Scenopinusjerei, a new species of window fly (Diptera, Scenopinidae) from Finland

Abstract A new species of window fly (Diptera: Scenopinidae), Scenopinusjereisp. nov., with characteristic bicoloured legs and completely black halteres, is described from Finland. To exclude potential previously named species, a survey of the relevant type specimens as well as original descriptions of the Palearctic and Nearctic Scenopinus species has been conducted, including old Scenopinusfenestralis (Linnaeus) synonyms. Scenopinusjereisp. nov. is likely to be an overlooked, boreal forest specialist living in the nests of cavity-nesting birds. An identification key to the European species is provided.

In this paper we finally provide a formal description of the previously undescribed species as well as information about its distribution and biology. We focus on differentiating the species from Scenopinus fenestralis and S. vitripennis, as these are most likely to be confused with the new species due to the variability of all three species as well as because of the diagnostic characters used in older literature. We also compared it with written descriptions of other known species in the Holarctic and the old synonyms of Scenopinus fenestralis to rule out existing names for the candidate species. Finally, a key for the identification of the European Scenopinus species is provided.

Material examined
Apart for the two old museum specimens of Scenopinus vitripennis, most of the examined material were collected relatively recently by the authors and their close associates. Except for the reared specimens, majority of the examples have been collected indoors (see notes) directly to vials and killed by freezing, ethyl acetate or potassium cyanide prior to mounting them on entomological pins.
Label data of newly collected specimens are given verbatim using the following symbols: / end of a line and beginning of the next; // end of label and beginning of the next (from top to bottom on the same pin). The specimens are deposited in the following collections and are indicated with the given acronym in the text: Male terminalia were dissected and prepared for examination essentially as described by O'Hara (2002). The dissected terminalia are preserved in glycerol in a small plastic vial pinned together with the specimen. Location abbreviations refer to the geographical provinces (e.g., Kb = Karelia borealis, explained when mentioned for the first time, see https://laji.fi/theme/emk for more details) and coordinates (NNNN:EEEE) on the labels are mostly given in the old national Finnish map grid coordinate system (YKJ; see Ollikainen and Ollikainen 2004), which is still in common use for biological sampling. As the first number of the longitude coordinate is always 3, this is often left out in the collection labels to save space. Sampling coordinates in decimal degrees as well as additional notes or commentaries about the specimen are given in brackets.

Classification and terminology
The classification follows Herting and Dely-Draskovits (1993). The morphological terminology used in this study follows Cumming and Wood (2017), except for the features of male terminalia, where Winterton and Gaimari (2017) is used.

Microscopy and imaging
The images were taken with a Leica Z6APO stereomicroscope and a Leica DFC450c (5MPix) camera, MSV266 motorised focus and using the Leica Application Suite 4.6.0 software for Z-axis stacking. Images were cropped, colour-and contrast-enhanced but not manipulated otherwise.

DNA extraction, PCR, and sequencing
Cytochrome oxidase subunit 1 (COI) DNA barcoding was performed as a part of the Tachinidae project of Finnish Barcode of Life initiative (FinBoL). The 5´-terminal part of COI was amplified using the routine barcoding primers LepF1 and LepR1 (Hebert et al. 2004). The sample identifiers in the barcode of life database (BOLD) are given for each barcoded specimen.

Sequence comparisons and COI tree
Sequence comparisons were performed using MUSCLE alignment (Edgar 2004) and Bayesian inference phylogenetic tree generated using MrBayes 3.2. (Ronquist et al. 2012), applying GTR substitution model with gamma-distributed rate variation across sites and a proportion of invariable sites, and 1,000,000 MCMC generations. The tree was visualised using FigTree 1.4.4. (Rambaut 2009).

Results
We report here a new species of window flies, Scenopinus jerei sp. nov. from Finland based on the following material and diagnostic characters. Diagnosis. Scenopinus jerei sp. nov. belongs to the S. fenestralis group and is easily recognisable from the other species in this group based on the contrasting colour differences between the femora and the yellow to orange tibiae. The coxae as well as the knob of the halteres are always uniformly black or dark brown, similar to the colour of the thorax. Head ( Fig. 2A, B). Black with greasy-looking shine, including oral margin and occiput, apart from weak grey microtomentum around occipital foramen, mouth edge and antennal base; semi-circular, height 1.5-1.7 [1.6] (n = 6) × its maximum width in lateral view. Antennal insertion slightly below mid eye level. Antenna dark brown with pedicel and anterior part of flagellomere paler; scape short and subrectangular; pedicel short and cylindrical, [0.8]-1.0 (n = 6) × as long as wide, flagellomere laterally flattened, 1.8[1.9]-2.0 × as long as high, subrectangular, narrowing apically and 4.8-[5.4] (n = 6) × as long as pedicel and with subcircular, subapical, sensory pore on outer side. Eyes large and bare; fronto-orbital plates meeting at [0.25]-0.3 (n = 6) length of frons; no frontal vitta; gena reduced to narrow strip between lower eye margin and mouth edge. Diameter of ommatidia on upper half of compound eye, above antennal base, 2-3 × diameter of ommatidia on lower half. Ocellar triangle acute, distance between posterior ocelli distinctly shorter than their distance to anterior ocellus. Frons bare but     Krivosheina (1981). patterned with minute pits. No setae or setulae on head, apart for short brown setulae at lower posterior part of gena behind mouth edge. Mouthparts, including palpus, black.

Scenopinus jerei
Thorax (Fig. 1A). Dorsally and laterally black with greasy-looking shine. Scutum patterned with small rugae and minute, barely distinguishable setulae. Pleura with similar patterning but with more distinct, short, sparse, greyish to brownish setulae. Hirsuteness most developed on anepisternum, where longest setulae are approximately same length as width of flagellomere.
Legs (Fig. 1A). Coxae black; femora brown and apically paler. Fore and mid tibiae [pale brown] to dirty orange, clearly paler than femora. Hind tibia otherwise of similar colour as femora but paler at base and apex. Hind coxa with thin black posterior setulae, longest setulae as long as width of coxa at its base. Femora with thin posterodorsal setulae, longest being 0.5 × as long as width of femora. Fore tibia preapically with [2]-3 ventral setulae. Mid tibia with two short, ventral, preapical setae and two adjacent setulae. Hind tibia with one ventral preapical setula and thin posteroventral setulae covering proximal half, longest of which are as long as width of tibia. Apart from aforementioned setae and setulae, all legs covered in minute setulae that provide rugous texture.
Female (Figs 1B, 3A, B). Differs from male as follows: Body length: 4.3-5.7 mm (n = 4). Head (Fig. 3A, B). Frons broad, at its narrowest point 0.62-0.69 (n = 4) × as wide as an eye in dorsal view. Frons shiny black with minute longitudinal rugae on frontal stripe and transverse rugae on sides. Orbital plates smooth and shiny. Ocellar triangle equilateral. No obvious size difference between ommatidia of upper and lower half of compound eye. Thorax (Fig. 1B). Very weak whitish grey microtomentum at anterior parts of postpronotum and proepisternum. Abdomen (Fig. 1B). Dorsally flattened along its entire length. Terminalia. Last visible tergite 9 bluntly triangular at its posterior edge and not divided into hemitergites.
DNA barcode divergence among Scenopinus. Scenopinus spp. are poorly covered in the DNA barcode databases, such as Barcode of Life Database (BOLD, www. boldsystems.org) or GenBank. It is noteworthy that all S. fenestralis specimens in the databases from Europe to North America have almost identical COI sequences and represent the same barcode index number (BIN). The DNA barcode of Scenopinus jerei sp. nov. differs markedly from the other northern European species, its closest match being Scenopinus fenestralis from which it is separated by 12.48% sequence difference (Fig. 5). There are no other closer matches among the barcode sequences in the BOLD or GenBank.
Notes on the biology and distribution of Scenopinus jerei sp. nov. The larvae of Scenopinus species are predators of other invertebrates living in dry organic substrates, such as in animal nests. In Finland, Scenopinus jerei sp. nov. has been collected inside sheds, attics and indoor storages as well as reared from nest boxes of birds. These rearings produced large numbers of tineid moths (Lepidoptera: Tineidea), especially Monopis laevigella (Denis & Schiffermüller), but also other Monopis spp., Niditinea striolella (Matsumura), and Tinea spp. Other insects observed from the same nest boxes included Ceratophyllus fleas, various beetles (Histeridae, Dermestidae) and flies (Piophilidae, Fanniidae, Heleomyzidae). Apart for two male specimens found dead on a windowsill in an attic of an old house in Kelovaara on July 24 (see type specimens), most observations are from third week of June. According to the observations of Jere Kahanpää (pers. comm.), Scenopinus jerei sp. nov. hibernates as full-grown larvae and the adults emerge in a couple of weeks in room temperature rearing conditions. Based on the collection locations, it is likely that Scenopinus jerei sp. nov. is a boreal forest specialist. Like other Scenopinus spp., Scenopinus jerei sp. nov. is not very active flier, does not visit flowers and therefore is rarely collected by active netting or traps. Judging from the few Finnish observations, the species appears widespread in the southern and central parts of the country. We are certain that Scenopinus jerei sp. nov. can also be found in boreal forest biotopes in the other Nordic countries and Russia but has been until now overlooked.
Etymology. This species is named after Mr. Jere Kahanpää, Helsinki, who was to first to discover that the taxon is new to science and kindly agreed with the current arrangement for its formal description.

Provisional key to the identification of European Scenopinidae species
Because the existing literature on the European species of Scenopinidae are outdated or difficult to obtain, we felt necessary to provide a key for the known European species of Scenopinidae. We must emphasise that we have been only able to examine the species with specimens listed in this paper, for which the identification key should work well. For the remainder, our approach was to go through the written species descriptions and pick features which we judged, by our collective species identification experience, to be useful for determination. To us this approach was better justified than reproducing the keys given in earlier literature, which are often difficult to follow or focus on limited number of poorly defined features. The diagnostic features for the key have been obtained from the descriptions in Kröber (1925), Trojan (1956), Kelsey (1969), Krivosheina (1981), Narchuk (1988), and Carles-Tolrá (2001). Fortunately, the European species separate into three easily recognisable species groups, each with relatively few species. The species groups appear in the key in alphabetical order, enabling fast navigation when one is familiar with the groups. Although result appears satisfactory, the key might not capture all the variations seen within each species and we strongly encourage DNA barcoding of specimens for future reference. In any case, we hope that this key can form a basis for forthcoming work with this interesting family of flies.

1
Cell r 5 closed and petiolate (Fig. 6B) Generally small in size (< 3 mm); R 4 branching from R 5 beyond the middle of cell r 5 (Fig. 6D); aedeagus in males not concealed by the epandrium ....albicinctus group 4 -Larger and more robust flies; R 4 branching from R 5 at or before the middle of cell r 5 (Fig. 6C)

Survey of candidate species among the Scenopinus fenestralis group
Scenopinus jerei sp. nov. was originally confused with Scenopinus vitripennis (Haarto 2000;Kahanpää and Winqvist 2005). The mistake occurred because Scenopinus jerei sp. nov. easily keys out as S. vitripennis using the key in Kelsey (1969), due to the dark halteres and relatively smooth female frons. As pointed out by Kelsey (1969), S. vitripennis was treated as a synonym of S. glabrifrons by many authors prior to Trojan (1956), who established it as a valid species based on female characters. In retrospect this conclusion is problematic, as the original type specimen of Meigen was a male and is presumed lost. At the time of Kelsey's work (1969), the type was the only reported male specimen of the species. Interestingly, there is one male identified in 1964 by Kelsey as S. vitripennis in SMNS. It is unknown to us why Kelsey did not include this specimen in his work on world Scenopinidae. The specimen is only 3.4 mm long, dark, has entirely grey-brown legs and, apart for some dirt and broken antennae, is in good condition (Fig. 1E). The colour of the legs, including the pale coxae, and general appearance of the specimen does not agree with our concept of Scenopinus jerei sp. nov. and neither does the examined female specimen in the ZIM collection (Fig. 1F).
Notably, the halteres of these specimens are grey or reddish brown, much paler than the thorax, whereas those of Meigen's description of the colouration of the legs as well as the underside of the head make it clear that S. vitripennis is not conspecific with our Scenopinus jerei sp. nov. However, these features are also not evident in the small, dark male specimen in the SMNS collection either. It may be that Kelsey disregarded this specimen from his work for the same reason. In fact, the male of S. vitripennis was later redescribed and illustrated by Krivosheina (1981). The shape of the antennae, terminal sternites as well as the visible genitalia features in Krivosheina's illustrations differ markedly from those of Scenopinus jerei sp. nov. However, the comparison with S. glabrifrons is not very detailed in Krivosheina's work. We also note that the female specimen in MIZ, collected and identified by Trojan, is very similar to S. glabrifrons in general appearance, including the flagellomere, which is cylindrical and parallel sided, compared to the basally more robust, apically narrowing flagellomere Scenopinus jerei sp. nov.
For some reason, the male genitalia of Scenopinidae have been traditionally dissected only partially, the only visible parts being the proximal parts of the aedeagus as well as the terminal segments, which makes the comparison of the published illustrations prone to interpretation errors. When fully dissected, the aedeagus has very distinct, species-specific features (Fig. 4), which could be better utilised for species identification. Unfortunately, S. vitripennis is very rare in collections, and clearly more work is needed to fully understand the extent of interspecific variation between it and S. glabrifrons. These future efforts should also focus on applying modern sequencing methods that allow the genotyping of old museum specimens (Staats et al. 2013;Prosser et al. 2016;Nakahama 2021).
To further validate our interpretation of the new species status of Scenopinus jerei sp. nov., we also checked the potential candidates among the known species of Scenopinus outside Europe. Since the revision of the world Scenopinidae by Kelsey (1969), Note the three stripes on the abdomen caused by the white integument protruding between the tergites. These should not be confused with white bands of microtomentum on tergites of some Scenopinus species. Photograph by J. Pohjoismäki B illustration of Caenoneura wing. Arrow pointing the petiole on r 5 . Drawn after Kelsey (1969) C generic Scenopinus wing. Modified from Winterton and Gaimari (2017) D wing venation in albicinctus-group with R 4 branching from R 5 beyond the middle of cell r 5 (arrow; compare with C) E Wing of Scenopinus griseus (Kröber) with narrow cu cell. Drawn after Narchuk (1988). relatively few Scenopinus species have been described (Pape and Thompson 2021), most of which can be excluded by their exotic location. However, as several Eastern Palearctic species in other families extend their range to Finland (e.g., Cannings and Kahanpää 2013;Stuke et al. 2020;Pohjoismäki and Bergström 2021), we wanted to rule out the following species from the Russian Far East: • Scenopinus mariensis Kelsey, 1981. This species is close to S. lesinensis but has notable colour patterning on the thorax.
• Scenopinus zhelochovtsevi Krivosheina, 1982. Legs uniform in colour. Illustrations of the antenna shape and male genitalia are dissimilar to those of Scenopinus jerei sp. nov.
No potential candidates were found among the known Nearctic species.

Survey of the old Scenopinus fenestralis synonyms
Scenopinus fenestralis is a common, variable, and widespread species. It also lacks clear morphological features, such as strong setae that are used as diagnostic characters for many fly groups, making it difficult to devise generalised descriptions or identification keys for both sexes. For example, the colour of the halteres can vary from dark brown to white and the colour of the legs from pale brown to orange. It is probable that developmental factors and possibly age of the fly play a big role in the morphological variation. Among others, it is known that hoverflies (Diptera: Syrphidae) developing in cool and humid conditions are typically darker than the ones developed in warm and dry conditions (e.g., Ottenheim et al. 1996). Similarly, small specimens developed under insufficient nutrition tend to be darker and morphologically more plain than larger specimens, which could also explain the aberrative habitus of the putative S. vitripennis male specimen in SMNS, mentioned earlier.
This intraspecies variation has contributed to the wealth of synonyms for S. fenestralis, which raised the question if some of the names could correspond to our concept of Scenopinus jerei sp. nov. However, checking all the old, synonymised type specimens dispersed over several collections would have been impossible. Fortunately, the colouration of Scenopinus jerei sp. nov. legs and halteres are so distinct that we were able to validate these diagnostic characters even from the quite brief old descriptions. Pape and Thompson (2021) list the following synonyms for S. fenestralis, whose original descriptions differ from the characters of Scenopinus jerei sp. nov. as indicated: • Musca tarda Linnaeus, 1761 -White halteres, yellow legs. • Musca saltitans Scopoli, 1763 -White halteres, red legs. • Musca spoliata Scopoli, 1763 -Male specimen of the latter. • Musca senilis Fabricius, 1794 -Legs yellow-red, head white from below. • Atricha fasciatus Schrank, 1803 -Greyish specimen with milk-white bands on the abdomen. While the abdominal banding sounds like an unusual feature for S. fenestralis, we have observed that the soft white integument can be protruding between the tergites in newly hatched Scenopinus specimens, giving the abdomen a banded appearance (Fig. 6A). Legs entirely olive brown.
• However, Kelsey attributes the synonymy to Schiner (1862), where S. fuscinervis Zetterstedt is given as a synonym of S. fenestralis. It is obvious that in this context, S. fuscinervis is a misspelling of S. furcinervis. In fact, the spelling is later corrected in Schiner (1864).
Based on this survey, we are confident that the species presented here as Scenopinus jerei sp. nov. is not among the accepted species nor hidden among the synonyms of S. fenestralis. We hope that the species discovery reported here, together with the provisional indentification key we have provided, will encourage more research towards this exciting but poorly known family of flies. suu, Finland) provided valuable laboratory assistance. We are grateful to Dr. Davide Badano (Sapienza Università di Roma) for the excellent editorial process, and Dr Steve Gaimari (CDFA, Sacramento, CA, USA) and Dr Paul Beuk (Natuurhistorisch Museum Maastricht, the Netherlands) for their constructive and highly insightful reviews, which helped to improve our manuscript. Finally, we would like to thank the Finnish Ministry of Environment for supporting our book project on the Asiloidea of Finland and adjacent countries.