﻿Karyotype and COI gene sequences of Chironomusmelanotus Keyl, 1961 from the Yaroslavl region, Russia, and the difficulties with its identification using GenBank and BOLD systems

﻿Abstract Karyotype and COI gene sequences of Chironomusmelanotus Keyl, 1961 from the Yaroslavl region (Russia) were analyzed. A low level of chromosomal polymorphism has been confirmed, eventually eight banding sequences were found: melA1, melB1, melC1, melD1, melE1, melF1, and melG1; only melD2 was found in two larvae from the Sunoga river. Analysis of phylogenetic tree and estimated genetic distances has shown not all COI gene sequences of Ch.melanotus in GenBank and BOLD to belong to this species. The lower distance of 0.4% was observed between two sequences from the Yaroslavl region and Finland, apparently these are true Ch.melanotus sequences. The distances between true Ch.melanotus and other sequences from Finland were 9.5% and 12.4%, and from Sweden it was 11%. The average genetic distance between studied sequences of 9.1% is out of the range of the 3% threshold previously determined for chironomids. According to our estimates, there are two sequences with a distance of 2.9% that may belong to Ch.annularius Meigen, 1818, and one sequence with a genetic distance of 2.1%, may belonging to Ch.cingulatus Meigen, 1830, which has been confirmed karyologically. Another two sequences form a separate cluster. We suggest that they either belong to a known species, but are not present in the databases, or belong to a distinct, undescribed species.


Introduction
Chironomus melanotus Keyl, 1961 is one of the most widespread and well-known species. It does not belong to any sibling species group (Kiknadze et al. 2010). The first finding and description was in Germany (Keyl 1961(Keyl , 1962Degelmann et al. 1979). Ch. melanotus is in demand in classical hydrobiology (Fjellheim, Raddum 1996) and in toxicology (Grebenjuk 1994;Grebenjuk and Tomilina 2014). The main problem in the investigation of Chironomus Meigen, 1803 is the difficulties with the species identification by larval morphology. Due to the presence of the giant chromosomes in the salivary gland of Chironomus larvae, it is more convenient to identify cytogenetically (Kiknadze et al. 1991(Kiknadze et al. , 2016. The karyotype of Ch. melanotus was described by Keyl (1961) as a "cytospecies" that belongs to "thummi" cytocomplex and mapped chromosomal arms A and F (Keyl, 1962). It has been shown that the level of polymorphism in Ch. melanotus is very low (Wülker 1973;Kiknadze et al. 1991;Jabłońska-Barna et al. 2013). Only in polluted water bodies a high spectrum of somatic rearrangements and a case of trisomy were observed (Jabłońska- Barna et al. 2013). Finally, due to the development of new techniques in molecular biology, for species identification/delimitation the fast and cost-effective technology DNA barcoding is commonly used and for massive analysis in biomonitoring metabarcoding is used. In recent years, many works on this theme have been published. This is a barcoding of invertebrates, including chironomids from Canada (Hebert et al. 2016), Germany (Morinière et al. 2019), Finland (Roslin et al. 2022), South Korea (Kang et al. 2022), Montenegro/Albania (Gadawski et al. 2022) and others. The disadvantage of this approach is the presence in the databases of genetic information (GenBank and BOLD) from unidentified or incorrectly identified specimens, so-called "dark taxa" (Morinière et al. 2019). The next problem is the understanding that a sequence divergence threshold is not suitable for all Chironomus species and depends on intraspecific and interspecific sequence divergences. Interspecific -varied for COI gene sequences in most cases from 9 to 20% and in rare cases from 1 to 4% (Proulx et al. 2013). Due to the fact that we cannot fully estimate intra-and inter-specific sequence divergences, here we will use the average value of this parameter -3% (Ekrem et al. 2007;Proulx et al. 2013;Kondo et al. 2016).
We could not find any studies of Ch. melanotus involving approaches of morphology, cytogenetics and DNA barcoding published in one article. In the GenBank and BOLD databases were found five and one COI gene sequences, respectively. These sequences were obtained from individuals collected in Finland and Sweden, and deposited during the preparation of this paper (Roslin et al. 2022). Preliminary examination has shown that not all of these sequences belong to Ch. melanotus.
The present study aims to calculate and compare the genetic distances between COI gene sequences of Ch. melanotus from Yaroslavl region identified by morphology and cytogenetics and the sequences obtained from GenBank and BOLD of Ch. melanotus from different populations identified by morphology or molecular-genetics (barcode), and additional sequences from GenBank and BOLD of several Chironomus identified by cytogenetics.

Materials and methods
Fourth instar larvae of Ch. melanotus were collected from a few places in the Yaroslavl region, Russia. Thirty-one larvae were found in a puddle on the Shumarovka river shore (58°02'25.5"N, 38°15'33.2"E) in October 2018. The depth is 0.5 m, and the bottom is black silt. Seven larvae were collected in the Sunoga river (58°03'20.3"N, 38°14'04.2"E) in August 2018. The depth is 0.1 -0.2 m, and the bottom is gray silt with sand. Four larvae were collected in a small stream (brook) in the shore zone of the Kotorosl' river (57°22'41.6"N, 39°50'08.5"E) in June 2016. The depth is 0.5 m, and the bottom is black silt and rotting wood.

Results and discussion
The karyotype of Chironomus melanotus Keyl, 1961 from the Yaroslavl region, Russia The species has a 2n = 8 set of chromosomes. By the chromosome arm combination -AB, CD, EF and G, the species belongs to Chironomus "thummi" cytocomplex. The chromosomes AB and CD are metacentric, EF is submetacentric, and G is telocentric. The nucleus and Balbiani ring were found in arm G. The peculiarity of the karyotype of Ch. melanotus is a heterochromatinized centromeric region that forms an unstable chromocenter ( Fig. 1), also observed only in Ch. cucini Webb, 1969, Ch. pilicornis, Ch. athalassicus Cannings 1975, Ch. magnus White et Ramsey, 2015and Ch. hyperboreus Staeger, 1845(Wülker and Butler 1983Int Panis et al. 1994;Kiknadze and Istomina 2000;Wülker and Martin 2000;Kiknadze et al. 2010).
Arm E. One banding sequence: melE1 1a-3e 5a-10b 4h-3f 10c-13g C. Arm F. One banding sequence: melF1 1a-8f 9a-11i 12a-15i 16a-23f C. Arm G. One banding sequence: melG1. Not mapped. The level of polymorphism in Ch. melanotus is known to be very low (Kiknadze et al. 1991;Jabłońska-Barna et al. 2013). At the moment, we know of two alternative banding sequences, melB2 and melD2, and the sequence melD2 is more typical for Western Siberia populations (Kiknadze et al. 2016) and a Finnish population (Hirvenoja et Michailova 1991). We found melD2 in a heterozygous state in two larvae from the Sunoga river. Any deviations in the karyotype characteristics like a trisomy, rearrangements, insertions and deletions, more typical for polluted water bodies (Jabłońska-Barna et al. 2013) were not found.

DNA-barcoding and phylogenetic analysis
The obtained COI gene sequence for Ch. melanotus from the Yaroslavl region was deposited in the GenBank with accession number OL546775; the length of the sequence is 658 bp (percentage of nucleotides A: 25; T: 38; G: 17; C: 19).
More interesting was the analysis of COI gene sequences. As was said previously, for the species name Ch. melanotus in the databases match six sequences of the COI gene from Finland (MZ659620, MZ657748, MZ658877, MZ657558, MZ658420) identified by molecular-genetics and Sweden (BSCHI737-17) identified by imago characters, and the average genetic distance between them of 9.1% is out of the range of 3% distances previously determined for chironomids (Ekrem et al. 2007;Proulx et al. 2013;Kondo et al. 2016). Low chromosomal variability of Ch. melanotus does not allow us to talk about a high level of genetic diversity. We can conclude not all sequences belong to Ch. melanotus species (Table 1). According to our estimation, the lower distance, about 0.4%, was between Ch. melanotus (MZ659620) from Finland and Yaroslavl reg. (OL546775). The distance between sequences of Ch. melanotus (OL546775) from Yaroslavl reg. and sequences from Finland (MZ657748, MZ658877) -9.5%, (MZ657558, MZ658420) -12.4%, and Sweden (BSCHI737-17) -11%. These values are greater than those between sequences from the Yaroslavl reg. (OL546775) and Ch. anthracinus (KF278222), identified karyologically (Proulx et al. 2013), with a distance of 4%. This still doesn't mean these species are really closely related, the analysis of one short segment of the COI gene is not enough to make such conclusions (DeSalle et al. 2005). However, a high similarity of their karyotypes has been noted, up to identity of some banding sequences (Keyl 1962;Kiknadze et al. 1991).
On the phylogenetic tree constructed by Bayesian inference (Fig. 2), we see the same situation as with the genetic distances. The sequence of Ch. melanotus (OL546775) from Yaroslavl reg., and Ch. melanotus (MZ659620) from Finland combined into one cluster, while the other sequences spread out into different branches. Two similar sequences (MZ657558 and MZ658420) from Finland and Ch. annularius (AF192189.1) combined in one cluster, with a support value of 1.0. The sequence (BSCHI737-17) from Sweden and Ch. cingulatus (AF192191.1) formed another cluster, with a support value of 1.0. Two similar sequences from Finland (MZ657748 and MZ658877) have formed a separate cluster, without including any other specimens.
All the obtained data show that several species are hidden in GenBank and BOLD under the name "Chironomus melanotus". First, there is a true Ch. melanotus cluster (Fig. 2) (MZ659620 and OL546775) the reliability of which is confirmed by karyotype analysis. Probably, two similar sequences (MZ657558 and MZ658420) belong to Ch. annularius (AF192189.1), and the genetic distance of 2.9% is very close to 3% accepted interspecific threshold (Ekrem et al. 2007;Proulx et al. 2013;Kondo et al. 2016), but does not exceed it. Another sequence (BSCHI737-17), with a genetic distance of 2.1%, likely belongs to Ch. cingulatus (AF192191.1).
Two similar sequences (MZ657748 and MZ658877) need special attention. The samples of Ch. melanotus from Finland were investigated during the project of FinBOL (Finnish Barcode Of Life), in the framework of which the authors tested the system Fin-PROTAX (Probabilistic Taxonomic Assignment Tool) (Roslin et al. 2022). As the authors report, the accuracy of taxonomic assignments at the level of species reached 88.5% (Roslin et al. 2022). Such precision is still insufficient, especially in a group rich in sibling species. This approach does not consider estimate intra-and inter-specific sequence divergences. For the COI gene, the estimated interspecific sequence divergences in most cases varied from 9 to 20%, but in a few cases with well-identified by cytogenetics species, this parameter varied from 1 to 4%, which overlapped between intraspecific and interspecific sequence divergences (Proulx et al. 2013). According to Morinière et al. (2019), the database of genetic information contains about 65% of sequences without species-level assignment, so-called "dark taxa" of all Chironomidae recorded from Germany. But we think that "superficial taxonomic impediment" (species are so poorly and unreliably named, they will need to be redescribed before they can be used) (Meier et al. 2022) is better applicable in this case. Thus, we can conclude that two sequences (MZ657748 and MZ658877) belong to well-known species that are absent in databases, or they can be considered as distinct species. A similar case was with Japanese Ch. nipponensis. At first, Yamamoto (2010, cited by Kondo et al. 2016) proposed dividing the "highland" and "lowland" populations of Ch. nipponensis by morphology, then Kondo et al. (2016) revealed the genetic distances between them at 9.1%, and confirmed the presence of two separate species.

Conclusions
On the example of Ch. melanotus, we confirmed that in Chironomus species identification we must use all available comprehensive approaches, involving morphological, cytogenetic and molecular genetic studies (mitochondrial and nuclear genes) (DeSalle et al. 2005;Ekrem et al. 2007;Proulx et al. 2013;Kondo et al. 2016). At least four species of Chironomus could be in the databases under the name "Ch. melanotus" from Finland and Sweden. This suggests that at the present stage of the collection of genetic data, it is impossible to trust only a computer algorithm. We agree with Zamani et al. (2022) that the use of DNA-based analyses for an initial sorting of new and known species is extremely useful as a first step, which significantly narrows the range of search before precise species identification. Nevertheless, despite the difficulties, the species identification of Chironomus greatly enriches our understanding of ecosystem functioning because this is an important part of it.