DNA-barcode and endophallus morphology delimit congruent species in a systematic revision of the oxyporine rove beetles of Russia (Coleoptera: Staphylinidae: Oxyporinae)

Rigorous species delimitation is a challenge in biology and systematics in particular. In insects, male genitalia traditionally, and the barcoding region of the CO1 gene recently, are the main markers to identify species, even though a standalone use of CO1 for that is often criticized. In our systematic revision of the mycophagous and in other ways peculiar oxyporine rove beetles of Russia, the legacy alpha-taxonomy could not be improved by traditional investigation of genitalia as they are unusually character-poor in this group. Using phylogenetic inference and ancestral state reconstruction, we demonstrate that CO1 and endophallus are useful markers for species delimitation in Oxyporus. We also show that many Contributions to Zoology 90 (2021) 344-407 Downloaded from Brill.com07/14/2021 10:14:47AM via University of Copenhagen

morphological traits previously used for species delimitation in Oxyporus are in fact highly variable and

Introduction
Oxyporus Fabricius, 1775, taken in its broad sense with Oxyporus s. str. and Pseudoxyporus Nakane and Sawada, 1956 as subgenera, is the only extant genus of Oxyporinae, a rove beetle subfamily that is noteworthy for many reasons. Oxyporinae are a relatively small lineage of at least 110 described, peculiar looking extant species that are confined to Eurasia and the Americas (Herman, 2001;Newton et al. 2001;Löbl & Löbl, 2015 and papers after). The subfamily is strictly mycophagous, tightly associated with fungal fruiting bodies and, due to its largely unique morphological and biological traits that must be quite ancient (Cai et al., 2017), it does not have a clearly established sister group within the family Staphylinidae (Tokareva et al., 2020).
Poor knowledge of the global species diversity of Oxyporus is the first and most important obstacle to the broad scientific study of this genus, a common manifestation of the now so-called "Linnean shortfall" (Possingham et al., 2007). This shortfall is further amplified by the properties of the genus itself. Oxyporus are relatively large, mostly colorful and attractive beetles that are active in short-lived, patchily distributed fruiting bodies of fungi. As a result, they are popular among taxonomists but are also rather rarely-collected insects, where large comprehensive series of specimens are available only for some species. As most Oxyporus are strikingly colored, taxonomists have tended to focus on differently colored, easy to distinguish singletons, resulting in a number of species described based on color patterns without a detailed consideration of other characters. Makranczy (2012) strongly cautioned against taxonomy based on external characters (mainly coloration) alone, which has led to an alarming overestimation of species diversity in Oxyporus, especially in the fauna of the Old World. Following earlier examples (Campbell, 1969;Smetana, 1989;Hwang & Ahn, 2000), Makranczy also stressed the importance of the morphological variation in the complex endophallus, which compensates for the very simple and thus poorly informative shape of the median lobe of the aedeagus in this genus. This structure is promising for species delimitation within Systematic revision of oxyporine rove beetles in Russia Downloaded  Oxyporus, but it is also known to be notoriously difficult to evert (dissect) for study and is thus under-explored.
Our own attempts to identify species of Oxyporus from Russia, especially from its more diverse Far Eastern fauna using for example a recent review work by Shabalin (2012), raised the same concerns of Makranczy (2012). Several 'species' that Shabalin (2012) had recognized in the Russian Far East using the state-of-the-art taxonomic literature for the region, differed from each other almost entirely by coloration alone.
The overall shape of the aedeagus is usually considered as sufficiently diagnostic for species in entomology. Its documentation (outline drawings, later photos) is the mainstream in the taxonomy of beetles, and rove beetles in particular, for many decades. The endophallus also has been noted as a species-specific marker and used for precise identification of pest beetle species long ago (Barber, 1935). A more regular use of the endophallus in beetle taxonomy started to increase after the 1970s (Denis Kassatkin, pers. comm.). In the contemporary literature, endophallic characters are frequently used for species diagnostics and in many cases they are believed to be even more informative in this capacity than the external aedeagal structures (e.g., Petitpierre & Anichtchenko, 2018). However, compared to the external structure of the aedeagus, the endophallus is used much less commonly and much less consistently for species diagnostics because of its hidden position inside the median lobe of the aedeagus. The examination of these hidden structures demands a special technique for their eversion. The above considerations urged us here to thoroughly examine the fully everted endophallus of Oxyporus, to ensure accurate comparison of homologous structures among species.
As a very significant area of the Palaearctic region with diverse geography and climate, Russia has a sizeable Oxyporus fauna that is suitable for exploring species limits. It is a region that includes very widespread and far more locally distributed species of Oxyporus. By reconciling this general exploratory aspect with the practical goal of a regional taxonomic revision, we aim to demonstrate immediate utility of rigorous species-delimitation tools. Naturally, in the revisionary part, we consider material and data from countries adjacent to Russia as well. This material is especially relevant to the Far East, where the fauna of Oxyporus is more diverse.
The Russian fauna of Oxyporus has never been reviewed in its entirety. Current knowledge on Russian Oxyporus began with single records for the widespread species Oxyporus (Oxyporus) maxillosus Fabricius, 1793, O. (O.) mannerheimii Gyllenhal, 1827 and O. (O.) rufus Linnaeus, 1758, which were earlier described from Europe (according to the extensive literature summarized in Herman, 2001). The first species described from Russia was O. (Pseudoxyporus) dybowskii, (Solsky, 1871) from Baikal region, followed by the large and spectacular O. (O.) procerus described by Kraatz (1879) from the Russian Far East. Decades later, Bernhauer (1935) described O. (O.) aequicollis from the Baikal region while Kirschenblatt (1938) described O. (P.) cyanipennis from Altai and O. (P.) melanocephalus from the Russian Far East. The paper by Kirschenblatt (1938) was an important contribution to the knowledge of Russian Oxyporus because, apart from two newly described taxa, it provided a key to all species of the genus known from the Russian Far East at that time. The peculiar eastern species O. (P.) cyanipennis, O. (P.) melanocephalus and O. (P.) dybowskii were redescribed much later by Makranczy (2012). Apart from these notable two papers, Kryzhanovsky et al. (1973), Ryabukhin (1999) and Shavrin and Gildenkov (2009) published new records for various species of Oxyporus in Russia. Only many decades after Kirshenblatt's tokareva et al. (1938) summary, a new review of, and an identification key to, Oxyporus of the Russian Far East was compiled by Shabalin (2012), who reported 14 species of the genus from that region. Shabalin's key was essentially a key to all Russian Oxyporus because only the three eastern species (O. (O.)  Most of these records are scattered in numerous faunistic accounts of particular nature reserves or areas of similar size, and only some cover larger regions. Of the more inclusive faunistic works covering larger regions, those that provide original detailed occurrence data (e.g., Semenov et al., 2015;Semionenkov et al., 2015) are more informative than generalised area records (e.g., Gusarov, 1989;Solodovnikov, 1998).
Overall, the knowledge of Russian Oxyporus is patchy and not well connected to regional works on the same species in the adjacent countries of Eurasia. Available diagnoses and keys for Oxyporus of Russia are based mainly on rather polymorphic traits, such as body coloration, shape and size and only rarely involve structural external characters. Problems, caused by the use of such characters are particularly pronounced in the species of the nominative subgenus  Lee et al., 2020, which are supposedly distinguished only by the body size, punctation of elytra and the number of basal abdominal segments with pale paratergites.
The current taxonomy of Russian Oxyporus is also a snapshot of how the complexity of a species problem (Queiroz, 2007) and developing species delimitation methods can intersect to create confusion and uncertainty, especially when only limited individuals are available for study within local faunas, a common practical constraint in systematic entomology. For example, recent progress in algorithm-based, repeatable molecular-based species-delimitation brought a rigorous framework, but also new challenges related to the disparity of this newer phylogenetic approach and the older biological species concepts on which the majority of insect species were based (Lukhtanov, 2019). Another challenge of dna-based species delimitation stems from unequal evolutionary rates and conflicting gene trees between various loci. A number of methods have been suggested to account for these complications (Talavera et al., 2013) and a general consensus has emerged that no single method is entirely sufficient and that robust species limits are best recognized through the congruence of multiple sources of evidence, even if they are few (Pante et al., 2015). Modern integrative revisions in systematic entomology show that single-locus dna markers including the commonly used mitochondrial CO1 barcode reveal otu s that are congruent with a plethora of 'good species' recognized by traditional taxonomists (Kress et al., 2015). In beetles, these are mainly species delimited by the shape of male genitalia (Schmidt et al., 2015;Hendrich et al., 2014). In some insects, including many beetles, the species delimitation is complemented by the study of the endophalus (inner sac). Moreover, it looks like simplicity of the external morphology of the aedeagus and complexity of the endophallus are correlated (Roig-Alsina, 1993;Jaloszynski et al., 2015;Petitpierre & Anichtchenko, 2018;Zhou et al., 2019). Presumably, the endophallus in these and possible other cases is the most, or only (e.g., Danilevski & Kasatkin, 2006;Hayashi & Yoshitomi, 2015) reliable morphological character for delimitation of closely related species. For example, the melyrid beetle species of the genus Laius occurring in sympatry exhibit greater differences in body size and sclerites of the endophallus than allopatric species (Yoshitomi, 2014), suggesting that these morphological features maintain species boundaries. In traditionally better-studied taxa like Carabidae, with generally simple external structures of the aedeagus, routine use of the endophallic characters in species diagnoses, even for amber fossils (Schmidt et al., 2016;Schmidt & Michalik, 2017) is becoming a common practice. However, since the various endophallic structures may evolve, also independently, in response to complex interactions between a male and a female or between males through intraspecific competition (e.g., Schmitt & Uhl, 2015;Jaloszynski et al., 2015;Matsumura et al., 2017, Dougherty et al., 2017Kyogoku & Sota, 2015) they may vary in their value as species-specific markers. Interestingly, Kataev (1996) reported very notable parallel intraspecific aberrations in the endophallic sclerites of some Carabidae.
Complexity of insect genitalia and their use for species delimitation aligns well with the classical evolutionary view of genitalic differences among species acting as isolating mechanisms i.e., the lock and key hypothesis.
A closer look at the subject (Eberhard, 1985), however, revealed more diverse factors underlying the evolution of genitalia and led to the reconsideration or even complete rejection of the lock and key hypothesis (Eberhard, 1985;Shapiro & Porter, 1989). Currently, sexually antagonistic coevolution and especially cryptic female choice (Eberhard, 2009) are considered major drivers of morphological evolution in genitalia, though these processes are not so directly associated with speciation. These advancements in evolutionary biology suggest that some structures of the genitalia, just as any other traits, may be more or less variable within and between species and thus perform better or worse as a marker for species delimitation. In this respect, Oxyporus is again noteworthy. Its obligate fungivory apparently determines complex mating biology because beetles of both sexes, of the same and different species, are often aggregated within the short-living fruiting bodies of mushrooms. For example, Tokareva et al. (2020) observed large gatherings of Oxyporus maxillosus (up to 40 specimens) in the fruiting bodies of Laetiporus sulphureus fungus, which they probably used for feeding and mating. After mating, females must move to other fungi to lay eggs as the fungal hosts are so short-lived. We also often observed that several specimens of Oxyporus of more than a single species can be found in one, even relatively small fungus. It was as early as in Heeger's (1853) paper that an observation of Oxyporus males subsequently mating with several females has been published. These observations suggest complex intra-and possibly interspecific sexual interactions in this genus, which may be functionally related to morphology of the male genitalia. As noted before (Makranczy, 2012) and can be seen from the survey of hitherto described species, Oxyporus (especially s. str.) have a very simple median lobe but complex, eversible endophallus. tokareva et al.
All these aspects motivated our paper, which aims to find reliable characters for robust species delimitation in Oxyporus in the course of a thorough taxonomic revision of this genus within the limits of the Russian fauna. For species delimitation we explored three character systems: 1) external morphology, including traditionally used characters of body coloration, 2) morphology of the aedeagus and 3) dna-barcoding. Our approach is inspired by the growing amount of evidence (Pante et al., 2015) that seeking congruence from multiple and alternative lines of evidence is the most robust way to capture the border between closely related species. For the aedeagi, we consistently explore their overall shape (shape of the phallobase, median lobe and parameres) and, as a separate character set, the shape and armature of their endophallus (also called inner sac).

Specimen collecting and deposition
This paper is based on specimens from the collections of the Zoological Institute, Russian Academy of Sciences, St. Petersburg (zisp); Federal Scientific Center of the East Asia Terrestrial Biodiversity, Russian Academy of Sciences, Vladivostok (feb ras); Canadian National Collection of Insects, Arachnids and Nematodes, Ottawa (cnc); Museum of Natural History, Prague (mnp); Natural History Museum, London (nhm); Natural History Museum of Denmark, Copenhagen (nhmd), and personal collections of Michael Schülke (cSch) and Kirill V. Makarov (cMk).
Some of the specimens were fresh material recently collected by at and as in the Czech Republic and in various regions of Russia. These specimens were collected either by the standard method of sifting leaf-litter with mushrooms, or by a specialized "plastic-bag" method. Each fungal fruiting body with round holes on the hymenophore, typical for Oxyporus presence, was picked up in a fast manner by hand covered with two plastic bags, one over another. Then, if Oxyporus specimen(s) were present, the pair of bags was everted around the fungus and tied on the top, with the air left inside to let the beetle(s) breathe and prevent early deterioration and destruction of the fungus. A field label for the sample collected was placed between the two plastic bags walls. The samples were processed not later than in three or four hours after being collected.

Specimen recording, georeferencing, and mapping
For the type specimens, labels are provided verbatim, separated by backslash; Cyrrilic labels are transliterated. For non-type specimens labels are interpreted and translated to English; only in ambiguous cases they are given verbatim. The majority of the examined specimens did not have geographic coordinates on their labels. Using all possible resources we found coordinates for most of the localities and added them in square brackets in the 'Material examined' data. We used these coordinates for our distribution maps made with SimpleMappr (Shorthouse, 2010).

Lectotype designations
In the interests of the stability of the zoological nomenclature, we have designated lectotypes for species with multiple syntypes that included at least one male. Our respective lectotype and paralectotype labels are added to the specimens.

Specimen sampling for dna-barcoding
To test the morphospecies hypotheses by dna-barcoding, CO1 barcodes were assembled for 97 specimens of Oxyporus. Of them, 45 specimens were 96% ethanol-preserved Oxyporus collected by at and as in the Czech Republic and in Russia (table 1). Barcodes for other 30 specimens were taken from the recently published paper on Korean Oxyporus (Lee et al., 2020). A single CO1 sequence of O. (O.) maxillosus used by Lee et al. (2020) in their analysis, when checked by us via the BLASTn algorithm, turned out to have 100% similarity with the barcode of Microcara testacea Linnaeus, 1767 (Scirtidae) and thus is not included in our research. Six additional specimens from the cnc were sequenced separately (see below) and uploaded to bold Systems (table 1). Finally, barcodes for yet another 17 specimens were taken from online databases, namely bold Systems (https:// v3.boldsystems.org/index.php/TaxBrowser_ Home) and GenBank ncbi (https://www. ncbi.nlm.nih.gov/nucleotide/). The sampling largely covers the morphological diversity of the genus across the entire territory of Russia and adjacent areas. Barcodes for outgroup taxa Lordithon fungicola Campbell, 1982 (Tachyporinae), Philonthus tenuicornis Mulsant & Rey, 1853, and Ontholestes murinus Linnaeus, 1758 (Staphylininae) were taken from bold Systems.

dna extraction, pcr amplification, sequencing, and alignment
Total dna was extracted from beetle legs using the Qiagen DNeasy Blood & Tissue Kit with the standard protocol (https://www.qiagen.com/us/resources). Samples were incubated at 56°C in 180 μl of atl buffer with 20 μl of proteinase K added afterwards for about 24 h. pcr was performed using Evrogen kit for Master Mix (https://evrogen.ru/products/PCR-kits/PCR-kits-polymerases/): 0,1 μl Taq polymerase in a 25 μl reaction mixture containing 1 μl of each primer, 2 μl dNTPs, 2.5 μl of Taq Buffer and 2 μl of genomic dna template. The primer pair LCO1490 (5'-GGTCAACAAATCATAAAGATATTGG-3') and HCO2198 (5'-TAAACTTCAGGGTGA-CAAAAAATCA-3') was used to amplify a 658 bp fragment of the CO1 gene (Folmer et al., 1994). The thermal cycling program consisted of 35 cycles of 94°C for 30 sec, 42°C for 30 sec and 72°C for 45 sec, followed by a final extension at 72°C for 10 min. Paired forward and reverse reads were assembled and edited in Geneious (v. 9.1). Amplified fragments were sent for sequencing to the "Molecular and Cell Technologies" resource center (St. Petersburg State University) or to Evrogen Inc. (Moscow). Additional specimens were sent to the Biodiversity Institute of Ontario (bio) (Guelph, Ontario, Canada) for extraction, amplification and sequencing.
Assembled sequences were checked via blast (https://blast.ncbi.nlm.nih.gov/Blast. cgi) to verify their identity and were aligned with CO1 fragments of Oxyporus (Oxyporus) rufus and O. (O.) maxillosus from ncbi. Sequences were aligned in Geneious 9.1 using mafft v. 7.450 (Katoh and Standley, 2013) under the most rigorous L-INS-i strategy (see alignment in supplementary material S3). The sequences were uploaded to Genbank (https://www.ncbi.nlm.nih.gov/genbank/). The accession numbers for all species used in our analysis are given in table 1.

Microscopy and illustrations
All photos of beetles were taken using a Nikon smz 1500 stereomicroscope equipped with a Nikon D700 digital slr camera. Photos of genitalia were taken either using a Leica M205C stereomicroscope equipped with a Canon ems 6D camera (images stacked with Zerene Stacker 1.04 (Zerene Systems, Richland, WA, USA) or using a Leica M205C stereomicroscope equipped with a Leica DFC495 camera (images stacked with Helicon Focus 6.2.2). Drawings of genitalia structures were made with a Leica dm 2500 microscope equipped with a camera lucida. All illustrations were tokareva et al. Downloaded

Endophallus preparation and terminology
For the preparation of the everted endophallus in relatively small beetles such as Oxyporus, we developed an osmotic technique based on the suggestions of E.A. Khachikov (Department of Zoology, Southern Federal University, Rostov-on-Don).
To evert a beetle endophallus with our technique, one needs high concentration acetic acid (above 90%), 10% koh solution, electric hot plate, small ceramic jars as ones used for chemicals, and warm (40-60°С) clean water. The step-by-step procedure is as follows: 1.
Place the ethanol-preserved or dry (removed from an insect pin) beetle specimen into a Petri dish with warm water. When the intersegmental membranes become soft enough, detach the tip of the abdomen (terminalia) starting with segment vii or viii. 2. Place a hot plate set to medium heat under a ventilation hood. 3. Under ventilation, drop 5-15 mL of acetic acid in an empty ceramic jar. The depth of the liquid in the jar should be approximately 8 mm. Place the terminalia inside the jar. 4. Heat the jar on the hot plate but avoid boiling the acetic acid by moving the jar off the plate for five to eight seconds whenever boiling occurs, or adjust the heat. If acid becomes nearly evaporated, remove the jar from the hot plate and pour more acid along the jar wall, distant from the sample. On average, the heating step takes from three to eight minutes. Its duration largely depends on the condition and size of the specimen.
To check if the terminalia are ready for the next step, place them in a drop of acetic acid on a slide and try to carefully compress the basal bulb of the aedeagus under a stereomicroscope. Sometimes the endophallus everts at this stage, usually in fresh, ethanol-preserved material. The specimen is ready for the next step if small (often sparkling) air bubbles are observed inside the aedeagus. 5. Remove surrounding tissue from the aedeagus and return to the jar with acetic acid for an additional 15-20 seconds. 6. Transfer the aedeagus to the jar with 10% koh solution prepared in advance and place on medium heat, under ventilation. Avoid boiling as above. On average, this step takes 6-20 minutes. Its duration depends on the degree of maceration of tissues inside the aedeagus. Sometimes, the endophallus spontaneously everts at this step. 7. Transfer the aedeagus to a Petri dish with warm distilled water. If the inner sac has not everted spontaneously in two minutes, gently press the basal bulb with blunt forceps and then with another pair of blunt forceps, keep pushing the dorsal membrane of the median lobe. The endophallus should evert easily if the osmotic pressure is high enough and the maceration in koh solution was long enough. The aedeagus with the everted endophallus should be kept in warm water to dilute all the active chemical agents. Unfortunately, the fully everted endophallus does not stay like that for a long time, and one should take a photo as soon as possible, preferably the same day as it was everted. In most cases the membranous sac shrivels down in glycerin, but can be straightened out again by washing in water and heating in koh solution for a few minutes. We did not find a better solution than to store aedeagi with the everted endophallus in a microvial with glycerin, pinned under the respective specimen.
Our terminology of the endophallic structures is purely descriptive, without any assumptions about possible homology to similar structures observed in other Staphylinidae or beetles in general. It is based on relative positions of the structures and presented in the scheme of the Oxyporus s. str. endophallus in fig. 2A-C. Terminology of the main parts of the aedeagus follows Blackwelder (1936).

Morphological character matrix
To test the robustness of the dna barcode-based trees, we conducted a total-evidence analysis with a morphological partition (for details see 2.7) and explored the reliability of particular morphological traits as species-specific characters (for details see 2.8). The set of specimens coded in the morphological matrix is the same as for the dna extraction, except for sequences provided by Lee et al. (2020) and those downloaded from ncbi. Moreover, there are no specimens included in the matrix without having a barcode. Thus, polymorphic traits which were demonstrated by some dna-barcoded specimens are included into the matrix as well. The set of characters describing male genitalia is extrapolated to female specimens assuming that if being males, they would had the same characters as males of the same species. All the specimens included into the species tree reconstructions are listed in the table 1 with the museum individual numbers and ncbi accession numbers provided. The morphological matrix includes characters of external morphology (1-9), of which 1-6 were suggested by Nakane and Sawada (1956) for subgenera delimitation; non-endophallic aedeagal characters (10-22); endophallic characters (15-22); and characters of body coloration (23-30), which were traditionally used in species delimitation in Oxyporus. The entire list of character is as follows: 1.
figure 1 CO1-based molecular phylogeny of Oxyporus from Russia and adjacent lands. bi 50% consensus tree on the left and ml tree on the right. Node numbers correspond to species as shown in the trees and used in the text. Each node is documented with posterior probability (bi tree) or bootstrap support (ml tree).
12. Median lobe, shape: (0)    Ancestral states reconstruction under Mk1 model for three characters of general external morphology: labrum structure (1), mesepisternum-epimeron surface (7), and pleural suture of mesopleuron (8). Pie charts represent the likelihood indices for each node. Character numbers as in section 2.6 of the paper.  (0) 3G). The morphological matrix is available as supplementary material S1.

bi and ml analyses, genetic distances
Taxonomic hypotheses of species were tested with three types of analysis: Bayesian Inference (bi) and Maximum Likelihood (ml) analyses of the molecular barcodes only and then the total evidence bi analysis of the barcode data combined with the morphological partition (see section 2.7). For all types of analysis the same set of specimens was used (table 1), and for the combined bi analysis a morphological partition for 30 characters (supplementary material S1) was added to the same set of specimens. bi analyses (Ronquist et al., 2011) were performed in MrBayes on xsede (3.2.7a), cipres Science Gateway (Miller et al., 2010). Four runs with four chains each were running simultaneously for 10 million generations with 0.1 temperature setting; burn-in was set at 25%. We applied the most complex General-Time-Reversible (gtr) with gamma-distributed substitution rates and invariable sites (gtr+I+G) model to the CO1 partition, as recommended by Arenas (2015) and Abadi et al. (2019), and the only available Mk model with a gamma distribution (Lewis, 2001) for morphological partition. Partitions of CO1 (574 bp) and morphological characters (30) were concatenated in Sequence Matrix 1.7.8 (Vaidya et al., 2011). Chains were sampled every 1000 generations and the respective trees were written to a tree file. After the analyses, the stdout files were checked to ensure that the average standard deviation of split frequencies was below 0.01 in each case. Fifty-percent majority-rule consensus trees and posterior probabilities of clades were calculated using the trees sampled. The ml analysis (Stamatakis, 2016) was launched in RAxML-hpc BlackBox (8.2.12), cipres Science Gateway with an auto-generated number of bootstrap iterations and a subsequent thorough ml search, using gtr+I+G model as well. Nodes with bootstrap supports lower than 85 were collapsed. The resulting CO1-based bi and ml trees with posterior probabilities (left phylogram) and bootstrap supports (right phylogram), respectively, are provided in fig. 1. To illustrate apomorphies that characterize each species clade, we performed acttran character optimization in Winclada (Nixon, 2002), based on the total-evidence bi tree as an input. The matrix of genetic distances was calculated using the Kimura 2-parameter (K2P) model of base substitution with mega X (Kumar et al., 2018) (supplementary material S2).

Ancestral character state reconstruction analysis and parsimony-based character evaluation
Ancestral character state reconstruction was performed by ml with the Mkv model of morphological evolution (Lewis, 2001) in BayesTraits v2, MultiState (Pagel et al., 2004) using the total evidence bi tree as input. The "Generate BayesTraits Input" function in TreeGraph2 (Stöver et al., 2010) was used to speed up the calculations of character states likelihoods for each node of the respective Bayes tree, and reconstructed states were obtained for each character (figs 6-8; supplementary material S4-S24). Additionally, the parsimonious characteristics (ci, ri, rc) for each character were calculated on the tree topology from the bi total evidence analysis with paup 4.0 (Swofford, 2003). They are presented in table 2.

Morphology of the endophallus of Oxyporus
Based on the pool of specimens available from our study region, we observed significant diversity of endophallic structures among species of Oxyporus and a strong difference between its two subgenera. While potentially homologous structures could be identified in the membranous endophallic lobes between the subgenera, the apical sclerites were remarkably difficult to homologize. This was especially true for Pseudoxyporus where the diverse endophallic structures were difficult to homologize even within the subgenus. Within the subgenus Oxyporus, on the contrary, they are more homogenous and thus more comparable. The general scheme of the Oxyporus s. str. endophallus is given in fig. 2 to aid with standardized comparisons between species. In the absence of any comparative study of the endophallic structures across Coleoptera or Staphylinidae, we did not attempt to tie our scheme here to any of a few works of this kind devoted to particular groups of Staphylinidae (Khachikov, 2006(Khachikov, , 2015Jałoszyński et al., 2015;Zhou et al., 2019). They all demonstrate very different types of endophallic structures which would require a comprehensive comparative morphological study across the family to establish homologies. Without such a study, we use as simple and as descriptive terms as possible, naming endophallic membranous lobes and sclerites according to their collocation ( fig. 2). We also figure 8 Ancestral states reconstruction under Mk1 model for three characters of coloration: black-and-yellow color pattern on elytra (26), abdominal tergites and sternites iii-v coloration (28), and abdominal paratergites coloration (30). Pie charts represent the likelihood indices for each node. Character numbers as in section 2.6 of the paper.  fig. 1, respectively. The total-evidence bi tree is given in fig. 5 with morphological characters plotted under fast optimization (acctran). Major nodes representing species in our revision are numbered from 1 to 10. Within the subgenus Pseudoxyporus, both bi and ml recover three clades corresponding to species as they are currently recognized: O. (P.) melanocephalus, O. (P.) dybowskii, and O. (P.) occipitalis. The K2P genetic distance between sequences of different species from the same subgenus varies from 0.12 to 0.17, whilst between species from different subgenera it varies from 0.16 to 0.22 (supplementary material S1).
Together with O. (P.) cyanipennis, all four Pseudoxyporus species are well recognized morphologically and none of them called for further testing of species limits. Their sequences and morphological characters were included in the analysis, mainly to test subgenus-specific characters. Both bi and ml trees here reinforce their current concepts presented in detail in the Taxonomy section.
Within the subgenus Oxyporus s. str. our analysis reveal seven clades, which we interpret as species as following: Node 4 unites O. (O.) mannerheimii from the European part of Russia, Siberia and the Russian Far East (bi: 1.00; ml: 100; zero genetic distance). This species is defined by two homoplasious characters: shagreened mesepisternum-epimeron surface (7-1) and evenly colored elytra (25-0).
Node 7  ) parvus, a recently described species from South Korea (bi: 1.00; ml: 99; genetic distance variation: 0.00 to 0.03). The clade is supported by two homoplasies: elytra yellow with black spots on lateral sides, elytral suture with wide black line (26-4) and black abdomen with yellow paratergites iii-v (30-1). Despite previously defined, significant distinctions in coloration between these 'species' , all specimens share the same set of aedeagal characters. According to the structure of the endophallic apical sclerites of the type specimen of O. (O.) basicornis, clade 7 represents a single species, which corresponds to the revised concept of O. basicornis.

Morphological characters: ancestral state reconstruction and diagnostic value for species delimitation
To test the diagnostic value of morphological characters at the species level, an Ancestral States Reconstruction (asr) with Maximum Likelihood, Mk1 model, was performed for all characters. Reconstructed ancestral states are shown in figs 6, 7, and 8 for several characters that were the most illustrative for each of three categories: general external morphology, genital morphology and body coloration. Reconstructed ancestral states for all other characters were added as supplementary material S4-S24. Parsimony-based indices (ci, ri, rc) for all 30 characters optimized on the total-evidence tree are shown in table 2. According to the ci, ri, and rc indices, as well as the likelihood distribution for character states ancestral to each species, all explored characters can be arranged into four groups with respect to their performance as species-specific (  fig. 7) were stable and allow for unequivocal species recognition within the subgenus Oxyporus once an endophallus is everted. 2. Characters 5, 14, 17, 18, 21 (supplementary material S7, S14, S17, S19, respectively) concerned the general or aedeagal morphology and also had relatively high parsimony indices (ri: 0.95-0.98; ci: 0.33-0.75). These characters were species specific and did not display interspecific polymorphism, which was supported by high asr likelihoods for most probable states of each species (not lower than 0.99), and thus can be reliably used in identification. 3. Characters 23-25 (ri: 0.83-0.90; ci: 0.50-0.66) describe coloration of head, pronotum and ground color of elytra, respectively. They did not expose polymorphism for species and asr likelihoods for most probable states of almost all species at nodes were relatively high ( (Nakane and Sawada, 1956). The same applies to characters 27-28 (coloration of abdominal segments iii-iv, supplementary material S23 and v, fig. 8, respectively). The character describing abdominal pleurite coloration (30) (Shabalin, 2012), our analysis showed low utility of this character set for species delimitation. Although asr likelihoods for the most probable states for each species (not lower than 0.97) and parsimony indices (ri: 0.92; ci: 0.67) for the character 29 (abdomen coloration, tergite and sternite vi) were relatively high, due to the presence of only two specimens forming a single cluster with alternative coloration on the tree (supplementary material S24), we do not recommend to use this character for species delimitation. In the much wider dry material sample we observed specimens with a brightly colored sixth abdominal segment but possessing a maxillosus-type endophallus ( fig. 13D). Intermediately colored specimens showed no geographically structured pattern and were found in the West and East Palearctic (see Discussion for details).

Discussion
Our data showed that all CO1-based clades that could be considered as separate species because of molecular distance are supported by morphological characters that indicate a hiatus between them. This supports the validity of a single locus approach for dna-based species delimitation, at least for the barcoding CO1 fragment for Palaearctic Oxyporus. In the course of checking traditional and novel morphological characters against the molecular results, we revealed seven new morphological traits (7, 14, 17-22) that can be reliably used for species delimitation in Oxyporus. Most of them are endophallic characters that are the most reliable for species diagnostics in Oxyporus. For some species, e.g., O. germanus and O. basicornis, the endophallic characters are the only diagnostic traits that we could observe, especially if the degree of impression on pronotum is variable. This confirmed the earlier warnings of Makranczy (2012) against using only body coloration patterns, which were unfortunately broadly used to describe species of Oxyporus. Among these body coloration characters, partial utility for species diagnostics was shown for three characters (23, 24, 25), namely the coloration of the head, the pronotum and the ground color of the elytra.
Data integration helped to resolve taxonomic problems, such as discrimination of the morphologically nearly identical species O. (O.) basicornis and O. (O.) germanus. In their case, nothing but the CO1 barcode and the endophallic structures can tell that these otherwise identical-looking species are not even sister to each other. In this respect, the congruence of CO1 and endophallus for the delimitation of Oxyporus species is remarkable. It is also clear how misleading external characters hitherto used for species delimitation were. In particular, O. via University of Copenhagen species body length can vary by up to 5 mm and the head width can be greater or less than the pronotal width. Moreover, we did not find any geographically structured pattern for these polymorphic traits, except coloration of the abdomen of western and eastern specimens of O. (O.) maxillosus. In that species, specimens with an entirely brightly colored (pale) abdomen seem to be more frequent in Europe east to the Urals, while those with an entirely dark abdomen mostly occur in the Russian Far East and adjacent territories. Given such variability of external characters, the rather uniform shape of aedeagus is poorly informative for species diagnostics in clusters of closely related species. Especially in Oxyporus s. str., contrary to many other staphylinids and beetles in general, the shape of the aedeagus is not diagnostic. In the case of O. (O.) basicornis and O. (O.) germanus, it is in fact identical. That makes the difficult procedure of everting the endophallus a desired or even a mandatory step for separating some species of Oxyporus from each other, unless CO1 barcoding is performed. Our study demonstrates that in these cases, CO1 barcoding is a rather straightforward shortcut to identification when dna-grade or recently collected dry pinned material is available.
The remarkable congruence of the species delimited by CO1 barcodes and endophallic structures, especially for otherwise extremely similar Oxyporus species stresses the need to explore the biological role of the endophallic structures. There must be something that promotes diversification of these structures among species. As far as known for some species of Oxyporus (Tokareva et al., 2020), multiple males and females of the same and several species can encounter each other in the same large and short-living fungal caps during the mating season. Such aggregations provoke complex intra-and interspecific sexual interactions. Given the above reviewed complexity of the evolutionary phenomena that drive morphological diversification of insect genitalia, Oxyporinae might be a suitable model group for the investigation of sexual competition, female choice and the "lock and key" hypothesis, especially with respect to the endophallus. It is noteworthy that the female genitalia in Oxyporus remain unexplored. With the approximately 120 species of Oxyporus known globally, much taxonomic work remains to be done. Unfortunately, even recent descriptions of new species rarely include photos of the aedeagus and almost never include at least some figures of the endophallus, not to mention one that is properly, fully everted. Until very recently, very few Oxyporus species have been barcoded. The only publication that both extensively used barcoding and provided illustrations of the endophallus for a new species is Lee et al. (2020 ). It is a very progressive and useful study of the Korean fauna which, unfortunately, was performed with coloration-based taxonomy of Oxyporus in Japan because the authors did not check all their Korean species on their possible conspecifity with Japanese congeners (for details see below). To progress with the taxonomic study of Oxyporus, we suggest endophallus eversion and dna barcoding as a routine techniques of species delimitation in this genus. We here demonstrate the utility of such an approach in the following revision of the Russian fauna of this genus. Another important taxonomic problem concerns a significant difference observed between the subgenus Pseudoxyporus and the nominative subgenus Oxyporus, which were originally considered as genera before Campbell (1969) down-ranked them to subgeneric status. Makranczy (2012) argued that this down-ranking was not justified due to the numerous morphological and bionomical traits indicating a notable distance between Oxyporus and Pseudoxyporus. Makranczy's conclusion that these taxa represent separate genera is supported by the large genetic distance between Oxyporus and Pseudoxyporus (0.16-0.22) and the significant difference between their endophallus types found in our work. However, a problem is that the genus Pseudoxyporus was originally erected on the too small, geographically and taxonomically, subsample of Oxyporinae (Nakane & Sawada, 1956). To make the final decision on this taxonomic issue, the world fauna, and especially species which share some characters of both subgenera, such as Oxyporus smithi Bernhauer, 1910, must be assessed.
Diagnosis. Antennae filiform or clavate, inserted at side of head near anterior margin of eyes; mandibles prominent, sickle-shaped; apical labial palpomere large, transverse, strongly securiform; procoxae large, conical, prominent, protrochantin broadly exposed; middle coxae widely separated; posterior coxae transverse; tarsi 5-5-5; abdomen with six visible sterna and two pairs of paratergites per segment. In Russia, the genus includes 10 species which cannot be confused with any other genus of rove beetles.
Note. Soon after Nakane and Sawada (1956) described the genus Pseudoxyporus for a number of Japanese species formerly in Oxyporus, Campbell (1969) (Nakane & Sawada, 1956).  Lee et al. (2020;as O. parvus), paratergites of first three or four visible abdominal segments yellow while respective tergites and sternites black. Endophallus: distal lobe bent dorsad, consisting of dorsal, ventral and additional pair of thin bow-like spicules and central sclerite of characteristic shape ( fig. 9A, B), all covered with membrane; dorsal and ventral pairs of thin spicules bent ventrad and then sharply dorsad to form notable sclerotized frame supporting walls of secondary gonopore.

Oxyporus (Oxyporus) basicornis Cameron
Redescription. Body length 5.7-10.1 mm (n = 10). Head subovate to round, smooth, brown to black. Pronotum without transverse elevation, smooth, brown to black. Elytra black with yellow pattern. Mesepisterna and epimera with rough punctation, black. Abdominal tergites and sternites black, lateral abdominal sclerites of segments i-iii or iv yellow. Aedeagus: median lobe with round apex; endophallus with twin-coned spherical dorsal lobe covered with medium-sized denticles, pair of spherical lateral lobes, and transverse ventral lobe covered with tiny granulation; apical distal lobe membranous, with tiny granulation, bent dorsad, with four pairs of thin long bow-like sclerites under membrane and one thin ligament sclerite dorsally. Basal sclerotization disjunct, additional spiculae of a peculiar form (figs 3E, F, 9A, B).
Comparison. Oxyporus basicornis is identical to O. germanus in coloration and body proportions. Both species can be distinguished only by CO1 barcode and endophallus structure. In particular, the endophallus of O. basicornis has two additional spiculae and the additional central sclerite missing in O. germanus. From all other species of Oxyporus in Russia, O. basicornis can be distinguished by the structure of endophallic sclerites and coloration of head, thorax, and elytra.
Distribution. Oxyporus basicornis is reliably known from Japan (Honshu), South Korea (Lee & Ahn, 2020) and Russia, where it is recorded here for the first time. Russian records come from Primorsky Territory and Buryatia; it is presumably widely distributed in the Russian Far East and East Siberia ( fig. 10C).
Bionomics. This species can be usually found in large fungal caps together with O. germanus and/or O. rufus, or alone in Suillus americanus (Peck) Snell in the Russian Far East.
Comments. Oxyporus basicornis was described from Japan as a variety of O. lewisi (Cameron, 1930) based on the dark median line on the last six antennomeres and yellow paratergites on the first three (and not four as in O. lewisi) abdominal segments. Nakane and Sawada (1956) raised it to the status of a species. In Russia, O. basicornis was recorded for the first time in Primorsky Territory, diagnosed only using color (Shabalin, 2012). Our study, however, has revealed that O. basicornis overlaps in coloration with O. germanus, another species described from Japan, and distributed in the Russian Far east as well (see below). Both species reliably differ from each other only by CO1 barcode and structure of the endophallus. Therefore, it was unclear which species was recorded by Shabalin (2012) under the name O. basicornis, while our data here, based on a reexamination of Shabalin's material as well, are the first reliable records of this species from Russia.
Oxyporus aequicollis was described from a single specimen sampled in 'Irkutsk' (Bernhauer, 1935) as a species close to O. mannerheimii. In fact, the original description of O. aequicollis is nearly identical to the original description of O. germanus, earlier described from Japan. Kirschenblatt (1938) (Lee et al., 2020). The authors showed the difference of O. parvus from O. germanus based on CO1 and provided a scheme of the endophallus of their new species, also mentioning and partly illustrating diagnostic characters of O. parvus, namely the presence of the tomentose patches on tergites iii and iv in O. parvus (on tergite iv in O. germanus), brown abdominal tergites iii-v in O. parvus (iii-vi in O. germanus) and the presence of a shallow and broad depression on anterior third of pronotum in O. parvus (presumably absent in O. germanus studied by the authors). Regarding the tomentose patches, our sample of material shows variability of the size, number and location of these patches even within a species, at least in O. rufus, O. germanus, O. basicornis, O. maxillosus, O. procerus, O. mannerheimii, and O. niger. The fine scratch-like impression in the anterior third of the pronotum of O. germanus appears to be a varying trait as well. Unfortunately, Lee et al. (2020) did not compare O. parvus with O. basicornis. Inclusion of their molecular data in our analysis, as well as matching the molecular clades with the respective (not sequenced) type material of O. basicornis via examination of the endophallus, clearly shows that O. parvus is conspecific with O. basicornis and thus the former is placed in synonymy with the latter. near Medvezhye lake, [53.6139N 142.9568E], 28.viii.1979, in fungi, Okulov leg. (feb ras).

Oxyporus (Oxyporus) germanus Sharp, 1889
Diagnosis. Pronotum smooth, paratergites of first three or four visible abdominal segments yellow while respective tergites and sternites black. Endophallus: distal lobe of apical sclerites consisting of ventral pair of spiculae and dorsal pair of complex sclerites flattening and widening in the middle of distal lobe and forming characteristic structure supporting walls of distal lobe ( fig. 9C, D).
Redescription. Body length 5.2-9.8 mm (n = 10). Head subovate to round, smooth, brown to black. Pronotum without transverse elevation, smooth. Elytra black with yellow pattern. Mesoepisterna and epimera with rough punctation, black. Abdominal tergites and sternites black, paratergites of segments iii-v or iii-vi yellow. Aedeagus: median lobe with round apex; endophallus with twinconed spherical dorsal lobe covered with medium-sized denticles and transverse, relatively large ventral lobe covered with tiny granulation, pair of spherical lateral lobes, and apical distal lobe membranous, with tiny granulation, bent dorsad; one pair of thin long bow-like ventral spiculae and paired complex additional spiculae in center flattening and widening towards apex under membrane and small obscure ligament sclerite dorsally; basal sclerotization obscure ( fig. 3G, 9C, D).
Comparison. Oxyporus germanus is identical to O. basicornis in coloration and body proportions. Both species can be distinguished only by CO1 barcode and endophallus morphology. In particular, the endophallus of O. germanus has paired additional central spiculae flattening and widening towards apex, which are absent in O. basicornis. From all other species of Oxyporus in Russia O. basicornis can be distinguished by the structure of endophallic sclerites and coloration of head, thorax, and elytra.
Bionomics. This species can be usually found in large fungal caps together with O. basicornis and/or O. rufus, or alone in Suillus americanus (Peck) Snell in the Russian Far East.
Redescription. Body length 8.5-10.7 mm (n = 10). Head subovate to round, smooth, brown to black. Pronotum without transverse elevation, smooth. Elytra black. Mesoepisterna and epimera shagreened, black. Abdomen black. Aedeagus: median lobe with round apex; endophallus with twin-coned spherical dorsal lobe covered with medium-sized denticles and transverse ventral lobe covered with tiny granulation, pair of spherical lateral lobes, and apical distal lobe membranous, with tiny granulation, bent dorsad; it includes one pair of relatively short dorsal and ventral spiculae and paired additional spiculae in center disposed in apical portion of lobe, and ligament sclerite closely connected with dorsal spiculae; basal sclerotization connected only to ligament sclerite (figs 3C, 11A, B).
Comparison. Among all other species of Oxyporus in Russia, O. mannerheimii is most similar in coloration to O. niger, from which it differs by the smooth pronotum and shagreened, not punctured surface of mesepisterna and mesepimera, and by the structure of the endophallus (compare fig.  11A, B and 11C, D).
Distribution. Oxyporus mannerheimii is a Transpalaearctic species and its distribution is confirmed here by matching dna barcodes of the European and Far Eastern specimens. In Russia, it occurs from Leningrad Region, through Urals and Irkutsk Region to Primorsky and Sakhalin Regions ( fig. 12A).
Bionomics. In Leningrad Province, Russia, it was collected from fungal caps of Pleurotus spp.
Diagnosis. Pronotum black, with transverse elevation followed by impression; mesepimeron and mesepisternum without rough punctation; elytra black and yellow; abdomen at base yellow-brownish and black apex, or entirely black. Endophallus: apical distal lobe modified into sclerotized hook, spiculae forming this hook adjunct to common basal sclerotization ( fig. 13A-E).
Among all species of Oxyporus in Russia O. maxillosus is most similar to O. niger from which it clearly differs by the presence of yellow spots on elytra and by shorter and wider bow-like, joined spiculae forming the apical hook of the endophallus.
Redescription. Body length 6.9-13.8 mm (n = 10). Head subovate to round, smooth, brown to black. Pronotum with transversal elevation followed by impression. Elytra black and yellow. Mesoepisterna and epimera with rough punctation, black. Abdomen at base yellow-brownish and black apex, or entirely black. Aedeagus: median lobe with round apex; endophallus with twin-coned spherical dorsal lobe covered with medium-sized denticles and transverse ventral lobe covered with tiny granulation, pair of spherical lateral lobes, and apical distal lobe modified into relatively wide sclerotized hook, bent dorsad; dorsal side of hook formed with relatively big ligament sclerite; spiculae pairs which form hook joined to basal sclerotization (figs 3A, 13A-E).
Bionomics. Data on larval stages, life cycle duration and known fungal hosts, as well as a relatively complete account of all earlier published data concerning bionomics of this species, can be found in Tokareva et al. (2020).
Comments. Oxyporus maxillosus is a morphologically variable and widespread species with a very large pool of published data of varying quality and reliability. Our data here are sufficient to clarify its identity across the Russian territory, but it still requires a revision as far as the fauna of Japan is concerned. As in Europe, from where it was described (Germany) (Fabricius, 1793), Oxyporus maxillosus occurs in Siberia and Far East of Russia as two forms: a nominative form with bicolored abdomen and a form with the entirely black abdomen described by Gebler (1829) as O. angularis from Altai Region (Barnaul) and described by Ganglbauer (1895) as O. maxillosus var. amurensis from 'Amur' . Based on the examination of extensive material here, some of which was dna barcoded and dissected to examine the aedeagus including endophallus, both forms are conspecific and occur sympatrically. Therefore we corroborate their current widely accepted status as synonyms of O. maxillosus. We also note that the form with the entirely black abdomen is much more frequent in the Eastern Palaearctic than the form with the bicolored abdomen. Another conclusion is that there is no sharp border between these two color forms, as sometimes a greater or lesser number of abdominal segments can be dark, or some segments can be transitionally bicolored. Analysis of the CO1 barcodes of a sample of these variously colored specimens from a broad geographic area including the Czech Republic, Russia (from Urals, Buryat and Altai Republic, Khanty-Mansy Autonomous Area, and Primorsky Territory), Japan, and South Korea revealed 0.03 as a maximum genetic distance among them. Molecular and morphological examination of the material from Japan matching descriptions of O. basiventris, O. aokii, and O. japonicus, all species described from Japan (Jarrige, 1948;Dvořák, 1956;Sharp, 1889, respectively), confirmed our suspicion that they are conspecific with the morphologically variable O. maxillosus, and thus all should be synonymized with the latter when the respective type material is examined. For the purposes of this revision, we conclude that the form of O. maxillosus with bicolored abdomen was wrongly reported from the Russian Far East as O. aokii by Shavrin and Gildenkov (2009) and O. basiventris by Shabalin (2012). Although less relevant for our revision, we should note that O. (O.) kobayashii Hayashi, 2015 from Japan is highly likely a synonym of O. maxillosus as well, judging from the illustrations and characters given in the original description (Hayashi, 2015). Again, to check this hypothesis, a proper investigation including the type material should be done. Sharp, 1889 (figs 3C, 11A, B) Oxyporus niger Sharp, 1889: 407;Kirschenblatt, 1938: 531;Shavrin and Gildenkov, 2009: 123;Shabalin, 2012: 414.
Diagnosis. Body black, except pale brown tarsi, apical parts of labial palps and lateral parts of antennae. Pronotum with transverse ridge followed by impression; mesepimera and mesepisterna with scattered small punctation. Aedeagus: median lobe apically flat; endophallus: apical distal lobe modified into relatively narrow sclerotized hook, spiculae forming this hook adjunct to common basal sclerotization ( fig. 11A, B).
Redescription. Body length 9.1-12.9 mm (n = 10). Head subovate to round, smooth, brown to black. Pronotum with transverse elevation followed by impression. Elytra black. Mesoepisterna and epimera with scattered punctation, black; abdomen black. Aedeagus: median lobe with round apex; endophallus with twin-coned spherical dorsal lobe covered with medium-sized denticles and transverse ventral lobe with tiny granulation, pair of spherical lateral lobes, and apical distal lobe modified into relatively narrow sclerotized hook, bent dorsad; dorsal side of hook formed by relatively big long ligament sclerite; pairs of spiculae forming hook adjunct to prolonged common basal sclerotization (figs 3C, 11A, B).
Comparison. Among all other species of Oxyporus in Russia, O. niger is most similar to O. maxillosus from which it clearly differs by the lack of yellow spots on the elytra as well as by the distinctly longer and narrower hook-forming spiculae of the endophallus. Oxyporus niger is somewhat similar to O. mannerheimii in coloration but differs from the latter by the smooth pronotum without transverse ridge or impression, by the scattered and small punctation of the mesopleurites, not to mention the different structure of endophallus.
Comparison. Among all species of Oxyporus in Russia, O. procerus is easy to distinguish by the large size and unique coloration.
Bionomics. Data on larval stages, life cycle duration and known fungal hosts for this species can be found in Tokareva et al. (2020).
Comments spherical dorsal lobe covered with medium-sized denticles and two slightly transverse ventral lobes covered with tiny granulation, pair of relatively big spherical lateral lobes, and membranous apical distal lobe bent ventrad, with pair of ventral spiculae tapering towards apex and pairs of dorsal and additional spiculae forming a complex, broadening into weakly sclerotized oval supporting walls of the distal lobe (figs 3D, 14C, D).
Comparison. Among all other species of Oxyporus in Russia, O. rufus is easy to distinguish by its stable and distinct coloration. In case of color aberration, this species is easily distinguished by the apical distal lobe of endophallus, which is curved ventrad and has central sclerites widening towards apex.
Bionomics. A relatively complete account of all earlier published data concerning bionomics of this species can be found in Tokareva et al. (2020).
Comments. Oxyporus rufus is here confirmed as a single, widespread species based on the examination of dna barcodes from a large sample of specimens. from which it differs in the wider triangular black band on elytra and in the structure of the apical sclerites of the endophallus (e.g., the absence of the strongly sclerotized apical hook).
Comments. Oxyporus triangulus, described from Japan, was only recently recorded for Russia (Shabalin, 2012). Examination of the endophallus of the Japanese and Russian specimens confirmed their conspecificity and thus corroborated Shabalin's record. Unfortunately, none of the available dried specimens of this species produced barcode sequences.
Redescription. Body length: 7.5-8.4 mm (n = 3). Head round to subovate, smooth, black and red. Pronotum smooth, without distinct protuberances on posterior angles. Scutellum with several faint punctures. Elytra with irregular rough punctation, metallic blue with red-yellow humeri. Mesepisterna and epimera smooth, red. Abdominal segments i-vi pale yellow to dark red, sometimes with black areas on tergite and sternite vi, segments vii-ix dark brown to black. Aedeagus: median aedeagal lobe with oval aperture with membrane at dorsal side of basal bulb and with acute apex; endophallus: with single dorsal circular lobe with fine granulation and twin-coned ventral lobe subdivided by transverse line covered with medium-sized denticles; apical distal lobe ventrally curved and consisting of three solid sclerites forming two shells and apical hood, apical membranous shell with massive distinct sclerite inside protruding from hood (figs 4B, 15C, D).
Comparison. This species is easy to distinguish from any other Oxyporus in Russia due to its spectacular coloration alone.
Comments. This rare and remarkable species, described from Russia, was later revised by Makranczy (2012), where the photo of habitus, detailed description, and figures of endophallus inside the median lobe were presented for the first time. Even though at the moment it is very easy to distinguish this species based on the coloration alone, here we report the structure of its fully everted endophallus and CO1 sequence for the first time. Diagnosis. Pronotum red, with conspicuous protuberances on posterior angles, elytra tricolored with reddish-yellow markings on humeri and pale yellow posterior edges. Aedeagus: median lobe with two-pointed apex; endophallus small, with vermiform ventral lobe and four sclerotised shells not surrounded with membrane as apical sclerites.

Oxyporus (Pseudoxyporus) dybowskii
Redescription. Body length: 6.1-8.5 mm (n = 4). Antenna piceous to light brown, longer than head, narrow; antennal segments reverse semi-triangular. Head round to subovate or slightly rectangular, smooth, black; frons with two longitudinal depressions behind antennae. Pronotum red, with distinct protuberances on posterior angles and thus slightly trapezoid. Scutellum smooth. Elytra black with reddish-yellow point on humeri and pale yellow posterior edge. Mesoepisterna and epimera smooth, black. Abdominal segments iii-vi reddish to dark red, sometimes with black areas on middle line of each tergite and almost fully black tergite vi, segments vii-ix dark brown to black. Aedeagus: median lobe with two-pointed apex; endophallus: rather small, with single ventral circular lobe with tiny granulation and vermiform dorsal lobe without granulation; apical distal lobe ventrally curved and consists of four solid sclerites forming two pairs of shells, ventral pair bigger than dorsal, each sclerite of this pair apically curved (figs 4E, 16A, B, 17).
Comparison. Oxyporus (P.) dybowskii is close in coloration to O. (O.) rufus, but can be easily distinguished by the elevated posterior angles of pronotum (figs. 17).
Comments. This rare species, described from Russia, was revised by Makranczy (2012), who provided a photo of the habitus, detailed description, and the figures of endophallus inside the median lobe for the first time. Even though at the moment this is a very easy to distinguish species based on coloration alone, here we report the structure of its fully everted endophallus for the first time because of its high potential for species diagnostics. Based on the description of external morphology, habitus photo and drawing of the aedeagus of O.(P.) pulchellus in its original description (Huang et al., 2006), we suspect that this species, described from two specimens from the Mt. Changbai from Jilin Province of China, may be conspecific with O.(P.) dybowskii. Kirschenblatt, 1938 (figs 4C, 16C, D) O. melanocephalus: Kirschenblatt, 1938: 529, 534;Makranczy, 2012: 122;Shavrin and Gildenkov, 2009: 123;Shabalin, 2012: 415;Tokareva et al., 2020: 257. Diagnosis. Head black and pronotum and elytra piceous red. Aedeagus with oval aperture with membrane at dorsal side of basal bulb; endophallus with relatively big vermiform ventral lobe and apical sclerite surrounded with membrane.

Oxyporus (Pseudoxyporus) melanocephalus
Redescription. Body length: 8,1-12,8 mm (n = 3). Antennae black, longer than head, narrow; antennal segments distinctly widened apicad. Head round to subovate or slightly rectangular, smooth, black; frons with two longitudinal depressions behind  Siberia to Altai in 2019 where some material used in this paper was collected. That expedition was funded by the Carlsberg Foundation grant. Special thanks to Maria Gaianova (Ludwig-Maximilians-Universität) and team members of the resource center "Chromas" at Saint Petersburg State University (spbu) for consultation of and support for the molecular part of the research. We acknowledge the spbu resource centers "Chromas", "Molecular and Cell Technologies" and "Centre for Microscopy and Microanalysis" for access to laboratory equipment and technical assistance. Thanks to three anonymous reviewers and comments from the editor, the manuscript has been considerably improved after the submission. This paper is part of the project 20-14-00097 funded by the Russian Science Foundation.