The taxonomic status of Cossus cossus afghanistanus ( Lepidoptera , Cossidae ) from Afghanistan : insights from molecular and morphological data

In our study we use a 658 bp fragment of the COI gene to analyze a taxon from Afghanistan usually treated in literature as Cossus cossus afghanistanus (Daniel, 1953). The previous conclusions on taxonomy and nomenclature were not supported by molecular data therefore the question of identity of this taxon has remained unverified. Phylogenetic analysis revealed C. c. afghanistanus to be strongly differentiated from nominotypical Cossus cossus (Linnaeus, 1758) (p-distance: 6.7% ± 1.5%). Cossus c. afghanistanus forms a distinct well-supported clade in ML and BI trees. This fact, together with prominent morphological differences (wing color and genitalia structure) shows that C. c. afghanistanus represent a separate species rather than a subspecies of Cossus cossus.


Introduction
Cossus cossus afghanistanus (Cossus cossus afghanistana in original description) was described by Daniel (1953) as a separate subspecies on the basis of two male specimens from Central Afghanistan (Wardik (holotype) and Kabul (paratype)).Afghan specimens differ from those found in Central Europe in having light brown wing color.Wing pattern was considered as a major diagnostic character to distinguish between the two taxa (Daniel, 1953) (Figs. 1a,b).Recently, a morphological analysis showed that male genitalic structures of C. c. afghanistanus differ from those found in nominotypical C. c. cossus (Linnaeus, 1758) in having a smooth ridge on costal edge of valva, short processes of transtilla, and a slender phallus (Figs.2a, b) (Yakovlev, 2009).Based on wing coloration and male genitalic structures, C. c. afghanistanus was treated by Yakovlev (2009) at the species level.Cossus cossus afghanistanus is known from the following specimens from Central Afghanistan (Yakovlev et al., 2015): Cossus cossus has not been reported from Afghanistan, however this species is known from adjacent countries (large series of specimens are available from southern Kazakhstan, Kyrgyzstan, northwest China, Uzbekistan, and Tadzhikistan).Apparently, the material mentioned above can be assigned to subspecific taxon C. c. tianshanus Hua, Chou, Fang et Chen, 1990 (type locality: Manas, Xinjiang Uyghur Autonomous Region of northwestern China [90 km NWW Urumchi]).It should be noted, that analyses based primarily on external morphological characters fail always clarify the phylogenetic position and identities of many Lepidoptera taxa (Hajibabaei et al., 2006;Lukhtanov et al., 2008;Dincă et al., 2013;Shapoval & Lukhtanov, 2016).Cossus c. afghanistanus has never been studied genetically, therefore its identification, taxonomic status and phylogenetic position have remained unconfirmed.Our paper provides first molecular analysis of this taxon.

Material and methods
One 10-20 ng genomic DNA and 0.5 mM each of forward and reverse primer, 1 mM dNTPs, 10x PCR Buffer (0.01mM Tris-HCl, 0.05M KCl, 0.1% Triton X-100: pH 9.0), 1 unit Taq DNA Polymerase (Thermo Fisher Scientific, Lithuania), 5mM MgCl2.The temperature profile was as follows: initial denaturation at 94°C for 1 min, followed by 30 cycles of denaturation at 94°C for 45 s, annealing at 50°C for 45 s, and extension at 72°C for 1 min with a final extension at 72°C for 10 min.Amplified fragment was purified using GeneJET Gel Extraction Kit (Thermo Fisher Scientific, Lithuania).Purification was carried out according to the manufacturer's protocol.The success of PCR amplification purification was evaluated by electrophoresis of the product in 1% agarose gel.Purified PCR product was used for direct sequencing.Sequencing of the double-stranded product was carried out at the Research Resource Center for Molecular and Cell Technologies (St.Petersburg State University).
Representatives of nominotypical C. cossus were included in our analysis to clarify the taxonomic position and phylogenetic relationships of C. c. afghanistanus.We use obtained in the present study sequence of Dyspessa salicicola (Eversmann, 1848) from Azerbaijan (GenBank accession number MF596152) as an outgroup to root the phylogram.A complete list of specimens included in this study is given in Table 1.
A Bayesian approach and maximum-likelihood (ML) analyses were used for estimating the phylogeny.Bayesian analysis was performed using the software MrBayes 3.1.2with the nucleotide substitution model GTR+G+I.jModelTest was used to determine optimal substitution models for Bayesian inference (BI) analysis (Posada, 2008).TRACER, version 1.4 was used for summarizing the results of the Bayesian phylogenetic analysis (http://beast.bio.ed.ac.uk/Tracer).Maximum-likelihood (ML) analysis was performed using MEGA6 software (Tamura et al., 2013).We used non-parametric bootstrap values (Felsenstein, 1985) to estimate branch support on the reconstructed ML tree.Branch support was inferred from 1000 bootstrap replicates.

Results and discussion
Phylogenetic analysis of the gene COI resulted in a consensus phylogram which indicated a high level of posterior probability and bootstrap values for the clades (Fig. 3).Both (BI and ML) analyses demonstrated that all the Cossus individuals formed two discrete, statistically supported clades.The first clade includes specimen of C. c. afghanistanus.The second clade is formed by the representatives of nominotypical C. cossus.Moreover, analysis of a dataset of 29 specimens recognized C. c. afghanistanus as a highly differentiated and strongly supported lineage with a basal position.Cossus c. afghanistanus differs from nominotypical C. cossus by numerous nucleotide substitutions (p-distance is 6.7% ± 1.5%).The COI gene as a part of mitochondrial DNA (mtDNA) is a widely accepted and frequently used marker in molecular and phylogenetic studies.mtDNA has numerous advantages, including lack of recombination due to uniparental inheritance and multicopy status in most cells (Avise, 2000).Furthermore, due to relatively higher substitution rates in the mitochondrial genome, mtDNA is generally less conserved than many nuclear genes, providing higher resolution for lower level phylogenies and species identification through DNA barcoding (Hebert et al., 2003).Nevertheless, several authors have pointed out that the phylogenetic studies, or species identification based on mtDNA alone, can be misleading as trees inferred from single markers sometimes display relationships that reflect the evolutionary history of individual genes rather than the species being studied (Ballard & Whitlock, 2004;Bensch et al., 2006).Moreover, mitochondrial introgression (Zakharov et al., 2009) and Wolbachia infection (Ritter et al., 2013) also can lead to erroneous phylogenetic reconstructions and species misidentification.Despite these limitations, level of genetic distances can provide indirect evidence for conspecifity/non-conspecifity of species being studied.Two allopatric taxa can be considered as different species if the COI distance exceeds the "standard" 2.7-3.0%DNA-barcoding threshold (Lambert et al., 2005).Although this level is not an absolute threshold to distinguish between species, it was demonstrated that such a deep level (ca.3%) of differentiation between COI barcodes is practically always associated with species level of the taxa compared (Hebert et al., 2003).Thus, considerable level of genetic distances can be used as a useful criterion while deciding on the taxonomic status of a group under analysis and inferring hypotheses about species borders (Lukhtanov, 2015).Furthermore, the presence of significant genetic distances between two taxa accompanied by solid morphological differences (e.g. in wing pattern, and/or in genitalia structure) can be considered as an additional independent evidence of the existence of two distinct species.
Cossus c. afghanistanus and C. cossus appear on our reconstruction as a two strongly differentiated monophyletic groups, and the p-distance between their COI barcodes (6.7% ± 1.5%) significantly exceeds the 2.7-3.0%DNA-barcoding threshold.Additionally, C. c. afghanistanus and nominotypical C. cossus differ in characters of wing coloration and male genitalic structure.Thus, in accordance with the criteria mentioned above they should be considered as a separate species.
The financial support for this study was provided by the grant from the Russian Science Foundation no.14-14-00541 to the Zoological Institute of Russian Academy of Sciences.The work was partially performed using equipment of Centre for Molecular and Cell Technologies of St. Petersburg State University.The authors are grateful to Prof. Boris Kondratieff (Fort Colins) for language editing.
specimen of C. c. afghanistanus (RYB) was collected by I. Pljushtch, O. Pak and Ju.Skryluik on 09.VI.2016 in Kabul (Afghanistan).The specimen (GenBank accession number MF596151) was processed at the Department of Karyosystematics of the Zoological Institute of the Russian Academy of Sciences.DNA extraction from a single leg removed from the specimen was accomplished using the QIAamp DNA Investigator Kit (Qiagen, Netherlands) following the manufacturer's protocol.Standard lepidopteran barcode primers(Hebert et al. 2004) were used for DNA amplification and resulted in a 658 bp fragment of the mitochondrial cytochrome oxidase I gene (COI).The PCR amplification was performed in a 50 μl reaction volume containing ca.

Figure 3 .
Figure 3.The Bayesian tree of Cossus afghanistanus and Cossus cossus based on analysis of the cytochrome c oxidase subunit I gene from 28 specimens.Numbers at nodes indicate ML bootstrap/ Bayesian posterior probability respectively, with nonmatching clades using different analyses indicated by '-'.Scale bar = 0.3 substitutions per position.Cossus afghanistanus and Cossus cossus clusters highlighted in pink and blue respectively.

Table 1 .
List of studied material (29 specimens).(*) -sequence obtained in the present study.(**) -sequence obtained in the present study and used as an outgroup.