Taxonomy of the poorly known Quedius mutilatus group of wingless montane species from Middle Asia (Coleoptera: Staphylinidae: Staphylinini)

. The Quedius mutilatus group, a very poorly known presumably monophyletic complex of wingless, possibly hypogean species confined to the Tien-Shan Mountains, is characterized as such for the first time. Newly available material clarified the identity of Q. mutilatus Eppelsheim, 1888 and Q. kalabi Smetana, 1995, each hitherto known from a handful of non-conspecific and vaguely georeferenced specimens only. Additional material is reported for Q. equus Smetana, 2014 and one species, Quedius kungeicus sp. nov., is described. All available data on the taxonomy, distribution and bionomics for all these four species of the group are summarized.


Introduction
The mega-diverse rove beetle genus Quedius Stephens, 1829 includes a small number of hypogean species with characteristic morphological modifications such as: notably reduced eyes, weaker cuticular pigmentation, dorso-ventrally flattened body with long legs, pronotum strongly narrowing anteriad, as well as reduced non-functional wings with corresponding shortened elytra and absence of a palisade fringe on the abdominal tergite VII.Hypogean species of Quedius are phylogenetically heterogeneous, while their distributions, bionomics and sister group relationships remain largely unexplored (Solodovnikov & Hansen 2016).The poor systematic knowledge of these specialized Quedius is unfortunate, as the genus represents an interesting model to study speciation within the context of mountain orogenesis.In this respect, the similar species Q. mutilatus Eppelsheim, 1888, Q. kalabi Smetana, 1995and Q. equus Smetana, 2014 (grouped here as the 'Q.mutilatus group') of the subgenus Microsaurus Dejean, 1833 seemed noteworthy, as they were the only members of the genus with such distinct morphological traits in the Tien-Shan, a vast mountain system of the Palaearctic region.As expected, the Q. mutilatus group was hitherto known from very scarce, mostly non-georeferenced material lacking bionomic data.Our examination of various collections of Quedius yielded new material that was collected over many decades and that sheds light on the composition, distribution and bionomics of this peculiar Asian species group.This paper reports these findings with, among other discoveries, a species new to science.

Material and methods
This paper is based on the examination of material from the following institutions and private collections: The holotype of the new species is deposited at ZIN.
Morphological examination and preparation of specimens were conducted using Lomo MSP-2 v. 2 and Leica M125 stereo microscopes.Measurements were recorded at 4.5 × magnification using an ocular micrometer.
Abbreviations: EL = length of elytra (from humerus to the most distal part of the posterior margin) EW = width of elytra (maximum, with elytra closed along suture) FB = forebody length HL = head length (from base of labrum to neck constriction along the head midline) HW = head width (maximum, including eyes) PL = pronotum length (along midline) PW = pronotum width (maximum) Overall body length was measured from the apex of the labrum to the apex of the abdomen.All measurements are in mm unless otherwise indicated.
The photographs of the habitus and aedeagi were obtained with a Nikon D700 digital SLR camera attached to a Nikon SMZ 1500 stereo microscope.Photos and line drawings of the male genitalia are based on preparations of their soft parts in glycerin (obtained via dissecting, maceration in 10% KOH and rinsing in distilled water).Pencil sketches for subsequent computer-based inking were made using a drawing tube attached to a Nikon SMZ 1500 binocular microscope.
The species distribution map in Fig. 1 was produced with QGIS 2.12.0 free software.It is based on the approximate geographic coordinates that we were able to find with the aid of various printed maps or online systems (Google Maps, Google Earth and others) for the toponyms indicated on the specimen labels.Our friend and colleague Ilya Kabak (St.Petersburg, Russia) kindly communicated the coordinates for the specimens he collected.Since none of the geographic coordinates are indicated on the original locality labels for any of the examined specimens, they are given in square brackets in the 'Material examined' sections.Specimen labels with very ambiguously indicated localities are cited verbatim and given between quotation marks.All our interpretations of such labels are given in square brackets.

Diagnosis
The Quedius mutilatus group of species can be recognized as follows: relatively large Microsaurus (body length 8-11.6 mm) with brown to dark brown, dorso-ventrally flattened body, with notably small eyes and pronounced signs of brachyptery such as short elytra and absent palisade fringe on abdominal tergite VII; head with posterior frontal puncture situated very close to neck; pronotum with distinct anterior and posterior angles, normally with three punctures in dorsal rows with anteriormost (smaller) puncture located latero-anteriorly at pronotal anterior margin, and with one to two sublateral punctures situated slightly anterior to large lateral puncture; scutellum impunctate; aedeagus robust, symmetrical, with apical portion of median lobe slightly curved towards paramere, with characteristic tooth near apex (in lateral view), with paramere widest shortly before apex (in parameral view) and with four distinct groups of sensory peg setae: two apical and two lateral.

Distribution
All species of the group seem to be allopatric (Fig. 1): Quedius mutilatus Eppelsheim, 1888 is restricted to the central part of Terskei Alatau, south of Issyk-Kul Lake; it is replaced eastwardly by Q. kalabi Smetana, 1995 andthen Q. equus Smetana, 2014, the latter probably being distributed throughout NE Terskei Alatau to Xinjiang Province in China.Quedius kungeicus sp.nov. is known from the type locality in eastern Kungei Alatau only.There is a single female from the high altitudes of Atbashi Mountain in Central Tien-Shan, which presumably belongs to the Q. mutilatus group, but cannot be further identified without associated males.Three ambiguous specimens from Toksanbai in Dzungarian Alatau, damaged by dermestids, are presumably mislabelled and in fact originate from Terskei Alatau, because a single fully preserved male among them displays an aedeagus of the Q. kalabi type.All specimens of the Q. mutilatus group with recorded bionomic data indicate a specialization to high elevations.

Systematics
Some authors affiliated Q. mutilatus with Q. przewalskii Reitter, 1887(Reitter 1887;Boháč 1988) or with the Q. przewalskii group (Smetana 2014); the latter, however, lacks a clear definition (Solodovnikov & Hansen 2016).For example, Smetana (2014) affiliated Q. mutilatus and Q. equus to the Q. przewalskii group, contrary to his (e.g., Smetana 2001) earlier opinion of the composition of this group.Based on body chaetotaxy and the shape of the aedeagus, it could be related to Q. puncticollis (Thomson, 1867), as proposed by Coiffait (1978).The Q. puncticollis group of Coiffait (1978) is phylogenetically heterogeneous, based on the diverse morphology and scattered distributions of species.We establish the Q. mutilatus group here for Q. mutilatus and related species from Tien-Shan.Members of the group SALNITSKA M. & SOLODOVNIKOV A., Taxonomy of the Quedius mutilatus group markedly differ from Q. puncticollis and related species in a microphthalmous, brachypterous habitus, with weakly pigmented coloration and in the shape of the aedeagus, with a pointed apex of the median lobe and sensory peg setae arranged in apical and lateral groups.The Q. mutilatus group differs from the Q. przewalskii group, in the more restricted sense of Smetana (2001), in having three (contrary to one or two) punctures in the dorsal rows of the pronotum and in the structure of the aedeagus, which is symmetrical, with a median lobe having a pointed apex and a more basally positioned ventral tooth, and with wide parameres having two groups of sensory peg setae (apical and lateral).
All species of this group are individually treated below.In addition to the description of a new species, a redescription is provided for Q. mutilatus, the most commonly cited species of the group with which other, later described species have been compared.Eppelsheim, 1888 Figs
Body.Length 8-12 mm; brown to light brown including appendages, elytra and abdominal tergites, becoming slightly paler toward apex, sometimes head and abdomen slightly darker than pronotum and elytra; distinctly flattened dorso-ventrally.
European Journal of Taxonomy 401: 1-17 (2018) Head.About as long as wide [HL/HW: 0.93-1.22(1.03)], with broadly arcuate posterior angles; surface with distinct microsculpture of transverse waves on disc, becoming more or less isodiametric on frons between eyes, and with distinct sparse non-setiferous micropunctation.Eyes small and flat, not protruding over lateral contour of head; temples about 2.00-3.18(2.54) times as long as longitudinal diameter of eye.Head disc on each side with setiferous punctures as following: anterior frontal puncture at inner margin of eye, posterior frontal puncture very close to neck, and pair of smaller vertical punctures immediately medio-posteriorly from the latter puncture.Each temple with two temporal punctures of which posterior temporal puncture closer to posterior margin of head than to posterior margin of eye; additionally, temples with some fine setiferous punctures bearing short pale setae.aBdoMen.With fine and dense punctation; interspaces with minute irregularities; posterior margin of tergite VII without palisade fringe (Fig. 2A).
Male secondary sexual cHaracters.First three segments of front tarsus more strongly dilated than in females; second segment slightly wider or as wide as apex of tibia.Sternite VIII with distinct roughly triangular medio-apical emargination; tergite X triangular, strongly narrowed apically with few strong and numerous smaller apical setae; sternite IX moderately elongate with wide and glabrous basal, and sparsely setose and vaguely bilobed apical portion.aedeagus (Fig. 2B).Median lobe: (in parameral view) parallel-sided, with obtusely pointed apex; (in parameral or lateral view) with minute tooth situated near its apex.Paramere (in parameral view) as wide as median lobe, lanceolate, with slightly incised apex; underside with irregular groups of sensory peg setae, including a pair of apical groups with ca 3-7 peg setae in each group, and a pair of lateral groups with ca 4-6 peg setae each; additionally, parameral apex with normal setae including pair of apical setae on each side of apical incision and ca 1-4 lateral setae on each side below.Internal sac without large, strongly sclerotized structures.Female secondary sexual cHaracters.First four segments of protarsi dilated, but slightly narrower than in males.Tergite X triangular, gradually narrowed apically, pigmented and with long setae medioapically.

Distribution and bionomics
Quedius mutilatus is endemic to the central Terskei Alatau mountain range, where it is known from Karakol Gorge in the west (type locality) to the Chon-Kyzyl-Suu River in the east.Examined material has been collected from the second half of May through the end of August, all at high elevations around 2000-3000 m.At least some specimens without bionomic records were presumably found under stones as a by-catch by various entomologists, including specialists targeting Carabidae and Elateridae who, as a rule, intensively turn stones at elevations above the timber line.A few specimens with recorded bionomics have been found in coniferous leaf litter and other ground-based debris like grass or moss.2018) Some specimens from NMW bear clearly written, but ambiguous geographical labels with 'Turkestan Akinin-Tepe', which we could not interpret in spite of consulting historical literature about I.Y.Akinin's travels in Middle Asia (Jacobson 1902).Based on the structure of the aedeagus, though, these specimens clearly belong to Q. mutilatus.

Comparison
Quedius mutilatus differs from Q. kungeicus sp.nov.by the rhomboid (as opposed to ovoid) apical portion of the paramere and a slight, but distinct incision on its apex (as opposed to indistinct incision; aedeagus in parameral view), as well as the less curved apical portion of the median lobe, with stronger ventral sub-apical tooth (aedeagus in lateral view).From Q. kalabi it differs by the more robust and less curved apical portion of the median lobe with its shorter apical part (aedeagus in lateral view) and, usually, by the larger number of sensory peg setae in lateral groups on the paramere.From Q. equus it differs by the less incised apex of the paramere and a distinctly greater number of sensory peg setae in the lateral groups on the paramere.

Notes on the type material
Quedius mutilatus Eppelsheim, 1888 was described based on two specimens with unclearly recorded collecting localities.Of these, a male comes from 'Fluss.Tamgi' which we interpret as River Tamga in the Karakol Gorge of Terskei Alatau, while a female comes from 'Lake Issyk-Kul', an even less precisely outlined area around a large lake in Kyrgyzstan.Smetana (1998), who designated the male as a lectotype, interpreted the female paralectotype label literally as the Lake Issyk-Kul at ca 1500 m of elevation and suggested that, unlike the high altitudinal Q. kalabi, Q. mutilatus is confined to lower elevations.However, such an interpretation is not confirmed here with new and better georeferenced material.Before Smetana's lectotype designation and his latest taxonomic treatment of this species, the male syntype was examined by Gridelli (1924), who noted some morphological characters, and by Coiffait (1978), who first illustrated the aedeagus for Q. mutilatus.
It is noteworthy that the paralectotype female of Q. mutilatus from 'See Isyk-Kul [Lake Issyk-Kul]' is teneral, slightly smaller than the lectotype and notably smaller than the size range of the non-type specimens.Given its small size and a lack of precise locality information, the species identity of the paralectotype of Q. mutilatus remains unclear.Therefore, we list it among other undetermined material of the Q. mutilatus group at the end of this paper.Smetana, 1995 Figs 1, 3, 4C-D Quedius kalabi Smetana, 1995: 77 (original description).

Distribution and bionomics
Previously, Q. equus was known from the holotype and two paratypes (all males) only, collected together at the type locality in Xinjiang Province in China.New material examined here matches well with the original description and significantly expands the distribution of this species westwards (Fig. 1).All newly studied specimens were collected at high elevations around 2000-3400 m in July.For only two of them does the label specifies alpine meadows as the habitat, which is the same as for the type material.

Comparison
Quedius equus differs from all other species in the Q. mutilatus group by the deep incision on the apex of the paramere and by fewer (1-3) sensory peg setae in the lateral groups arranged in longitudinal rows.

Etymology
The specific epithet 'kungeicus' is a latinized adjective derived from the mountain range where this species was collected.
Body.Length: 10 mm; dark brown, hind margins of abdominal tergites slightly paler; appendages of the same coloration as body; body flattened dorso-ventrally.Overall external morphology as in Q. mutilatus, but eyes slightly protruding over lateral contour of head (Fig. 5 A). 4 e-F, 5B-d).Apical portion of median lobe (in lateral view) with characteristic 'hump' (shown by arrow in Fig. 5B) more pronounced than in other species of the group and with relatively more elongate apical portion that does not have a pronounced ventral sub-apical tooth; paramere (aedeagus in parameral view) as wide as median lobe, apically ovoid and with hardly visible apical emargination; parameral underside with apical groups of ca 7-8 sensory peg setae on each side of emargination and with two lateral groups of 2-3 peg setae each; parameral apical contour with two pairs of apical setae and two pairs of lateral setae.Internal sac without strong and obvious sclerotized structures.

Female
Unknown.

Distribution and bionomics
Quedius kungeicus sp.nov. is known from the holotype only, which was collected in the Kungei Alatau mountain range at an elevation of around 2600-2700 m in spruce forest.According to Ilya Kabak, who collected the holotype of Q. kungeicus sp.nov., it was most likely taken from under stones.This and the habitus of the species suggest an association with talus.For reasons why the paralectotype of Q. mutilatus and the paratype of Q. kalabi, both females, are likely non-conspecific with the primary type material and cannot be identified, see the comments for the respective species above.
Three specimens from Dzungarian Alatau in ZIN are damaged by dermestids and among them only one of the two males has the aedeagus intact.Based on external morphology, these specimens, collected during the same collecting event, appear conspecific.The structure of the aedeagus preserved in one of them suggests they may be conspecific with Q. kalabi or at least represent a very similar taxon.This, in turn, suggests their origin from Terskei Alatau and therefore a case of mislabeling.The presence of the Q. mutilatus group in Dzungarian Alatau, which is outside the known distribution pattern of the group, needs confirmation by other material.
A female specimen from NMW from an indefinite locality recorded as 'Turkestan', and a female from ZIN from the Atbashi mountain range cannot be identified further than being members of the Q. mutilatus group without associated males.It is noteworthy that a female from the Atbashi Range so far represents the southwesternmost and a relatively isolated record of the Q. mutilatus group.

Discussion
The shared, very uniform external morphology (similar habitus and chaetotaxy), shape and chaetotaxy plan of the aedeagus, as well as biogeographic considerations suggest that the Q. mutilatus group is monophyletic.The phylogenetic affinities of this species group within the subgenus Microsaurus remain unclear.Beetle morphology and the currently known distribution records suggest that the Q. mutilatus group consists of hypogean (sensu Solodovnikov & Hansen 2016) allopatric species.Both assumptions, however, need to be explored and confirmed by robust data, currently unavailable because the group has very rarely been collected.At present, four species of the Q. mutilatus group can be separated SALNITSKA M. & SOLODOVNIKOV A., Taxonomy of the Quedius mutilatus group from each other by clear morphological differences in the structure of the (Fig. 1).The examined material displays notable intraspecific variability, suggesting that species-level taxonomy and the distribution pattern of the Q. mutilatus group may appear more complex with denser sampling.In future cases of specimens where morphology and biogeography would not give a clear answer of their species identity, our suggestion would be to downgrade current species of Q. mutilatus group to the subspecies level and treat ambiguous material as Q. mutilatus ssp.incertae sedis.Such an approach would allow effective documentation of new material until the next level of taxonomic understanding is reached.We would like to stress that our current, exclusively morphology-based species concepts at most set an initial taxonomic frame for a deeper phylogeographic investigation of the Q. mutilatus and similar groups of montane species.
cRyv = Private collection of A.B. Ryvkin, Moscow, Russia cSch = Private collection of M. Schülke, Berlin, Germany (to be deposited at the Natural History Museum of Berlin) NMW = Natural History Museum, Vienna, Austria (H.Schillhammer) ZIN = Zoological Institute, Russian Academy of Sciences, St. Petersburg, Russia (B.Korotyaev) ZMUC = Zoological Museum, University of Copenhagen (part of the Natural History Museum of Denmark), Copenhagen, Denmark (A.Solodovnikov, S. Selvantharan)

Fig. 1 .
Fig. 1.Distribution and aedeagus variability of the Quedius mutilatus group in the Tien-Shan Mountains.Numbers are specified in Table1.Empty symbols = type spcimens (for Q. kalabi precise locality of the holotype unknown); ?= ambiguously labelled or undetermined material.Scale bar: 0.5 mm.

Table 1 .
Geographical data for material mapped in Fig.1.Geographic coordinates found for localities indicated on the labels are given in the square brackets, those communicated by collectors without brackets.Segments as following: second and third combined as long as first; third slightly longer than second; fourth and fifth, each, as long as wide; sixth to tenth, each, wider than long and gradually widening towards apex of antenna; last segment slightly shorter than two preceding segments combined.