New species of robber fly of the genus Andrenosoma Rondani, 1856 (Diptera: Asilidae) from Iran

Andrenosoma lottae Mortelmans sp. nov. is described based on males and females from Fars province, Iran. Pictures of the habitus are provided, together with details of the male terminalia to facilitate species identification. The status of Andrenosoma violacea is discussed and hereby put forward as a species inquirenda. http://www.zoobank.org/urn:lsid:zoobank.org:pub:2E6AB0F2-7227-4BDA-BA1C-3BA032B7A84D

Species of Andrenosoma Rondani, 1856 are medium-to large sized robber flies (8-20 mm). Adults are often found lurking on trees looking for prey and commonly can be found in coniferous forest clearings, where they feed on xylophagous insects' larvae for example on fallen trees (García et al., 2017). Identification of Andrenosoma remains challenging as diagnostic characters are situated in male terminalia and often poorly described, whereas external morphological characters are variable and reference material is needed for identification.

Material and Methods
Morphological analyses and diagnoses of the specimens were performed with a Euromex NZ1903b stereomicroscope with 60x magnification. Diagnoses of the male genitalia, general measurements, and photographs of the specimens were made by use of a Leica M205 stereomicroscope with a maximum of 160x magnification, at the LifeWatch Marine Observatory (Flanders Marine Institute, Ostend, Belgium). With prior approval from the RBINS collections manager, Wouter Dekonick, we dissected one male specimen to enable examination of the terminalia. To prepare the male terminalia for examination, the authors employed a technique commonly used to study dipteran terminalia: (1) completely removing the abdomen, (2) soaking it for 10 minutes in warm KOH, (3) soaking it for 20 minutes in tap water, (4) soaking it for 10 minutes in EtOH-HCl (acidified ethanol), and finally (5) soaking it for 20 minutes in tap water. After examination, the macerated abdomen is placed in a plastic microvial containing a few drops of glycerin, and the microvial is pinned beneath the rest of the specimen. The abdomen was examined by placement in a spot plate with the depression filled with tap water. For photography, the abdomen was immobilized by placing it in a spot plate with the depression filled with glycerin.
Terminology used in this paper follows that of Cumming & Wood (2009). The map (Figure 1) is created with the open-source software R (RStudioTeam, 2019) The free raster map data used as background for the maps are accessible via Natural Earth (www.naturalearthdata.com). Species records are based on published data (Richter, 1985;Hayat et. al., 2008;Tomasovic & Saghaei, 2009;Ghahari et. al., 2014).

Andrenosoma Rondani, 1856
Andrenosoma is a typical member of the tribe Andrenosomatini, subfamily Laphriinae. The genus is recognised by the following combination of characters: antennae without an obvious arista, but a small sensory element placed in a small cavity at, or near the apex of the postpedicel (Geller-Grimm, 2003). Antennal postpedicel virtually as long as scape and pedicel combined in Andrenosoma. Vertex well-excavated with a broad frons and sharply marked ocellar tubercle (Hull, 1962). Facial swelling rather weakly developed (Hull, 1962). Proboscis shorter then height of head, dorsoventrally flattened and bent only slightly upwards apically (Hull, 1962). Proboscis tapering apically (compare to e.g., Pilica Curran, 1931) (Hull, 1962). The two-segmented palpi flattened (not round as in Laphriini). Metanotal callosity without setae (Hull, 1962). Costa complete (Fisher, 1986). Male genitalia rotated 45-90°, cerci and epandrium separated. A complete genus description can be found in Hull (1962) and Fisher (1986 Completely black in ground colour, with heavy silver pubescence; narrow longitudinal shiny strip from occiput to antennae, with slight pubescence between ocelli and antennae; face below antennae convex, slightly protruding the beyond the eyes in lateral view; mystax with white setae and smaller white setulae; ocellar tubercle with slight greyish pubescence, bearing two white setae pointing anteriorly; hairs on face and frons white; occiput with many with long white to yellow setulae, including some heavier yellow setae dorsally; antennal sockets located at one fourth from top of head, scape and pedicel black with very light greyish pubescence; antennal postpedicel black with slight reddish tinge; antennae bare, except antennal pedicel with several white setae; postpedicel distinctly longer than scape and pedicel combined. Small sensory element placed in a small cavity at the apex of the postpedicel; proboscis black with white setulae, wide on its basal part and constricted at the apex, flattened dorsoventrally; proboscis without stout setae on middorsal margin; palpi twosegmented, both segments black, with white setulae on outer side, flattened laterally and leaf-like. -Thorax ( Figure 2): Scutum completely black with slight whitish pubescence, except for dark reddish anterior humeri; postsutural area of scutum between suture and posterior end of scutum, on each side of the dorsocentral bristles, shiny and without pubescence; presutural area of scutum, just in front of the suture to the posterior end of postpronotal lobe, on each side of the dorsocentral bristles, shiny and without pubescence. These four areas are bordered by areas with heavier pubescence. Median line of scutum very narrow, heavily dusted, bordered by two lines of shinier areas; area just anterior of scutellum heavily pubescent; scutum with many thin, semi-erect, white setulae. proepisternum, upper half of anepisternum, posterior margin of anepisternum, anterior margin of anepimeron, creating a reversed U-shape; meron and anatergite also with white pubescence; all pleura with white setulae, especially on top of the pubescent zones, except for the anatergite that hold a row of strong yellow setulae; halter yellow; postmetacoxal area not fused. -Legs ( Figure 2): Legs black with reddish tinge, especially on tarsi, tibiae and coxa; posterior surface of tibiae 1 and 2 deeply red-black, barely distinct to the black of anterior surface; coxa 1 and 2 heavily white pubescent with strong white setulae anteriorly; coxa 3 slightly pubescent with weak white setulae; trochanters 1-3 shiny with weak white setulae; femora enlarged, heavy and club shaped compared to slender tibiae; femora with several weak white setulae, except for the tip of femur 3 bearing several setae posteriorly; tibiae with several weak white setulae, in addition to several white setae distributed over length of tibiae; tip of tibiae 1 and 3 with two small, but strong setae ventrally; tarsi with several weak white setulae, but also several strong white setae, especially near the top of each tarsal segment; Pulvilli white; claws long, yellow basally, black on top half. Ecology: One specimen of the type series was caught in dense bushes in an otherwise bare landscape. Laphriinae are a dominant forest-inhabiting group of robber flies, distributed in wooded areas throughout the world (Fisher, 2009) of which females deposit eggs in openings of dead tree trunks and stumps, as the larvae will live in rotten wood feeding on wood-inhabiting species (Lavigne & Bullington, 1984). Therefore, the occurrence of Andrenosoma in a rather bare landscape is unusual. However, also Lehr (1958) found adults in such environments perching within shrubs and bushes.

Discussion
Morphology of male terminalia. Species of Andrenosoma are identified by abdominal colour patterns (ranging from entirely black to entirely red, with all intermediates), the construction of the first posterior cell (open, closed or stalked), colour of mystax, and especially by the morphology of the male terminalia. Practically every publication on Andrenosoma describes the gonostyli (= dististyli, surstyli or gonopods by some authors) in lateral view for species identification (Engel, 1930;Séguy, 1952;Richter, 1985;Fisher, 1986;Kovar & Hradsky, 1996;Geller-Grimm & Taylor, 1999;Maldès, 2004;Tomasovic & Ghagaei, 2009;Tomasovic & Bartolozzi, 2018;Garcia et al., 2018). While studying material of A. lottae sp. nov., problems arose to annotate two structures of the male terminalia as the gonocoxite bears an inner, moveable, lateral process (which is relatively concealed in lateral view of the gonocoxite) as well as an outer, non-moveable, subapical process (which is well visible in lateral view of the gonocoxite). Initially, it was difficult to annotate these structures -which of these structures is the gonostyli? For A. lottae sp. nov. and A. farsicola we considered the inner process as the gonostyli as this structure is moveable, whereas the fixed process is considered a secondary distal lobe of the gonocoxite. For A. lottae sp. nov. and A. farsicola the secondary distal lobe is easily seen from a lateral view and entomologist might erroneously annotate this structure as gonostyli. Remarkably, for other species of Andrenosoma we could examine (i.e., A. atra, A. albibarbe, A. aff. albibarbe and A. cyrtoxys) it was clear both inner and outer structures of the gonocoxite are articulate and moveable, which is different compared to A. farsicola and A. lottae sp. nov. in which the outer structure is fixed.
In all literature that depicts male terminalia of Andrenosoma, only the outer structure on the gonocoxite is drawn (and annotated as gonostylus), whereas the inner process is rarely depicted. Only Kovar and Hradsky (1996) and Richter (1985) illustrate a small segment of the 'inner process' that protruded from behind the 'outer process'; and Fisher (1986) who successfully illustrates the entire shape of the inner process of A. atra. As both inner and outer structures are moveable, it remains unanswered whether we can annotate the 'outer process' as gonostylus in these taxa. Careful interpretation of the figures and annotations in literature remains essential for identification of Andrenosoma and a revision of Palaearctic Andrenosoma is necessary in order to detect whether authors draw the inner or outer structure on the gonocoxite. We assume the inner process is of value for species identification. Relationship to other species of Andrenosoma. Andrenosoma lottae sp. nov. is clearly related to A. farsicola, based on male terminalia: both species have a unique epandrium that has an pointy ventro-distal end and a fixed secondary distal lobe of the gonocoxite. Furthermore, both species occur in the same steppe-like habitat, which is distinctly different to other (Palaearctic) Andrenosoma. It can be argued these differences are of generic value, but again, only a revision of Palaearctic Andrenosoma might answer such question. Andrenosoma lottae sp. nov. differs to A. farsicola by the shape of gonocoxite, the shape of the secondary distal lobe, the shape of the gonostyli, by the smaller size (10m in A. lottae sp. nov. to 19 mm in A. farsicola) and by the entire red abdomen (black in A. farsicola).
Andrenosoma valentinae Richter, 1985 is especially related to A. lottae sp. nov. based on the abdominal colour patterns: both species have red tergites and red sternites, which is unique for Palaearctic Andrenosoma. It is remarkable that Richter (1985) described its genitalia to be rotated 180°, which is typical for Mactea . Mactea is closely related to Andrenosoma and occurs in the far east of the Palaearctic realm and extends southwards into the Indo-Malayan region (i.e., Japan, China, Russia Far East, Taiwan, and Indonesia) (Hradsky & Geller-Grimm, 2003). The differences between both genera are listed in Hradsky and Geller-Grimm (2003), Lehr (1989Lehr ( , 1999 and  and can be summarised as such: the apical end of proboscis is bent apically in the top third [only slightly turned upwards or even straight in Andrenosoma], the male genitalia are rotated 180°C (45-90° in Andrenosoma), cerci and epandrium fused (separated in Andrenosoma). Andrenosoma valentinae is found close to Iran, just north of Tabriz. A translation of the Russian description of A. valentinae Richter, 1985 is given in the Supplementary Annex 1.
Finally, the closed and stalked first posterior cell is an interesting character, only shared with A. atra and A. farsicola. The first species is a widespread species in Europe, the second is an Iranian endemic, only known from Jahrom, Kerman province. Both species are characterised by the completely dark abdomen, distinctly different to A. lottae sp. nov. Taxonomic status of Andrenosoma violacea. During the comparison to species of Andrenosomatini, we were unable to study nor interpret the species concept of Andrenosoma violacea (Fabricius, 1777), a species not recorded since its description. This enigmatic species is described from Hattorf, central Germany, as Asilus violacea Fabricius, 1777. A small part of the description was later copied into Fabricius (1781), and later again copied into Fabricius (1794). This triplicate description, dealing with the same species, resulted in a widespread mistake: practically all sources refer to Andreno-soma violacea (Fabricius, 1781) [e.g., Engel (1930 or Andrenosoma violacea (Fabricius, 1794) [e.g., Von der Dunk, 2007; and not to Andrenosoma violacea (Fabricius, 1777). The description by Fabricius (1777) reads as: "Hirsutus ater, abdomine violacea. Corpus totum pilis erectis rigidis hirtum, atrum. Abdomen ovatum, violaceum, nitidum. Alae fuscae", translated as "Black setae, purple abdomen. Entire body covered with erect, stout black setae. Abdomen oval (?), purple, shiny. Wings darkened". The description by Fabricius (1781) reads as: "Hirßutus ater, adbomine violacea", translated as "Black setae, purple abdomen". The description by Fabricius (1794) Meigen? See Pont (1996)] and Meigen found a very close relationship between Andrenosoma atra and A. violacea, but neither suggest synonymy. Séguy (1952) lists Laphria violacea Meigen [no date given] as a synonym of A. atra. This is remarkable, as Meigen never published a species named L. violacea. We assume Séguy (1952) meant to say Laphria violacea Fabricius, a species that was synonymized by Meigen. In that case, Séguy (1952) misinterpreted Meigen, as Meigen did not synonymized these species. Finally, we could not recover the source for the generic switch to Andrenosoma. All of this suggest a troublesome history that can only be solved through examination of type material. Examining type material is troublesome, as Zimsen (1964) published on Asilus violacea Fabricius, 1777 that only a name label exists in the Fabricius' collection, and no specimen is found (which is confirmed by the current currators, T. Pape and M. Kuhlmann, pers. comm. 2021). Without type, it is not possible to interpret Fabricius' short descriptions.
More recently, Von der Dunk (2007) implies a synonymy of Andrenosoma violacea (Fabricius, 1794) [the reference should be Andrenosoma violacea (Fabricius, 1777)] with A. atra; and says the name A. violacea has no validity. Von der Dunk (2007) does not gives references for these statements, but we assume this is based on Meigen (1820) which is, yet again, misinterpreted. Moreover, it remains troublesome von der Dunk (1996) presents a key to Asilidae and presents two previously unknown characters for A. violacea: the 1 st posterior cell closed and size maximum 12 mm (at this time, the type material was already lost). Yet again, Von der Dunk (1996) is not giving references nor is he listing his studied material. We were unsuccessful to contact Von der Dunk to inquire about these facts. As nothing else is known on this enigmatic species, we consider this name a species inquirenda, which article 75.5 of the ICZN restricts to a nominal species-group taxon that cannot be determined from its existing name-bearing type; and not a synonym of A. atra, as not a single character can be interpreted correctly.

Supplementary Material
Supplementary Material is given as a Supplementary Annex, which is available via the "Supplementary" tab on the article's online page.