A molecular approach to the genus Alburnoides using COI sequences data set and the description of a new species, A. damghani, from the Damghan River system (the Dasht-e Kavir Basin, Iran) (Actinopterygii, Cyprinidae)

Abstract The molecular status of nine species of the genus Alburnoides from different river drainages in Iran and additionally by seven species from Europe was assessed. mtDNA COI gene sequences from freshly collected specimens and available NCBI data revealed four major phylogenetic lineages. Based on the results, a distinct taxon from the Cheshmeh Ali (Ali Spring), a Damghan River tributary in the endorheic Dasht-e Kavir basin, northern Iran, which is the closest sister to Alburnoides namaki (Namak Lake basin) + Alburnoides coadi (Nam River in the endorheic Dasht-e Kavir basin) is considered as a new species, Alburnoides damghani sp. n. It is distinguished from other Alburnoides species in Iran by a combination of character states including: a weakly-developed, variably-scaled, ventral keel from completely scaleless to completely scaled, a short snout with the tip of the mouth cleft on a level with the lower margin of the pupil or slightly lower, a small eye (eye horizontal diameter slightly to markedly less than interorbital width), commonly 8½ branched dorsal-fin rays, commonly 11−12½ branched anal-fin rays, 40−46(47) total lateral-line scales, 2.5–4.2 or 2.5–4.1 pharyngeal teeth, gill rakers short and widely spaced, 6−8 in total, 39−41 (commonly 40), total vertebrae, (19)20(21) abdominal vertebrae, 19−21 (most commonly 20) caudal vertebrae, abdominal vertebral region most commonly equal to or longer than caudal region, and most common vertebral formulae 20+20 and 21+19.


Introduction
The genus Alburnoides, a member of the family Cyprinidae, is found in Europe, Asia Minor and Central Asia with 28 species so far considered valid , Mousavi-Sabet et al. 2015a, b, Coad 2015. Alburnoides bipunctatus (Bloch, 1782) was the name applied to most populations throughout Europe and the Middle East from north of the Alps (France) eastwards to the Black, Caspian and Aral Sea basins but ongoing research has revealed a much greater diversity , Seifali et al. 2012, Mousavi-Sabet et al. 2015a.
Ali, a Damghan River tributary in the Dasht-e Kavir drainage, could not be identified with any of the named species and represents a new species. Hence, the aim of this study was to describe this new species based on a wide comparison with known Iranian species of the genus and investigate phylogenetic relationships among the major Alburnoides lineages by analyzing sequence variation of the mitochondrial COI gene.

Morphological examinations
After anesthesia, fishes were fixed in 5% formaldehyde and later stored in 70% ethanol. Counts and measurements follow Hubbs and Lagler (1958). Measurements were performed using digital calipers to the nearest 0.01 mm. Standard length (SL) was measured from the tip of the upper jaw to the end of the hypural complex, total length (TL) was measured from the tip of the upper jaw to the end of the longest caudal-fin lobe. Head length and interorbital width were measured to their bony margins. Fin ray counts separate unbranched and branched rays. The last two branched rays articulated on a last compound pterygiophore in the dorsal and anal fins and are noted as "1½". Mean and standard deviation were calculated without the "½". Lateral-line scale count includes pierced scales, from the first one just behind the supracleithrum to the posteriormost one at the base of the caudal-fin rays (i.e., posterior margin of the hypurals) excluding 1 or 2 scales located on the bases of the caudal-fin rays, total number of lateral-line scales is also provided. Counts of unpaired fin rays and vertebrae were done from radiographs. The character states of the ventral keel scale cover were estimated by direct measurements as shown in Bogutskaya et al. (2010). Statistical calculations and the multidimensional scaling (MDS) analysis were performed using software packages Statistica 6.0 and Primer v6.1.9.

Molecular data analysis
The haplotypes were compared to published Alburnoides sequences using (BLASTn) basic local alignment search tool (Altschul et al. 1990). All sequence data were aligned using MEGA6 software (Tamura et al. 2013). To unify the length of the sequences, the common 620 bp length segments were selected and used for phylogenetic analysis. Modeltest (Posada and Crandall 1998), implemented in the MEGA 6 software (Tamura et al. 2013), was used to determine the most appropriate sequence evolution model for the given data, treating gaps and missing data with the partial deletion option under 95% site coverage cut-off. We generated maximum likelihood phylogenetic trees with 10,000 bootstrap replicates in RaxML software 7.2.5 Stamatakis (2006) under the GTR+G+I model of nucleotide substitution, with CAT approximation of rate heterogeneity and fast bootstrap to explore species phylogenetic affinities. Bayesian analyses of nucleotide sequences were run with the parallel version of MrBayes 3.1.2 (Ronquist and Huelsenbeck 2003) on a Linux cluster with one processor assigned to each Markov chain under the most generalizing model (GTR+G+I) because overparametrization apparently does not negatively affect Bayesian analyses (Huelsenbeck and Ranala 2004). Each Bayesian analysis comprised two simultaneous runs of four Metropolis-coupled Markov-chains at the default temperature (0.2). Analyses were terminated after the chains converged significantly, as indicated by the average standard deviation of split frequencies <0.01.
Sequenced Screening for diagnostic nucleotide substitutions relative to Oryzias latipes was performed manually from the resulting sequence alignment. Estimates of evolutionary divergence over sequence pairs between species were conducted in Mega6 (Tamura et al. 2013). Analyses were conducted using the Kimura 2-parameter model (Kimura 1980). The rate variation among sites was modelled with a gamma distribution (shape parameter = 1). Codon positions included were 1st+2nd+3rd. All positions containing gaps and missing data were eliminated.
As appropriate outgroup to root the constructed phylogenetic hypothesis, Alburnus alburnus (accession number: KM373683), was included.

Results
COI barcodes were generated for a total of 36 Alburnoides specimens. Two phylogenetic approaches Bayesian Inference (BI) and Maximum Likelihood (ML), gave the same tree topologies and thus one is presented (Fig. 2). Tables 2-3 list the diagnostic nucleotide substitutions and estimates of the average evolutionary divergence found in the mtDNA COI barcode region. The two different phylogenetic approaches produced almost identical tree topologies although Bayesian analysis Yang 1986, Yang andRannala 1997) has been empirically demonstrated to be the most efficient character-based method for accurately reconstructing a phylogeny (Simmons and Miya 2004). Two methods produced trees with 4 major lineages supported by high posterior probability and bootstrap values and seven groups (Fig. 2 Diagnosis. Alburnoides damghani sp. n. is distinguished by having a combination of character states which includes a weakly-developed, variably-scaled, ventral keel from completely scaleless to completely scaled; a stout short snout with tip of the mouth cleft on a level with the lower margin of the pupil or lower; a small eye (eye horizontal diameter slightly to markedly less than interorbital width); commonly 8½ branched dorsal-fin rays; commonly 11−12½, branched anal-fin rays; 40−46(47) total lateral-line scales (40-46 scales to posterior margin of the hypurals); 2.5-4.2 and 2.5-4.1 pharyngeal teeth; 6−8 total gill rakers in outer row on first left arch; 39−41, commonly 40, total vertebrae; 12−14, commonly 13, predorsal vertebrae; abdominal vertebral region most commonly equal to or longer than caudal region (vertebral formulae 20+20 and 21+19).
Description. Description of holotype (Fig. 3). The caudal-fin lobes are rounded and the fin is shallowly forked. A ventral keel between the pelvics and the anal fin is scaleless for 1/3 of the length in front of the anus. There is a pelvic axillary scale and scales extend over the proximal bases of the anal fin forming a sheath. The upper body profile is convex, similar to the lower profile. The body is relatively thick and the caudal peduncle short and deep (its depth enters the length 1.7 times).
The eye is small, its horizontal diameter enters interorbital width 1.2 times. The snout is short and stout, its length only slightly exceeds the eye diameter. The upper jaw slightly projects over the lower jaw. The mouth is small, terminal, the mouth cleft is slightly curved, and the tip of the mouth cleft is on a level with the lower margin of the pupil. The posterior end of the lower jaw is on a vertical with the anterior margin of the pupil. The body depth enters SL 3.2 times, HL enters 3.7, predorsal length 1.8, caudal peduncle depth 7.7, caudal peduncle length 4.7, length of longest dorsal fin ray 4.4, and length of longest anal-fin ray to scale sheath 6.6. Eye horizontal diameter enters HL 3.9 times, snout length enters 3.4, and interorbital width 3.2. Pectoral-fin length enters pectoral-fin origin to pelvic-fin origin distance 1.2 times and pelvic-fin length enters pelvic-fin origin to anal-fin origin distance 1.1 times.
Dorsal-fin rays are 4 unbranched and 8½ branched, anal fin rays are 3 unbranched and 12½ branched, pectoral-fin branched rays are 13, and pelvic-fin branched rays are 7. The anal-fin origin is on a vertical from the posterior end of the dorsal-fin base.  Total lateral-line scales number 46 and those to posterior margin of hypurals 44, scales around caudal peduncle 17, scales above lateral line to dorsal fin origin are 9, scales below lateral line to anal-fin origin are 4, scales below lateral line to pelvic-fin origin are 4, and midline predorsal scales are 27. Pharyngeal teeth 2.5-4.2. Gill rakers number 6, they are short and stubby, the longest touching the adjacent one when appressed. Total vertebrae number 40 (abdominal vertebrae 20, caudal vertebrae 20). Predorsal vertebrae number 13.
The peritoneum is silvery with fine melanophores. The lateral line is clearly delineated by darker pigment above and below, but this is obscured on the caudal peduncle by the flank stripe. Some pigment on flank scales above the lateral line give the impression of stripes. A mid-flank stripe is evident, darkest on the caudal peduncle. The back and top of the head are dark, the belly is light with almost no melanophores. Melanophores are dense dorsally on the flank becoming progressively less ventrally. All fins have melanophores lining the rays, and the dorsal, anal and caudal fins have melanophores on the membranes, with very few melanophores on the pectoral-and pelvic-fin membranes. The unbranched pectoral-fin ray is lined with melanophores on its inner margin.
Description of paratypes. General appearance of body is shown in Figures 2−4 and morphometric data are given in Table 3. Body compressed but thick, upper body profile clearly convex, similar to the lower profile. The eye is small, always less than interorbital width (eye horizontal diameter enters interorbital width 1.1−1.4 times). Snout short and stout, only slightly pointed, snout length about equal to eye horizontal diameter. Mouth short, posterior end of upper jaw commonly in front of vertical with anterior margin of eye, posterior end of lower jaw on about vertical with anterior margin of pupil. Mouth terminal, but mouth cleft more or less markedly curved and tip of mouth cleft is on or below a level from lower margin of the pupil. Upper jaw slightly produced over lower jaw in most specimens, especially larger-sized. Ventral keel between pelvic and anal fin not sharp and weakly pronounced, variably scaled (examined in 24 paratypes): completely scaleless (in 7 specimens), scaleless along 3/4 (4 specimens), 2/3 (4 specimens), 1/2 (5 specimens), 1/4 (2 specimens), 1/5 (1 specimen) of keel length in front of the anus or completely scaled (1 specimen). Pelvic axillary scale present extending over the proximal base of the anal fin. Caudal fin shallowly forked with rounded lobes. Anal-fin origin at the vertical of the posterior end of the dorsal fin base (Fig. 5) or in front of it (Fig. 4). The dorsal-fin outer margin is truncate to slightly convex and the anal-fin outer margin is slightly concave. For measurement and ratios see Table 4.
Mature males bear tubercles on the unbranched and branched fin rays, in a single row branching into two distally on the branched rays. These are most prominent on the pectoral, pelvic and anal fins. Tubercles line scale margins in a single row of up to six tubercles, in particular over the anal fin and on the lower caudal peduncle. Scales below the dorsal fin are also lined with tubercles but to a much lesser extent than those above the anal fin. Flank scales generally may bear tubercles but many do not and anterior flank scales may have only a single tubercle. Minute tubercles are present on the dorsal and upper head surface.
Coloration of live specimen. Pigmentation consists of a darker back fading to a silvery white belly, three to four rows of large dark spots above lateral line starting from posterior part of operculum to posterior level of anal fin, continuing with two rows behind anal fin to base of caudal fin, small black spots on the operculum, behind and below the eye, smaller and less dark spots between the eye and upper jaw, a lateral line demarcated by pigment above and below it (the typical "stitched" pattern in many Alburnoides species), base of anal, pelvic, pectoral and dorsal fins almost reddish-orange, caudal-fin base pale or faint yellow. Posterior free margin of dorsal, anal, caudal and pelvic fins whitish hyaline, faint pigmentation on the caudal-fin centre branching distally to follow the inner margins of the fin fork, and fine pigmentation on the proximal part of dorsal-and anal-fin rays, darker in dorsal-fin rays (Figs 3, 4).
Habitat (Fig. 7). At the Cheshmeh Ali sampling site, the spring was about 5−10 m wide, with substrate consisting of coarse gravel and boulders, good riparian vegetation   Table 5. Mean values of some meristic characters of Alburnoides species from the Caspian Sea basin and adjacent endorheic basins, used for the DMS analysis.
Numbers of samples as in Figs 1 and 8 (own data except for two indicated samples). and almost fast-flowing and transparent waters. The physicochemical parameters at the spot were: dissolved oxygen, 7.54 mg/L, total dissolved solids, 318 mg/L, salinity, 0.32‰, conductivity, 552 μm/cm, pH: 7.97 and water temperature 23.25 °C.

Species
Comparative remarks. Alburnoides damghani sp. n., together with other Iranian species of the genus, belongs to the south-eastern group of species distributed in the eastern area of the genus distribution and characterised by commonly 4 pharyngeal teeth in the long row on the right 5th ceratobranchial ). As most distinguishing characters for species identification are counts (numbers of branched rays in the dorsal and anal fins, gill rakers, lateral-line scales and vertebral counts), a MDS statistical analysis was performed based on mean values of these counts (Table 5) to visualize the level of similarity of individual samples (species) in the Caspian Sea basin localities and adjacent endorheic basins. Frequences of occurrence of individual counts by characters can be found in earlier publications , Mousavi-Sabet et al. 2015a. The map plotting each sample in two-dimensional space is presented in Fig. 9; stress value is 0.04 (very low) meaning that the results are highly reliable (Davison 1983). The proximity of the examined samples to each other indicate how similar they are, and Alburnoides damghani sp. n. stands far apart from all other species, being relatively closer to A. namaki, A. varentsovi and Alburnoides sp. (Amu Darya River), morphologically.
When compared to Alburnoides species distributed in the Caspian Sea basin and adjacent endorheic basins in Iran, A. damghani sp. n. is clearly different from A. parhami from the Atrek River drainage by having four teeth in the long row on the 5 th ceratobranchial (vs. 5). By having five pharyngeal teeth in the long row on the 5 th ceratobranchial (this character state is invariably present in all examined specimens), A. parhami stands apart from all other species in Iran. Besides the number of teeth, A. damghani sp. n. is distinguished from A. parhami by having three unbranched dorsal fin rays (vs. often four, found in 13 from 35 examined specimens), commonly a partly scaleless ventral keel (vs. sharp and commonly scaleless), a terminal mouth with the tip of the mouth cleft on or below a level from lower margin of the pupil (vs. an upturned terminal mouth with the tip of the mouth cleft on a level with the upper half of the pupil), and 40−46 lateral-line scales to the margin of the hypurals (vs. 45−51).
Alburnoides damghani sp. n. differs from both A. petrubanarescui (which is the most morphologically peculiar species in the area possessing the lowest number of anal-fin branched rays) and A. namaki (a species phylogenetically close to A. damghani, see Fig. 2) by a slightly pointed snout (vs. markedly rounded), a terminal mouth with the tip of the mouth cleft on or below a level from lower margin of the pupil (vs. subterminal, with the tip of the mouth cleft on or below a level from lower margin of the eye), and 40−46 lateral-line scales (to the margin of the hypurals) (vs. 42−51, commonly 44−48). Alburnoides damghani sp. n. further differs from A. petrubanarescui by commonly 8½ branched dorsal-fin rays (vs. commonly 7½), commonly 11−12½ branched anal-fin rays (vs. commonly 9½), abdominal vertebrae commonly 20 (vs. commonly 21), and a ventral keel commonly partly or completely scaleless (vs. completely scaled). From A. namaki, A. damghani sp. n. can be further distinguished by a smooth and sometimes partly scaled ventral keel (vs. sharp and completely scaleless) and a higher number of predorsal vertebrae (modally 13 vs. modally 12).
Alburnoides coadi (Fig. 9) is the phylogenetically closest sister to A. namaki and the two species are rather similar in shape of the head, mouth and body; however, the two species are different by a complex of meristic characters (Fig. 9). Alburnoides damghani sp. n. differs from A. coadi, first of all, by a lower number of the lateral-line scales to the margin of the hypurals (40−46 vs. 47−51), a higher number of gill rakers (8−10, modally 8 and 9 vs. 6−8, modally 7), and a lower number of total vertebrae (modally 40 vs. modally 41).
Alburnoides damghani differs from A. holciki and A. qanati by a relatively smallsized eye with horizontal diameter slightly to markedly less than interorbital width (vs. large eye with eye diameter about equal to or larger than interorbital width), a tip of the mouth cleft on a level with or below the lower margin of the pupil (vs. on a level with the upper half to the upper margin of the pupil), and a shallowly forked caudal fin with rounded lobes (vs. clearly forked caudal fin with pointed lobes). Alburnoides damghani sp. n. is further distinguished from A. holciki from the Hari River drainage  in northeastern Iran by a usually smooth and often partly scaled ventral keel (vs. sharp and scaleless), a lower number of total lateral-line scales (44−47 vs. 47-57), a lower number of anal-fin rays (commonly 11−12½ vs. 13-16½), a lower number of total vertebrae (39−41, usually 40 vs. 40-42, usually 41), an abdominal vertebral region most commonly equal to or longer than caudal region, and most common vertebral formulae 20+20 and 21+19 (vs. abdominal region shorter than caudal region, and most common vertebral formulae 20+21, 20+22 and 19+21). Alburnoides damghani further differs from A. qanati (the Pulvar River drainage of Fars Province in southern Iran) by modally 12½ branched dorsal-fin rays (vs. modally 11½).
As can be seen from Fig. 8, Alburnoides sp. from rivers in the south of the Talysh Mountains and west of the Safid River (examined samples are mostly from estuarine areas of the rivers in Gilan Province), A. samiii from the type locality (Safid River drainage), and A. tabarestanensis from different localities (other than the type one) cannot be clearly discriminated by their meristic character states. Also, they are rather similar by the head and body shape, having most commonly a horizontal, slightly curved mouth, with a tip of the mouth cleft often on a level below the lower margin of the pupil, a slightly to markedly rounded snout, a variably but commonly well forked caudal fin. River drainage (Namak Lake basin) than to the other river systems of the Dasht-e Kavir basin (Freyhof et al. 2014) which is supported here. The validity of A. eichwaldii from the Kura River is supported by the COI barcode region. Alburnoides bipunctatus armeniensis Dadikyan, 1972 from Rivers Arpa, Vorotan, Vedi, Marmarik, Kasakh, and their tributaries (Aras River system, Kura River drainage) is a synonym of A. eichwaldii according to Bogutskaya and Coad (2009) being supported here by using COI barcode region of four fresh collected specimens from two localities in the Aras River (near the cities of Poldasht and Parsabad, border of Iran and Azerbaijan (Fig. 1). Recently, the phylogenetic relationships and taxonomy in the genus Alburnoides have been examined by comparative sequencing analyses of mitochondrial and nuclear markers by Stierandová et al. (2016). According to these authors, a molecular analysis revealed 17 Eurasian lineages divided into two main clades, termed the Ponto-Caspian and European in accordance with the lineage distribution. According to Stierandová et al. (2016) the European clade is represented by A. bipunctatus, A. rossicus, A. tzanevi, A. maculatus, A. ohridanus, A. strymonicus, 4 unnamed or undescribed species and populations defined as the Alburnoides prespensis complex including A. prespensis s. stricto, A. fangfangae and A. devolli. However, they concluded that phylogenetic analyses present ambiguous results and do not support recently accepted taxonomy which presumes validity of three species: A. prespensis, A. fangfangae, and A. devolli supporting our results, considering A. fangfangae and A. devolli being part of an A. prespensis complex (Fig. 2). Furthermore, Stierandová et al. (2016) considered A. eichwaldii, A. fasciatus, A. kubanicus, Safid River population (now A. samiii) and Talar population (now A. tabarestanensis) in the Ponto-Caspian clade. Base on the current study, IV lineage can be considered in the Ponto-Caspian clade and I and II lineages both in the European clade. Moreover, the placements of A. strymonicus and A. sp. Sperchios, which were uncertain in Stierandová et al. (2016) appear to be well-supported here. From a biogeographical viewpoint, the locations of lineage richness in most cases correspond to confirmed glacial refugia (Stierandová et al. 2016).
To conclude, the genetic analyses supported the validity of many morphologically distinguishable species of the genus Alburnoides in Iran (i.e., A. damghani sp. n., A. eichwaldii, A. holciki, A. namaki, A. qanati) belonging to a distinct phylogenetic lineage. Two species of Tigris river basin, A. idignensis and A. nicolausi are very closely related and are not well supported as sister taxa (low posterior probability of 0.62) by the COI barcode region, however, the ancestral node for A. idignensis is 1.0, as is the ancestral node for A. nicolausi, which is strong support for monophyly of each of these species. The analysis also demonstrated the existence of four major phylogenetic lineages within the genus Alburnoides in general.
Ghasemian, S. Mirghiasi, and B. Parsi for helping with fish collection, and the Environment Departments of Semnan, Fars, Markazi, Qom, and Ardabil provinces for their kind cooperation in visiting the collection sites. We are grateful to O.A. Diripasko (Institute of Fisheries and Marine Ecology, Ukraine) for his valuable assistance with the statistical analyses. The research work was funded by Shiraz University (approved by the Ethics Committee of the Biology Department, ECSU-909789), Tehran University, and the Canadian Museum of Nature. We also thank M. Geiger and J. Freyhof from the FREDIE project.