uranorhagio and Uranorhagio asymmetricus, respectively.

KY Zhang et al. (2010b) described a new species referred a new genus Brevisolva daohugouensis and two additional Mesosolva species: Mesosolva jurassica and Mesosolva sinensis. Judging from the original descriptions, drawings and photographs (KY Zhang et al., 2010b, pp.76–79, figs 1–8), Mesosolva jurassica demonstrates very close resemblance in body structures and wing venation to Mesosolva sinensis, and then, both species should be united into a single one (see Figure 6A–C herein). It should be repeatedly emphasized: due to individual variation and/or sexual dimorphism, some minor differences in the wing venation usually occur within an archisargid species. Such differences may exist even between the left and right wings of a single specimen. It is evident that Mesosolva sinensis is closely similar in body structure and wing venation to Mesosolva daohugouensis JF Zhang et HC Zhang, 2003. Both species are from the same fossil site. It might be debatable whether the two species could also be united into one species. On account of the characteristic wing venationof Brevisolva daohugouensis (see Figure 6D) showing close similarities to Mesosolva sinensis it is difficult to see how the genus Brevisolva could be separated. The genus Brevisolva as defined by its authors (KY Zhang et al, 2010b) does not have diagnostic features that separate it from Mesosolva. As for the short Rs stem, short R5, the position of r-m which is close to d base, these characteristics in wing venation of Brevisolva could be treated as the difference between species, and are also similar respectively to some known species of Mesosolva, for example, in Mesosolva longivena Mostovski, 1996 and Mesosolva balyshevae Mostovski, 1996. Additionally, the short petiole of cell m3 is not the particular feature of Brevisolva. There is an undescribed impression of Mesosolva showing its petiole clearly shorter than the section dM3 (see Figure 7). Brevisolva daohugouensis could be regarded as a species of Mesosolva. A new specific name, Mesosolva zhangae (KY Zhang et al., 2010), nom. n., is proposed because the Mesosolva daohugouensis has already been occupied (JF Zhang and HC Zhang 2003).

Hong (1983) described two monobasic genera, Mesosolva and Prosolva from the Callovian–Oxfordian Haifanggou Formation in Beipiao, Liaoning, China. He assigned these to the family Xylomyiidae (originally Solvidae). JF Zhang et al. (1993) discussed the systematic position of Mesosolva parva Hong, 1983 and Prosolva huabeiensis Hong, 1983, pointed out that these probably belong in an unnamed group at familial level, which is probably related to the family Rhagionidae based on the characteristics of antenna; but the structures of antenna were mistakenly described: the so-called second segment (i.e. pedicel) is, in fact, the third segment (i.e. first flagellomere) (JF Zhang et al., 1993, p. 667). In the author’s collection of brachycerous flies from Daohugou biota, there is another nearly complete impression of male Mesosolva with antennae visible, which consist of the scape, pedicel, first flagellomere, and stylus (Figure 8). It is evident that the so-called pedicel described by Hong (1983) is the first flagellomere, in size and shape very closely resembling that of the present undescribed specimen (see Figure 8), although they are from different individuals at different fossil localities. Mostovski (1996a, 1996b) redefined Archisargidae and Mesosolva. Meanwhile, he transferred Mesosolva into Archisargidae, and described seven new Mesosolva species. He thought that Mesosolva parva and Prosolva huabeiensis probably belong to the same genus (Mostovski 1996b). JF Zhang and HC Zhang (2003) described the first record of Mesosolva from the Daohugou biota, and agreed with Mostovski’s (1996b) conclusion mentioned above. Recently, KY Zhang et al. (2010b) revised the diagnosis of Mesosolva proposed by Mostovski (1996b). Unfortunately, the redefinition is unsatisfactory. What is striking is the additional characteristic: CuA1 arising from cell bm, mouth of cell sc wide open, much wider than that of cells r1 and r2+3. They failed to explain how these features could be defined as the Mesosolva diagnosis. Actually, these that they added are common characteristics of archisargid genera, and occur in almost all representatives referred to various genera in the two subfamilies (Archisarginae and Uranorhagioninae) of Archisargidae. The author argues that these delineations proposed by KY Zhang et al. (2010) do not conform to the diagnoses of all the archisargid genera.

Sinallomyia ruderalis from the Lower Cretaceous Yixian Formation was originally regarded as a new genus and species of the subfamily Tabaninae within Tabanidae. Judging from the original illustration (Ren 1998, p. 69, fig. 6; Figure 9 herein) the R1 is some four-fifths of wing length; and the venation and body structure demonstrate, more or less, resemblance to Mesosolva zhangae (KY Zhang et al., 2010), nom. n.; hence, this genus and species can be transferred to the Archisarginae of Archisargidae (JF Zhang, 2012a).

Pauromyia oresbia from the same locality and horizon was previously assigned to the Rhagionidae (Ren 1998, p. 72, fig. 11; Figure 5 herein). This species can be moved into Sharasargus within Archisarginae, Archisargidae because its venation shares close similarity to Sharasargus spiniger Mostovski, 1996 referred to Archisarginae, Archisargidae (JF Zhang, in press).

Ren (1998) described two new species of a new genus: Helempis yixianensis and Helempis eucalla from the same locality and horizon. He considered these taxa having typical wing venation of Protempididae. On the basis of original drawings (Ren 1998, pp. 80, 81, figs. 22, 23; Figures 10, 11 herein), the two species which might be probably united into one species have very long R1, which is some four-fifths (or more) of wing length, relatively narrow and long wings, unsegmented arista, and the characteristic discoidal cell, which is distinctly shifted distally. All the characters contradict to including these species in the Protempididae. On the contrary, Helempis yixianensis and Helempis eucalla demonstrate close resemblance in wing venation to Ovisargus gracilis Mostovski, 1996, and then could be transferred to Ovisargus Mostovski, 1996 (Archisarginae, Archisargidae). A detailed discussion will be made in a separate paper.

The Origoasilus KY Zhang et al., 2011 previously erected as a new genus and assigned to a new family Origoasilidae KY Zhang et al., 2011 has been transferred to Archisarginae of Archisargidae. The Origoasilidae is a junior synonym for Archisargidae (JF Zhang, 2012b).

Ideally, these previously described species from the upmost Middle–lowest Upper Jurassic “Daohugou Formation” and the Lower Cretaceous Yixian Formation should be revised properly through re-examination of the type material, since the original drawings may contain details resulted from misinterpretation of insufficiently preserved structures. For this reason, until such time as reinvestigation of these specimens is possible, their taxonomic positions could be temporarily assigned to Archisarginae, Archisargidae based on original descriptions and drawings because the Mesozoic archisargid flies have characteristic wing venation (see revised diagnosis of Archisargidae mentioned above) which is easily separated from other extinct and extant families within the lower Orthorrhapha, Brachycera.