The first molecular insight into the genus Turanium Baeckmann, 1922 (Coleoptera: Cerambycidae: Callidiini) with a description of a new species from Middle Asia

This paper sheds the first light on the phylogeny of the Central Asian genus Turanium Baeckmann, 1922. By applying an integrative taxonomy approach, we revealed and described a new species from Kyrgyzstan—Turanium losi Karpiński, Plewa & Hilszczański sp. nov. Distinguishing characters from closely related Turanium pilosum (Reitter, 1891) are presented and their ecological associations are discussed. The key characters, including the male terminalia, were examined by means of scanning electron microscopy. High-quality stacked photographs of the habitus of the specimens are presented for both species and their geographical distributions are mapped. While the new species shows stable morphological characters that allow its differentiation from T. pilosum and the COI genetic distance between them is approx. 3%, the different species delimitation methods gave discordant results. Although the new species remained unrecognized for so long, it seems that these cerambycids are common in the region and both can be considered potentially invasive as they are apparently highly polyphagous. It has also been documented that they occur sympatrically in Kyrgyzstan. Both the Bayesian and maximum likelihood analyses of COI sequences confirmed the monophyly of the genus Turanium with strong support (PP 1 and BS 90, respectively). Moreover, the recently revealed polyphyly of the tribe Callidiini was supported by our analyses and, consequently, the discussion on the establishment of a new tribe Ropalopini is raised. This study further corroborates the effectiveness of DNA barcoding as a tool in detecting new species and provides some of the first sequences for Central Asian cerambycids, which remain almost completely unknown in terms of molecular studies.


Introduction
The genus Turanium Baeckmann, 1922 is a small group of the family Cerambycidae, subfamily Cerambycinae that is distributed in Central Asia-from northeastern Iran (Bojnord) to central and southeastern Kazakhstan (Dzungarian Alatau) and western part of Kyrgyzstan, including almost the entire territory of Tajikistan (Danilevsky 2001;Sama et al. 2008). This so far relatively poorly studied group comprises seven hitherto described species that were classified into two subgenera: Chalcoturanium Jankowski, 1934 and Turanium (Tavakilian and Chevillotte 2021). Five species (i.e. Turanium hladili Kratochvíl, 1985, Turanium pilosum (Reitter, 1891, Turanium rauschorum Holzschuh, 1998, Turanium scabrum (Kraatz, 1882, and Turanium tekeorum Danilevsky, 2001) were placed in the nominative subgenus, while Turanium badenkoi Danilevsky, 2001 and Turanium johannis Baeckmann, 1922 belong to Chalcoturanium. The latter species was hitherto commonly divided into two subspecies: T. j. johannis and Turanium johannis juglandis Jankowski, 1934. Interestingly, despite the latter taxon has been synonymised with T. johannis (Danilevsky 2001), it was later still listed as a valid subspecies by the same author (Danilevsky 2020). Plavilstshikov (1940) was the first who summarised information about this genus and proposed a key for four species that were known at the time. However, since he considered T. johannis and T. juglandis as valid species, according to the current taxonomy, his key includes in fact only three of the seven (now eight) described species. The most recent and very informative study of this genus, together with the description of two new species, was presented two decades ago by Danilevsky (2001). However, although all so far known species have been presented, some key elements of a revision are missing, such as the presentation of male genitalia or photographs of the type material for all species. Therefore, a new and more comprehensive revision containing all these elements and a reconstruction of a possible phylogeny based on morphological and/or molecular data is needed.
DNA barcoding is a technique that involves sequencing of 658 bp from the 5' end of the mitochondrial cytochrome oxidase subunit I gene (Hebert et al. 2003). Although there are various situations when the use of this method or mitochondrial markers in general can be seriously limited (e.g. hybridising species, maternally-inherited endosymbionts) (Raupach et al. 2020), numerous studies across a broad range of taxa have already proven that DNA barcoding provides a useful tool to speed up taxonomic procedures, such as automatic species identification by matching new sequences to existing taxa in reference libraries (e.g. Schmidt et al. 2015;Grebennikov et al. 2017;Coral Şahin et al. 2019;Kelnarova et al. 2019;Schmid-Egger et al. 2019;Çakmak et al. 2020), associating dimorphic sexes (e.g. Zhai et al. 2017;Zhang et al. 2019) or different life stages (e.g. Miller et al. 2005;Zhou et al. 2007;Stur and Ekrem 2011;Karpiński et al. 2021), as well as differentiating of closely related or sibling species (e.g. Hebert et al. 2004;Vieites et al. 2009). This method, however, should always be supported by other evidence (particularly morphology). International Barcode of Life Data Systems (BOLD; www.boldsystems.org) is a public and freely available database that can serve as a great tool for storage, acquisition, analysis and publication of DNA barcode records. Among various analytical tools implemented in the BOLD workbench, one of the key elements is Barcode Index Number (BIN) that allows barcodes to be analysed using the system that clusters them to produce operational taxonomic units (Ratnasingham and Hebert 2007).
During the routine barcoding of longhorned beetles distributed in Central Asia, which remain almost completely unknown regarding their molecular sequences, we noticed that the specimens of 'Turanium pilosum' that were sampled in Tajikistan and Kyrgyzstan formed two well-defined clades. Hence, this study aimed at investigating (morphology and species delimitation methods) the existence of a possible new taxon and at better understanding the taxonomic position of this group by revealing results of the first molecular-based (COI) phylogeny of the genus Turanium and providing the sequences for further studies. Taking advantage of the opportunity, we also decided to examine the phylogenetic position of the morphologically closest genera of the tribe in trying to test the recently questioned (Lee and Lee 2020) monophyly of Callidiini.

Material examined
The individuals that were used for the detailed morphological and molecular analyses were collected by the authors during the entomological expeditions to Tajikistan (2014), Kazakhstan (2017Kazakhstan ( , 2018 and Kyrgyzstan (2018Kyrgyzstan ( , 2019. Some of the localities that were posted on professional websites (www.cerambyx.uochb.cz; www. zin.ru/animalia/coleoptera) were taken into consideration and presented only when combined with photographs, enabling accurate identification. After publication, the holotype will be deposited in the collection of MIZ.

Morphological analyses
The beetles were examined using an Olympus SZH10 Stereo Microscope at 7-140× magnification, a PROLAB MSZ Stereo Microscope at 7-90× magnifications, and a Hitachi S-3400N Scanning Electron Microscope. To examine the sclerotized parts of the male terminalia, the specimens were relaxed in distilled water for 12-24 h at room temperature. Then, the genitalia and last abdominal segment were separated from the other abdominal structures using pins or forceps, without removing the rest of the abdomen. Separated genitalia were put into 15% KOH solution at room temperature for some 24 h.

Photography and preparation of the figures
The scanning electron microscope (SEM) images were taken using a Hitachi S-3400N SEM at MIZ. Photographs of the habitus were taken with a Canon EOS 50D digital camera equipped with a Canon 100 mm f/2.8 USM Macro lens. The images that were produced were stacked, aligned, and combined using Helicon Focus (www.heliconsoft.com) and Zerene Stacker (www.zerenesystems.com) software. Photographs of the habitats were taken with a Canon EOS 600D and a Nikon Coolpix AW110 cameras. All plates were prepared using Adobe Photoshop CS5 and GIMP v2.10.14. The distribution of the species was illustrated in Quantum GIS (QGIS) v3.6.0 'Noosa' (QGIS Development Team 2021) using National Geographic World Map (National Geographic et al. 2021) as the raster layer. Geographical coordinates of the localities are given in decimal format and WGS84 projection.

Molecular analyses
DNA barcoding, the analysis of a standardised segment from the 5' end of the mitochondrial cytochrome c oxidase subunit I (COI) gene, was performed on a representative selection of species (four of eight including members of both subgenera; six specimens) of the targeted group: T. pilosum (Kyrgyzstan), T. losi Karpiński, Plewa & Hilszczański sp. nov., T. scabrum (Kazakhstan), and T. johannis (Kyrgyzstan) (Table S1). Additionally, we barcoded specimens of two extremely rare Central Asian species from the most closely related genus Ropalopus Mulsant, 1839: Ropalopus nadari Pic, 1894 and Ropalopus mali Holzschuh, 1993, which belong to the same tribe (Callidiini) and occur in the same habitats. Sequencing of some additional specimens that were collected, killed and preserved under the same conditions was unsuccessful, however, each species tested has obtained at least one sequence of a satisfactory length. We chose this gene because it has proven very informative in our previous studies (e.g. Plewa et al. 2018;Karpiński et al. 2021), and COI sequences of many species representing the most closely related genera (Cerambycini, Clytini, Hylotrupini) already exist in BOLD (www.boldsystems.org) and GenBank (www.ncbi.nlm.nih.gov/Genbank). The individuals that were reared from the inhabited material collected in the field (T. losi Karpiński, Plewa & Hilszczański sp. nov., Kyrgyzstan) were preserved in 96% ethanol, which was subsequently replaced in order to avoid diluting the alcohol. For the remaining species, only dried specimens (collected between 2014 and 2019; killed with ethyl acetate) were utilised. The specimens were processed for DNA barcoding in 2020. The right mid femur was cut open on both ends to expose the muscle tissue and then partly crushed with forceps and placed in a sealed well that contained two drops of 95% ethanol on a standard 96-well microplate, which was used for the tissue submission.
All of the laboratory work for extracting, purifying, amplifying and sequencing the DNA was performed at the 'Canadian Centre for DNA Barcode' (CCDB, http:// www.ccdb.ca), University of Guelph, Ontario, Canada, following the standard protocol (Ivanova et al. 2006).
Due to the quality of the material, only eight specimens were successfully sequenced for a 619-658 bp long DNA barcoding fragment. The sequences were submitted to GenBank under the accession numbers OK073067-OK073074 (Table S1).
The obtained sequences and additional relevant information such as the specimen images, primers, gel images and trace files were uploaded to the 'Barcode of Life Database' (=BOLD, http://www.boldsystems.org) in the public online dataset 'Turanium Central Asia LK' (DS-LKCCAT; DOI: dx.doi.org/10.5883/DS-LKCCAT). All of the voucher specimens reported herein are part of the BOLD project 'LECHK Cerambycidae Central Asia L. Karpinski' and were deposited in the LK's Cerambycid DNA-grade specimen bank at MIZ and in the collection of Forest Research Institute, Poland.

Phylogenetic analysis
Phylogenetic trees were reconstructed using both Bayesian inference (BI) and maximum likelihood (ML) methods. The obtained trees were also used to infer possible species delimitations. In addition to the barcod ed specimens, sequences of the following species of Cal li di ini (11 species of 6 genera) and the most closely related tribes (16 species of 6 genera in 3 tribes) were obtained from Gen-Bank and BOLD as closest related outgroup: Callidiini-  (Rambaut 2014). Most of the selected species represent a wide range in the Palaearctic-from Europe (e.g. R. femoratus) to the Russian Far East, Korean Peninsula and China (O. cinnabarinum), and Japan (S. japo nicus), although a Nearctic species (R. san guinicollis from Canada) was also considered for broader taxon sampling.
Bayesian analysis was performed using MrBayes v3.2.7a (Ronquist et al. 2012): four chains (one cold and three heated) and two runs of 10 million generations with default prior settings were conducted. The stationarity and convergence of MCMC were assessed in Tracer v1.7.1 (Rambaut et al. 2018), as well as by the examination of Potential Scale Reduction Factor (PSRF) values and Average Standard Deviation of Split Frequencies in the MrBayes output.

Species delimitation
Distance-based and tree-based species delimitation methods were performed to investigate species boundaries of T. losi Karpiński, Plewa & Hilszczański sp. nov. In order to receive a robust estimate of its entities, three approaches were used: Assemble Species by Automatic Partitioning (ASAP; Puillandre et al. 2021), Poisson Tree Process (PTP; Zhang et al. 2013) and its modification-Bayesian implementation of the Poisson tree processes (bPTP).
The ASAP distance-based analyses were run on the 'ASAP web' server (https://bioinfo.mnhn.fr/abi/public/ asap) using Kimura 2-Parameter as a substitution model for matrices of genetic distances. Only partitions showing the lowest asap-score were considered.
The best-score ML tree from RAxML was used as input for PTP, with the most distantly related outgroup removed. ML inference was performed using the single-rate method and was run with mPTP v0.2.4 (Kapli et al. 2016). MCMC analysis was performed for 100 million generations, sampling parameters every 10 thousand generations and the first 10% generations were used as burn-in. The default value for the -minbr parameter was used, and analyses started from the ML random delimitation estimate.
The tree from MrBayes was used for bPTP analysis which was run on the 'bPTP server' (species.h-its.org). Similarly to the previous analyses, the most distantly related outgroup was removed. MCMC chains were run for 500 thousand generations, and all other settings were left as default.

Taxonomy
Initiated by the conspicuous divergence in COI sequences, detailed morphological studies of the specimens representing both sequenced populations from Tajikistan and Kyrgyzstan confirmed the existence of two closely related species. Subsequently, additional individuals representing populations from other localities in the region were investigated. The results of this study are presented as both SEM and stacked colour plates. The general habitus of the beetles is presented dorsally (Fig. 1) and ventrally (Fig. 2). Particular body parts illustrating the key characters are presented in Figs 3-6. Additionally, male terminalia (lateral lobes of tegmen) are shown in Fig. 7. The habitus of the holotype of Turanium pilosum (Reitter, 1891) is depicted in Figs 8 and 9. The comparison of our material with the Reitter's type (HNHM) revealed that the barcoded specimens that represent the population from Tajikistan (Takob) belong to the existing species, and the sequenced specimens from Kyrgyzstan (vicinity of Arkit village) represent a new species of the genus Turanium. However, subsequent examination of additional individuals and Internet resources not only revealed the pres-  Description. Morphology. Body relatively slender, strongly flattened dorsoventrally; BL in males: 10.5-14.0 mm (HT 12.0 mm), in females: 11.0-15.0 mm. Humeral width in males: 3.0-3.9 mm (HT 3.5 mm), in females: 3.0-4.2 mm. Integument black; legs and antennae black; elytra black (sometimes with slightly brownish areas on sides behind middle) with slight metallic luster. Pubescence of whole body made by sparse but distinct, long and usually erect whitish hairs and short black setae, being the longest on pronotum and anterior part of elytra, in its posterior part with many shorter and semi-decumbent hairs especially along epipleura; on ventral side distinct long whitish hairs, varying in intensity among specimens but usually denser on abdominal sternites; on femora semi-decumbent, relatively long and sparse whitish hairs; on tibiae and tarsi mostly replaced by denser blackish setae forming brush; on antennae very dense, relatively long, erect setae, especially rich on first five antennomeres and gradually disappearing towards last joint, on last antennomeres in form of single spike-like setae. Head broad, with coarse sculpture; frons with longitudinal furrow of variable depth between antennal tubercles; clypeus and labrum broad and well-pronounced; mandibles and palpi stout; eyes large, surrounding antennal tubercles; genae narrow, approx. 0.15-0.2 of eye width. Antennae thick and strong, relatively long, exceeding elytral apex by 1.5-2 last antennomeres in males, and clearly shorter, never reaching elytral apex in females; average length ratio of antennomeres in males: 1:0  shorter than elytral length and 1.2 (HT 1.2) in males and 1.26 in females times narrower than elytra at humeri. Pronotum rather regularly (less regularly in males) and en-tirely punctate, in males usually with smooth area at middle near base; punctation rather uniform (more uniform in females), relatively sparse but clearly finer and denser at sides. Prosternum finely and sparsely punctate, distinctly creased longitudinally. Prosternal process short and very thin, barely exceeding middle of procoxae, slight-ly curved dorsally. Elytra moderately long, approx. 2.25 (HT 2.22) in males and 2.33 in females times as long as humeral width, gradually tapering towards apex in males and usually almost parallel-sided in females, in both sexes with single indentation on each elytron about 1/3 of anterior length; elytral sculpture composed mainly by regular, sparse, fine, shallow punctures, with coarse surface between them (Fig. 4K-M), rather uniform at entire length; scutellum of variable shape, rather triangular but usually with rounded edges, covered with whitish hairs. Tarsomeres similar in both sexes; protarsomeres pronounced, lobes of protarsomere 3 elongate and cordate, protarsomere 5 slightly longer than 1; metatarsomere 1 about as long as 2 and 3 combined and slightly longer than 5; tarsal pads broad, with numerous compact and very dense setae (Fig. 5M). Femora flattened similarly in both sexes. Metatibiae in males clearly exceeding elytral apex, slightly shorter in females. -Male terminalia. Lateral lobes of tegmen (Fig. 7A, B) short and thin, generally slightly tapering towards apex but sometimes more or less variable in shape, with rather long and thick setae apically; manubrium elongated, with broad edges. Median lobe of variable shape, usually robust and relatively short, gradually tapering towards apex.  -F, 8, 9), principally by the different outer edges of the pronotum that are clearly rounded and not parallel-sided (Fig. 3B vs. E), the coarse punctation of the elytra (Fig. 3H vs. K), the different color of the pubescence (especially on antennae and elytra) that is made by whitish hairs and black setae (not brownish and russet as in T. pilosum) and its much stronger abundance on the ventral side of the body (Fig.  5G vs. H-J; although Kyrgyz individuals of T. pilosum seem to be slightly more densely pubescent on the ventral side: Fig. 2D), as well as the different color of the integument, which is black in the new species and usually brown or brownish in T. pilosum. Tarsal pads are broader in the new species, with pronounced, clearly more compact and abundant, longer setae, with a narrower separating line along middle, which sometimes may not be even visible on the protarsomere 2 (Fig. 5M). Moreover, regardless of body size, all studied male specimens of T. losi Karpiński, Plewa & Hilszczański sp. nov. have clearly visible and almost completely developed antennomere 12 (Fig. 6B), while its origin can be barely detectable even in the utmost males of T. pilosum (Fig. 6A). Female antennae are distinctly shorter in the new species, clearly not reaching the elytral apex. Generally, the body of the new species is more corpulent and of a bigger size, with the smallest male measuring 10.5 mm (avg. 13 mm) and the biggest females reaching 15 mm (avg. 14 mm), while all studied specimens of T. pilosum are within the range of 10-12 mm including females. Although this character appears to be variable, it is worth to mention that most specimens of T. pilosum lack the longitudinal furrow on the frons, or it is only slightly marked, while all studied individuals of the new species have this furrow present and it is strongly marked in the majority of specimens. Similarly with the prosternal process; although it is usually wider in the new species (Fig. 3M vs. N, O), this difference appears to result from individual variability and it requires more caution. Regarding male genitalia, while many individuals of both species show sufficiently strong differences in the shape and pubescence of the lateral lobes (Fig. 7A, B vs. D-F), the structures of some individuals differ significantly from the pattern (Fig. 7G), objectively making this character too variable for use in species separation. Similar situation was confirmed also in the individuals of the closely related genus Ropalopus (Karpiński et al. 2020) and in the genus Anoplistes Audinet-Serville, 1833 (Cerambycinae: Trachyderini) (Karpiński et al. 2021)-some individuals of closely related but clearly independent (morphologically, ecologically, and genetically) species may share almost identical parameres, not to mention the highly variable shape of the median lobe. The molecular distance in COI sequences of T. losi Karpiński, Plewa & Hilszczański sp. nov. and T. pilosum was also compared and it reaches almost 3%, with almost no intraspecific variability (Table 1).   The new species can also easily be separated from T. rauschorum (habitus of male paratype presented in Danilevsky (2001)), another 'blackish' species of the discussed group that is distributed in the region, by the much longer antennae in both sexes (in males of T. rauschorum about as long as body and reaching posterior elytral fourth in females), by the different shape of pronotum in males ('always without smooth areas' in T. rauschorum), and by considerably bigger body dimensions, which in huge series (75 specimens) of T. rauschorum vary in males between 7.9-12.4 mm, and in females between 11.0-12.0 mm (10.5-14.0 mm and 11.0-15.0 mm, respectively in T. losi Karpiński, Plewa & Hilszczański sp. nov.). Furthermore, T. rauschorum is considered a monophage restricted to Atraphaxis (Polygonaceae) and it is only known so far from southern Kazakhstan and one locality in northernmost Kyrgyzstan (Danilevsky 2001).
Other species in the genus clearly differ in body proportions and colouration, the punctation of the pronotum and elytra, and other characters. The habitus of T. scabrum and T. johannis is presented in Fig. 1 (I, J and K, L, respectively), and their lateral lobes of tegmen in Fig.  7. Their COI sequences have also been compared and the molecular distance is remarkable (Table 1), even despite rather poor divergence in their parameres (Fig. 7H vs. I). Karpiński, Plewa & Hilszczański sp. nov. is known to occur in northwestern Kyrgyzstan and easternmost Uzbekistan (Fig. 10), however, this species is most likely widely distributed in the  region, including southern Kyrgyzstan and northern Tajikistan since the high mountains of Tian Shan (Turkestan Range) seems not to be the sufficient distributional barrier for ecologically very close T. pilosum.

Bionomics.
Adults of the new species were collected from the first days of May to the third quarter of June, at altitudes between 1000 and 1900 m a.s.l. The earliest observation was done at the lowest altitude (May 6, at 994 m a.s.l.). The holotype and the main series of paratypes have been reared from larvae collected in mid-June at an altitude of 1500 m a.s.l. At the time, imagoes of T. pilosum (exclusively males) were collected in the same locality and despite careful investigation no adults of the new species were observed at this plot. At the higher altitude (1620 m a.s.l.), females of the new species were found on June 20, together with a single male of T. pilosum. In type locality in Kyrgyzstan the species is related to mountain deciduous open forests with a substantial share of Fraxinus L. (Oleaceae), Prunus L., and Malus L. trees (Rosaceae) (Fig. 11A, B). In Tajikistan (at 1850 m a.s.l.), adults of T. pilosum were mating in the mid-July in a mountain valley with Juglans L. (Juglandaceae) and Malus trees (Fig. 11C, D) (Kadyrov et al. 2016). Considering the above, it seems reasonable to conclude that T. losi Karpiński, Plewa & Hilszczański sp. nov. is a phenologically relatively early species emerging in the first decade of May at lower elevations, and surviving in nature to the end of June at higher altitudes, while T. pilosum occurs later-from the mid-June to the end of July or even to the first decade of August in higher located sites.
The new species seems to be common in the region and its larvae are most likely wide polyphages as the adults that were reared from Kyrgyz Prunus L. and Malus L. were able to continue breeding on the wood of European hornbeam Carpinus betulus L. (Betulaceae). Moreover, Danilevsky (2001) listed numerous host plants (both deciduous and coniferous) for 'T. pilosum' that certainly include some records of the new species (see more in Remarks). The depicted male specimen was collected on June 2 and the female was reared from Picea A. Dietr. (Pinaceae).
Based on own observations, the life cycle usually lasts two years, however in the laboratory rearing, a single male emerged from the wood material that was added already in Poland (II generation) after only eight months, which means that the cycle can be shortened in nature to one year under optimal conditions. Remarks. As T. losi Karpiński, Plewa & Hilszczański sp. nov. is rather common in Middle Asia, there were already some specimens imaged in the Internet or in scientific papers that present the new species but were clearly misidentified. One example could be a male of 'T. johannis' from Uzbekistan (www.cerambyx.uochb.cz; accessed on: 10.01.2021, the image has been replaced) and another a female of 'Turanium rauschorum' from Kyrgyzstan, close to the Kazakh border (www.zin.ru). Regarding the revision of the genus (Danilevsky 2001), although the quality of the images in the distributed version of this paper does not allow the reliable identification of the depicted specimens, it seems that both presented 'T. pilosum' (fig. Etymology. We are pleased to dedicate this species to a Polish entomologist, Krzysztof Łoś, our close friend and one of the main organizers of the 2017-2019 trips to Kazakhstan and Kyrgyzstan.

Phylogeny
Two following partitions were found: COI1 + COI2 and COI3. For MrBayes SYM+I+G model was selected as the best supported for the first one and HKY+G for the second. The maximum likelihood (Fig. 12) and Bayesian (Fig. 13) inference trees showed similar topologies, however, most likely due to insufficient data (COI only), some clades in the Bayesian tree were not resolved. Both the ML and BI analyses of the molecular sequences revealed the monophyly of the genus Turanium with strong support (BS 90 and PP 1, respectively). Turanium losi Karpiński, Plewa & Hilszczański sp. nov. from Kyrgyzstan and T. pilosum from Tajikistan formed a well-defined clade with almost maximal support in both trees. The resemblance of these two taxa is strongly supported morphologically, which clearly indicates close sister relationships between them. Turanium scabrum (Kazakhstan) of the same subgenus was revealed as more closely related compared to early-branching T. johannis (Kyrgyzstan) of the subgenus Chalcoturanium, which was placed as a sister group to all remaining Turanium species. Interestingly, the genus Ropalopus was revealed as paraphyletic in both analyses, although the nodes were moderately or weakly supported and the representation of this genus might not be comprehensive enough to draw far-reaching conclusions (see more in the Discussion). The analyses further indicated that both Central Asian Ropalopus: R. nadari and R. mali were clustered together with strong support (BS 91, PP 1), in the position of a sister group to the Turanium clade (BS 59, PP 0.79). It also indicated R. sanguinicollis as sister to the clade of all remaining Ropalopus and Turanium in both trees (BS 66, PP 0.90).
Callidiini was recovered as polyphyletic, separated in a few clades, although generally with weak support value of each subgroup: Phymatodes + Pyrrhidium as a part of Clytini, and Callidium and Ropalopus + Turanium as separate (ML) or unresolved (BI) clades.

Species delimitation
For additional verification, we also tested the putative new species using a few species delimitation methods. Different methods gave, however, discordant results, either confirming the species status of T. pilosum and T. losi Karpiński, Plewa & Hilszczański sp. nov., accepting them partly, or rejecting them as distinct taxa. Turanium losi Karpiński, Plewa & Hilszczański sp. nov., besides strong morphological support, was confirmed as a separate species in the ASAP method. On the contrary, the PTP method recognised both Turanium as a single taxon, while the bPTP method gave inconclusive results, supporting their divergence in 72% of analyzes (while the author of the method suggests treating results as sufficiently accurate with the support of approx. 0.90). It is worth mentioning, however, that two species of the genus Plagionotus Mulsant, 1842: Plagionotus arcuatus (Linnaeus, 1758) and Plagionotus detritus (Linnaeus, 1758) have also been revealed as a single species by this method, with very similar support (0.78) as for Turanium taxa. Similarly, in the PTP method, both Plagionotus species were revealed as a single taxon. There are no justified reasons not to consider P. arcuatus and P. detritus as two separate species. The COI genetic distance between them is approx. 3%, nearly the same as in case of T. pilosum and T. losi Karpiński, Plewa & Hilszczański sp. nov. (Table S2). We also investigated that the authors of the original paper (Hendrich et al. 2014) have identified those species correctly (their sequences were submitted also to BOLD, where images of exact specimens are available). A genetic distance threshold may exist below which the delimitation methods are not able to resolve relationships based on single specimens.
The results from each method are presented on the Bayesian inference tree (Fig. 13) for the entire Ropalopus + Turanium clade and for both Plagionotus species.

Plausible separation of the tribe Ropalopini
Callidiini is a cosmopolitan tribe comprising 30 genera (41 including subgenera) (Tavakilian and Chevillotte 2021). After exclusion of Callidium Fabricius, 1775 and Clades within Callidiini were marked by coloured branches as follows: purple-Turanium, blue-Ropalopus, red-remaining genera. Additional interpretation for the trib labels on the right side: purple-Hylotrupini, blue-Cerambycini.
Semanotus Mulsant, 1839, which in our study were separated from Turanium Baeckmann, 1922 and Ropalopus Mulsant, 1839 and clustered in Lee and Lee (2020) together with Oemini as a sister group to Phoracanthini + Dichophyiini, and after further exclusion of Phymatodes Mulsant, 1839, which was positioned in our study within Clytini and as a sister clade to Compsocerini + Hylotrupini in Lee and Lee (2020), Ropalopus is the oldest established genus of Callidiini. It was described in the same year as Phymatodes and Semanotus. In our study, the genera Ropalopus and Turanium formed a separate clade, which, however, was not resolved by Bayesian inference. Therefore, although we do not propose herein any changes to the tribal classification, if the distinctiveness of the discussed clade (and perhaps some other yet untested genera) is confirmed in more densely sampled multilocus molecular analysis or well-designed morphological analysis, it will be reasonable to establish a new tribe Ropalopini. The external morphology of adults seems to confirm this affinity, and according to Švácha and Danilevsky (1988), Turanium larvae are similar to those of Ropalopus macropus. Some representatives of molecularly unstudied genera: Callidiellum Linsley, 1940, Leioderes L. Redtenbacher, 1845, Pronocera Motschulsky, 1859, as well as additional species of Ropalopus (especially the type species-R. clavipes and some taxa of the 'ungaricus/insubricus' group) and Semanotus definitely need to be sequenced and compared to obtain a relatively objective phylogeny reconstruction.

Phylogenetic positions of Turanium and Callidiini
Besides the monophyly of the genus Turanium our analyses revealed the polyphyly of the tribe Callidiini. Although a COI-based phylogeny should not be overinterpreted, we decided to present and discuss some parts of our trees as very similar results were obtained by Lee and Lee (2020) in the analyses based on six molecular markers (mitochondrial protein-coding COI, mitochondrial ribosomal 16S rRNA, nuclear ribosomal 18S rRNA and 28S rRNA, and nuclear protein-coding wingless and CAD), who proposed the most sophisticated to-date tribal and generic relationships within Cerambycinae. In their study, most of the sampled cerambycine tribes were rendered as monophyletic except for Callidiini, Callidiopini, Cerambycini, Cleomenini, and Phoracanthini. Callidiini was recovered as polyphyletic and divided into two distinctly separated and distant clades with high support value of each subgroup: (I) Phymatodes and (II) Callidium + Oupyrrhidium + Semanotus. According to those authors, it will be reasonable to treat the Phymatodes-group as an independent tribe after finding morphological characters supporting the presumed monophyly of each group. In our analyses, Callidiini was separated into three clades: (I) Phymatodes + Pyrrhidium (the latter not tested by Lee and Lee), (II) Callidium, and (III) Ropalopus + Turanium (both genera not tested by Lee and Lee). To support these results, it is worth to note that, in total, six species representing five of six subgenera of the genus Phymatodes were analysed and only one (P. testaceus, type species of the genus) was common to both discussed projects. We also utilised the sequences of European P. pusillus (subg. Phymatoderus) and P. rufipes (subg. Phymatodellus), while Asian P. maaki and P. ermolenkoi (subg. Poecilium), and P. mediofasciatus (subg. Paraphymatodes) were processed in Lee and Lee (2020). Similarly, in the genus Callidium, only one (C. aeneum) of four species tested, representing all three subgenera, was utilised in both projects, however, the type species of this genus is C. violaceum, and not C. aeneum (subg. Callidostola) as it was erroneously stated by Lee and Lee (2020). We additionally processed sequences of the nominative C. violaceum and C. coriaceum (subg. Palaeocallidium). Formation of analogous clades despite the use of different markers and different species sampling seems suggestive. As it was mentioned above, neither Ropalopus nor Turanium were analysed in Lee and Lee (2020), and the inclusion of the representatives of these two genera in the combined tree would be particularly interesting, especially as Turanium is an exclusive Central Asian genus, representation of which region is absent in slightly biased towards Western and Eastern Palaearctic and Oriental regions study of Lee and Lee (2020). Moreover, the morphological traits that appear to merge the genera of Callidiini could have likely arisen as a result of convergence, as the flattened body seems very beneficial for adults when getting beneath the bark to hide from predators or unfavorable weather conditions, as well as wide, clavate femora may be useful for maintaining better adhesion when moving rapidly along branches.
Since we present here the first sequences of Turanium and, next to one Palaearctic and one Nearctic species available in GenBank/BOLD, the only sequences of Ropalopus (sole for Asian species), these results seem quite novel and except herein widely referenced paper of Lee and Lee (2020) there are unfortunately very few articles that investigate this issue. In the most recent study dealing with cerambycid phylogeny (Nie et al. 2020) the only presented Callidiini species are Semanotus bifasciatus and Pyrrhidium sanguineum (both were tested in the above discussed analyses). Moreover, as this paper aimed to solve higher-level phylogeny, the total number of taxa representing the subfamily Cerambycinae is too low to conclude on a tribal system. On the other hand, the study attempting to reconstruct a tribal level phylogeny of Lamiinae (Souza et al. 2020) revealed that a relatively high number of described tribes in this subfamily do not comprise monophyletic groups. Therefore, future studies with strong evidence in both morphological and molecular phylogenetic aspects are essential to resolve the taxonomic system of Cerambycinae, and particularly of Callidiini.

Invasive potential of Turanium
Turanium is an exclusive Central Asian genus that includes generally highly polyphagous species. For instance, according to Švácha and Danilevsky (1988), T. scabrum is a polyphage of deciduous trees recorded from Elaeagnus L., Prunus, Rosa L., Halimodendron Fisch. ex DC., Tamarix L., and others; T. pilosum is also polyphagous on deciduous trees and it was recorded from Juglans, Amygdalus L., Cydonia Mill., Malus, Armeniaca (Scop.) Koch, Cerasus Mill., Acer L., Salix L., Populus L., Ulmus L., Morus L., Sorbus L., and others; T. juglandis (=johannis) is polyphagous on both deciduous and coniferous trees: Picea, Abies Mill., Juglans, and 'apparently others'. Danilevsky (2001) provided some additional host plants for most of the species, however, the most important in this regard is T. pilosum, as T. losi Karpiński, Plewa & Hilszczański sp. nov. appears to be confused especially with this species and some locality and host plant records clearly refer to the newly described species. In the revisionary paper, T. pilosum is considered as polyphagous on both deciduous and coniferous trees, and Picea was added to its host plants list. Although species of this genus develop in branches and rather thin stems and, consequently, it is difficult to expect large economic importance, T. scabrum has been reported as a serious pest of Elaeagnus angustifolia and Populus diversifolia in tugai forests (Sinadskiy 1963). The imagoes of T. scabrum were observed in Kazakhstan mating on primarily oleaster old branches that formed a flood barrier (a com-mon solution in the region) (Karpiński et al. 2018), which in case of massive occurrence of beetles may also weaken the durability of this protection. It seems that also T. losi Karpiński, Plewa & Hilszczański sp. nov. and T. pilosum may have some economic significance, but mainly in young plantations of fruit trees.
Another issue is the invasive potential of Turanium. At least half of the species are common beetles in Central Asia and their adults, considering individual populations, are present in nature almost throughout the entire growing season. Moreover, it was documented that larvae are also able to develop in dry wood (T. scabrum; Karpiński et al. 2018). The life cycle can be shortened to one year under optimal conditions (T. losi Karpiński, Plewa & Hilszczański sp. nov., unpublished own observation). Despite the fact that the range of Turanium is restricted to the area of Middle Asia, an individual of T. johannis was reported from Ukraine (Kerch, Crimea) already in 2009 (www.insecte.org; accessed on: 24.02.2021). According to the International Monetary Fund (World Economic Outlook 2018), all countries in which representatives of this genus occur are still considered developing economies, thus, the invasive potential of these beetles may have not been fully disclosed yet, and as international trade develops in these countries, species of this genus are likely to spread. One example could be velvet longhorned beetle, Trichoferus campestris-a relatively closely related (Cerambycinae: Hesperophanini) species with a similar bionomy, native to this region but, unlike Turanium, also to highly industrialized China, that is considered a serious wood-boring pest and one of the most rapidly spreading cerambycids (e.g. Grebennikov et al. 2010;Dascălu et al. 2013;Keszthelyi et al. 2019).

Acknowledgements
We would like to dedicate this work to late Dr. Ottó Merkl, who for many years was the senior curator, and recently also the director, of the Department of Zoology, Hungarian Natural History Museum in Budapest and who was the SYNTHESYS+ host of LKa while studying the type material of Turanium. We thank our colleagues: Krzysztof Łoś (the main organizer), Tomasz Jaworski, Grzegorz Tarwacki, and Yerlan Shaperov for their participation in the entomological expeditions to Kazakhstan and Kyrgyzstan in 2017-2019. We would also like to show our gratitude to Dagmara Żyła (Museum and Institute of Zoology, Polish Academy of Sciences, Poland) for her assistance in conducting phylogenetic analyses and shared experience in molecular work. Finally, we thank an anonymous referee and the subject editor Steffen Pauls, whose detailed comments helped to improve the manuscript.