ISSN 0032-9452, Journal of Ichthyology, 2020, Vol. 60, No. 3, pp. 375–386. © Pleiades Publishing, Ltd., 2020. Taxonomic Status and Molecular Systematics of an Endemic Fish, Herklotsichthys lossei (Clupeidae) from the Persian Gulf: Insights into Non-monophyly of the Genus L. Purrafee Dizaja, H. R. Esmaeilia, *, T. Valinassabb, and A. Salarpouric a Ichthyology and Molecular Systematics Research Laboratory, Zoology Section, Department of Biology, College of Sciences, Shiraz University, Shiraz, Iran b Iranian Fisheries Science Research Institute, Agricultural Research, Education and Extension Organization, Tehran, Iran c Iranian Fisheries Science Research Institute, Agricultural Research, Education and Extension Organization, Persian Gulf and Oman Sea Ecological Research Center, Bandar Abbas, Iran *e-mail: hresmaeili@shirazu.ac.ir Received August 19, 2019; revised November 16, 2019; accepted December 5, 2019 Abstract—The endemic Persian Gulf herring, Herklotsichthys lossei, has similar characteristics with other congeners, especially H. punctatus, which makes it difficult to recognize. Freshly collected samples from the type locality of H. lossei assign morphologically to H. lossei, described by Wongratana in having a dark blotch on the dorsal fin and small dark spots on the flank. However, all the studied materials have a series of black spots along the back and both sides of dorsal–fin base which was not mentioned by Wongratana and was not found in the examined type materials deposited in Natural History Museum, London. Based on the available molecular data (mtCOI) for six species of Herklotsichthys, it is clear that the genus is not monophyletic. Herklotsichthys dispilonotus, H. punctatus and H. lossei make a sister lineage to other congeneric species including H. quadrimaculatus, H. lippa and H. spilurus and also with several phylogenetically distantly related genera Sardinella, Harengula, Amblygaster, Nematalosa, Anodontostoma, Alosa, Clupea and Clupeonella. This situation makes the taxonomic status of clupeid genera, especially Herklotsichthys more complicated. As H. dispilonotus the type species of the genus Herklotsichthys is nested with H. lossei and H. punctaus, these three species remain in the same genus and three others, H. quadrimaculatus, H. lippa and H. spilurus form a distinct genera. Keywords: DNA taxonomy, mtCOI sequences, Indo-Pacific region, endemism DOI: 10.1134/S0032945220030066 INTRODUCTION The genus Herklotsichthys Whitley, 1951, comprises 12 valid species (Fricke et al., 2019). It is characterized by having the gill opening with two fleshy outgrowths, upper jaw rounded when seen from the front, pelvic fin rays 8, back blue/green in live, fronto-parietal striae in top of head 3 to 8, lower portion of paddle– shaped second supramaxilla longer than upper, no dark spot at origin of dorsal fin, scale striae joined at center and no perforations on exposed portion of scale, no lateral line, and 29–52 gill rakers (Fischer and Bianchi, 1984). Two species of this genus have been reported from the Persian Gulf: Herklotsichthys lossei Wongratana, 1983 and H. quadrimaculatus (Rüppell, 1837) (Wongratana, 1983; Fischer and Bianchi, 1984; Fricke et al., 2019). Wongratana (1983) described H. lossei from the Persian Gulf, Bushire—Bushehr (Bushehr province) and Bender Shapur—Bandar Imam Khomeini (Khuzestan province), Northwest Persian Gulf, Iran, based on morphological characteristics of prominent ridges of teeth on palatines and pterygoids; intestine more or less straight, without a loop below stomach; no expansion of anteroventral edge of 3rd infraorbital; pyloric caeca 36 to 43; predorsal scales normal, median, not hidden; some dark spots along the flank and a dark blotch on dorsal fin. A closely related species is H. punctatus (Rüppell, 1837) of the Red Sea and Gulf of Suez showing several common morphological characters (Wongratana, 1983). But H. lossei differs from H. punctatus by having a dark blotch on upper part of dorsal fin, small black spots along flanks, and more pyloric caeca (36–43 vs. 29–34) (Wongratana, 1983). However, the validity of these characters has been under question in the recent years. Based on Randall (1995), Wongratana recognized lossei distinguishable on the basis of the blackish spot on the dorsal fin, supposedly lacking in punctatus (however, it is present, though faint); no blackish spots on the back as seen on punctatus (they are present on lossei). Randall 375 376 L. PURRAFEE DIZAJ et al. Table 1. Morphometric data of Herklotsichthys lossei from the Persian Gulf (n = 156) Value Character Standard length (SL) Head length (HL) Head depth/HL Head width/HL Upper jaw/HL Head depth/SL Eye diameter/HL Postorbital distance/HL Preorbital distance (snout length)/HL Interorbital distance/HL Caudal peduncle length/SL Length of caudal fin/SL Caudal depth/SL Body depth/SL Distance between base of caudal and anal/SL Distance between base of pectoral and ventral/SL Distance between base of anal and pectoral/SL Predorsal distance/SL Preanal distance/SL Prepelvic distance/SL Prepectoral distance/SL Body width/SL Maximum body depth/SL SD max min mean 70.23 19.59 91.72 54.99 56.04 28.70 40.23 45.19 36.04 27.42 15.17 33.22 19.31 33.07 25.21 32.92 62.43 51.84 83.69 59.94 31.58 14.25 33.07 51.99 14.14 78.42 40.84 38.42 20.13 28.65 32.46 22.35 19.02 3.43 23.93 7.29 24.19 15.54 22.00 48.39 40.56 73.21 46.22 23.77 9.31 24.19 58.91 16.56 85.17 45.76 45.74 23.96 33.03 37.94 25.78 22.26 8.16 28.68 13.30 28.18 19.83 28.28 54.29 45.38 79.21 54.39 27.06 11.98 28.18 3.51 1.09 3.66 2.59 3.19 1.50 2.42 3.03 2.29 1.88 2.56 2.04 2.72 1.85 1.99 2.37 3.28 2.17 2.88 2.60 1.75 1.00 1.85 SL—standard length, HL—head length, SD—standard deviation. (1995) expected further study may show that lossei is only subspecifically distinct from punctatus. Therefore, to clarify the validity of these characters, we (i) describe morphological characters of freshly collected samples from the type locality and other parts of the distribution range of H. lossei from the Persian Gulf, (ii) compare new material with the type materials and (iii) provide mtDNA COI gene sequences for this endemic clupeid fish to shed light on its phylogenetic relationship to other closely related species and other genera of clupeid fishes. MATERIALS AND METHODS Sampling and Preparation Collected fishes were either fixed in 5% formaldehyde, and stored in 70% ethanol, or directly fixed in 99% molecular grade ethanol. Measurements were made with a digital caliper and recorded to 0.1 mm. All measurements were made point to point, and never by projections. Methods for counts and measurements follow Randall and Di Battista (2012). Lengths given for specimens are standard length (SL), the straight–line distance from the anterior point of the upper lip to the base of the caudal fin (i.e., posterior end of the hypural plate). Head length (HL) is measured from the same anterior point to the posterior end of the opercular membrane, and snout length (preorbital distance) from the same point to the bony edge of the orbit. Body depth is the greatest depth from the base of the dorsal fin; body width is the greatest width. Orbit diameter is the greatest diameter, and interorbital width measured over the center of the eyes. Caudal–peduncle depth is the least depth, and caudal–peduncle length is measured horizontally from the rear base of the anal fin to the caudal–fin base. Predorsal, preanal, and prepelvic lengths are taken from the front of the upper lip to the origin of the respective fins. Counts and measurements, expressed as percentages of SL, are given in Table 1. Scale counts were possible on only a few specimens because the scales of species of Herklotsichthys are very deciduous. Most scales are missing from the majority of specimens and the scale pockets are not clearly JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 TAXONOMIC STATUS AND MOLECULAR SYSTEMATICS defined. Transverse scales are counted from the origin of the dorsal fin to the origin of the pelvic fin. The gill-raker counts include rudiments; the raker at the angle is contained in the lower limb count. The gill rakers were counted on all specimens, with no indication of an increase in number with growth. We received some photos of type materials of H. lossei from the London Natural History Museum (BMNH) for comparison with our samples. Examined Material All from Iran, Hormuzgan province, Qeshm Island, materials deposited in Zoological Museum Collection of Biology Department, Shiraz University, Iran (ZM–CBSU): ZM–CBSU X007–1 to X007– 100, 26°42′30.23′′ N, 55°58′22.75′′ E; Hormuzgan province, Bandar Lengeh: ZM–CBSU X060–1 to X060–40, 26°33′46.03′′ N, 54°53′19.24′′ E; Bushehr (=Bushihr) province, Asaluyeh area: ZM–CBSU X019–1 to X019–16, 27°28′2.63′′ N, 52°36′36.89′′ E. DNA Extraction and Amplification Total genomic DNA was extracted from muscle tissue using the Salt method (Bruford et al., 1992). For PCR, 1 µL of this genomic DNA was used in a 10 µL reaction with 0.5 U Bioline (“BioLine”, USA) Taq polymerase, 0.4 µL 50 mM MgC12, l µL l0× buffer, 0.5 µL of l0 mM dNTPs, and 0.3 µl of 10 µM of each primer FISH–BCL (5' TCAACYAATCAYAAAGATATYGGCAC 3') and FISH–BCH (5' TAAACTTCAGGGTGACCAAAAAATCA 3') (Baldwin, 2008). Thermal cycler program for PCR was one cycle of 5 min at 95°C; 35 cycles of 30 s at 95°C, 30 s at 52°C and 45 s at 72°C; one cycle of 5 min at 72°C, and hold at 10°C. Purification and sequencing of the PCR products were conducted in Germany. Molecular Data Analysis Vol. 60 size of the smallest fragment, resulting in a dataset of 653 base pairs (bp). No indications of unexpected stop codons or nuclear copies of mitochondrial fragments occurred in any sequence. New sequences are deposited in NCBI GenBank (www.ncbi.nlm.nih.gov) given with their respective accession numbers (Table 2). The most appropriate sequence evolution model for the given data was determined with Model test (Posada and Crandall, 1998) as implemented in jModelTest (Darriba et al., 2012). To explore species phylogenetic affinities, according to Modeltest, we generated maximum likelihood phylogenetic trees with 10000 bootstrap replicates in RaxML software 7.2.5 (Stamatakis, 2006) under the GTR + G + I model of nucleotide substitution, with fast bootstrap, and also Bayesian phylogenetic analysis (BI), using the Markov Chain Monte Carlo method (MCMC), with 6000000 generations under the most generalizing model (GTR + G + I) using MrBayes 3.1.1 (Ronquist and Huelsenbeck, 2003). Screening for diagnostic nucleotide substitutions was performed manually from the sequence alignment. Sequence divergence values between species were calculated using Kimura two Parameter (K2P) distance model implemented in MEGA 6.06 (Tamura et al., 2013). The reconstructed ML-based hypothesis of the mitochondrial relationships was used as input to infer putative species boundaries (molecular species delimitation approach) by using Poisson Tree Processes (PTP) and the refined multirate PTP (mPTP) models introduced by Zhang et al. (2013) and Kapli et al. (2017) respectively conducting in: http://mptp.h–its.org/#/tree. Version 09/2019. In both models, the aim is to find a group delimitation that maximizes the likelihood of the partition of branch lengths. In PTP model, a uniform evolutionary rate (lambda) is used while in the newer mPTP model, different rates for each group (species) are assumed (Kapli et al., 2017; Mousavi-Sabet et al., 2019; Sayyadzadeh et al., 2019). RESULTS We newly generated 10 COI gene sequences, 4 H. lossei, 3 Alosa braschnikowi, 1 Anodontostoma chacunda and 2 Sardinella albella and included already published data from NCBI GenBank for an additional 16 specimens belonging to 5 species of the genus Herklotsichthys including H. punctatus, H. dispilonotus, H. quadrimaculatus, H. lippa and H. spilurus and 35 other clupeid species (Table 2). We included Anchoa mitchilli (Linnaeus, 1758) (accession number: MH570230.1) as outgroup taxon (Plough et al., 2018). Data processing and sequence assembly were done with the software MEGA 6.06 (Tamura et al., 2013) and BioEdit (Hall, 1999). The Clustal W (Higgins and Sharp, 1988) was used to align the DNA barcodes after manually screening for unexpected indels or stop codons. Sequences of COI gene were trimmed to the JOURNAL OF ICHTHYOLOGY 377 No. 3 2020 Detailed Morphology of Herklotsichthys lossei Body compressed laterally, elongate, deepest at dorsal–fin origin. Body depth 24–33% of SL. Dorsal profile of head and body convex from snout tip to dorsal–fin origin and along dorsal–fin thereafter decreasing to uppermost point of caudal–fin base. Ventral profile of head and body slightly convex from lower jaw tip to pelvic–fin insertion, slightly convex or straight from pelvic–fin origin to anal–fin origin, and more or less straight along anal–fin base. Dorsal and ventral profiles of caudal peduncle almost straight. Belly rounded, covered by 26–33 sharp needle– like scutes anterior to insertion of anal fin (prepelvic and postpelvic scutes), no predorsal scutes. Anus situated just anterior to anal–fin origin. Caudal peduncle 378 L. PURRAFEE DIZAJ et al. Table 2. Accession number of sequences used in this study (https://www.ncbi.nlm.nih.gov/) Species name Alosa aestivalis (Mitchill, 1814) A. agone (Scopoli, 1786) A. alosa (Linnaeus, 1758) A. fallax (Lacepède, 1803) A. immaculata Bennett, 1835 A. sapidissima (Wilson, 1811) Anodontostoma chacunda (Hamilton, 1822) Amblygaster clupeoides Bleeker, 1849 Anchoa mitchilli (Valenciennes, 1848) Clupea harengus Linnaeus, 1758 Clupeonella cultriventris (Nordmann, 1840) Dussumieria acuta Valenciennes, 1847 D. elopsoides Bleeker, 1849 Escualosa thoracata (Valenciennes, 1847) Harengula clupeola (Cuvier, 1829) H. jaguana Poey, 1865 Herklotsichthys dispilonotus (Bleeker, 1852) H. lippa (Whitley, 1931) H. lossei Wongratana, 1983 H. quadrimaculatus (Rüppell, 1837) H. punctatus (Rüppell, 1837) H. spilurus (Guichenot, 1863) Nematalosa erebi (Günther, 1868) N. flyensis Wongratana, 1983 N. japonica Regan, 1917 N. nasus (Bloch, 1795) Sardinella albella (Valenciennes, 1847) S. aurita Valenciennes, 1847 S. fimbriata (Valenciennes, 1847) S. gibbosa (Bleeker, 1849) S. hualiensis (Chu and Tsai, 1958) S. janeiro? S. jussieu (Lacepède, 1803) S. lemuru Bleeker, 1853 Accession number KX459323.1 KJ552733.1 KC500192.1 KM286 449.1 KJ552756.1 KJ552592.1 KX459321.1 AP011614.1 EF607313.1 MH570230.1 KJ205365.1 KJ552938.1 EU014223.1 KC500 632.1 EF607377.1 GU224871.1 JQ842515.1 KX223910.1 HQ956377.1 KU170 602.1 KU508431.1 JF493637.1 JF493638.1 JF493639.1 JF493640.1 KF489612.1 KM538357.1 KM538358.1 KM538359.1 KM538360.1 KM538361.1 KR861530.1 JQ350053.1 KJ669557.1 KX274200.1 KU942898.1 JX983399.1 KJ566766.1 KM538516.1 HQ231356.1 JF494 406.1 KU942876.1 AM911174.1 HQ231360.1 HQ231367.1 JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 TAXONOMIC STATUS AND MOLECULAR SYSTEMATICS 379 Table 2. (Contd.) Species name Accession number S. longiceps Valenciennes, 1847 S. maderensis (Lowe, 1838) S. marquesensis Berry and Whitehead, 1968 KX988263.1 KY176607.1 MG816723.1 KJ968221.1 KX223945.1 KU942889.1 MG835 699.1 JX260883.1 *MT371660 * MT371661 *MT371662 *MT371665 *MT371666 *MT371667 *MT371668 *MT371669 *MT371663 *MT371664 S. melanura (Cuvier, 1829) S. sindensis (Day, 1878) S. zunasi (Bleeker, 1854) Tenualosa ilisha (Hamilton, 1822) Alosa braschnikowi (Borodin, 1904) Anodontostoma chacunda (Hamilton, 1822) Herklotsichthys lossei Wongratana, 1983 S. albella (Valenciennes, 1847) * Present study. compressed; depth about more than orbit diameter. Head compressed; head length 28% of SL. Snout tip rounded; snout length more than eye diameter. Upper jaw longer than the snout. Interorbital width less than eye diameter. Mouth small, terminal, dorsal to body axis. Two supramaxilla present; second asymmetrical; prominent ridge of teeth on palatine. Margin of mouth black. Eye large, round, covered with adipose eyelid, positioned laterally on head dorsal to horizontal through pectoral–fin insertion, visible in dorsal view, pupil round. Interorbital space flat. Nostrils are not close to each other, positioned anterior to orbit. Posterior margins of preopercle and opercle smooth. Subopercle is short with rounded posterior margin. Opercular membrane without serrations. Dorsal–fin origin behind to vertical through base of last pectoral fin ray, more close to head, predorsal length 45.38% in SL. Dorsal fin with 3 unbranched and 13–15 soft branched rays. First dorsal fin ray minute. Posterior tip of depressed dorsal fin not reaching to vertical through anal fin origin. Pelvic fin shorter than pectoral fin with 1 unbranched and 7 branched soft rays. Pelvic–fin axillary scale present. Pelvic–fin insertion posterior to vertical through dorsal–fin origin. Posterior tip of pelvic fin far from anus. Pectoral fin with 14–15 branched soft rays. Uppermost pectoral fin ray inserted below midline of body. Posterior tip of pectoral fin not reaching vertical through pelvic– fin origin. First anal fin ray minute. Two posteriormost anal fin ray more branched distally. Anal fin not JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 noticeably enlarged. Anal–fin origin well posterior to vertical through base of posteriomost dorsal fin ray. Caudal find forked. Posterior tips of caudal–fin lobes pointed. Gill rakers long, slender, rough, visible from side of head when mouth opened. Branchiostegal rays 6. Lateral line absent. Scales cycloid, thin and deciduous. Vertical striae on scales continuous across center of scale (Fig. 3). No scales on head. Dorsal and anal– fin bases with scaly sheaths. Flank silvery with 1 or 2 yellow spots behind gill opening and top of head in fresh specimens. A series of dark spots along flank and back, at both sides of dorsal fin (Fig. 1c). Dorsal fin with prominent black marking. Margin of caudal fin dusky. Color in Alcohol. Back blue/green in live but dark in fixed specimens with some dark spots. Flank silvery also with some dark spots. The margin of caudal fin dusky. The tip of snout is black. Distribution. Currently known only from the Persian Gulf. Remarks. As described by Wongratana (1983), this small herring–like species has a compressed body (body depth 26–28% SL; 24–33% SL in our study), vs. 18–30 for H. quadrimaculatus, 28–35 for H. spilurus and 24–30 for H. punctatus. Belly with 29–30 scutes (vs. 29–32 for H. quadrimaculatus and 25–33 for H. punctatus); number of lower gill rakers 31–34 (30–33 here) vs. 30–36 in H. quadrimaculatus and 30–37 in H. punctatus; number of pyloric caeca 36–43 380 L. PURRAFEE DIZAJ et al. (a) (b) (c) Fig. 1. Persian Gulf herring Herklotsichthys lossei (SL 66.92 mm) collected in the Persian Gulf: (a, b) lateral and (c) dorsal view. (36–42 here) (vs. 27–35 in H. spilurus, and 29–34 in H. punctatus). Freshly collected samples from the type locality of H. lossei assign morphologically to H. lossei, described by Wongratana (1983), in having a dark blotch on the dorsal fin and small dark spots on the flank. However, all the studied materials have black spots either along the back and the both sides of dorsal fin (Fig. 1) which was not mentioned by Wongratana (1983) and was not found in the examined type materials, holotype and paratypes (Fig. 2). H. lossei is closely related to H. punctatus (Rüppell, 1837) of the Red Sea, distinguished from it by presence of a blackish spot on the distal part of dorsal fin (vs. absent in some H. punctatus), however, it is present, though faint based on Randall (1995), presence of blackish spots on the back (vs. present but as irregular and bigger spots in H. punctatus), presence of a row of dark spots on flank (vs. absence in H. punctatus), lower gill rakers counts of 31–34 by Wongratana (1983) and also Randall (1995) and 30–33 here (vs. 32–36), and a higher number of pyloric caeca 36–43 (36–42 here) (vs. 29–34 in H. punctatus). According to Randall (1995), H. lossei might be subspecifically distinct from H. punctatus. Losse (1968) described two forms of H. punctatus: form A with a prominent black patch on dorsal fin and form B with dorsal fin dusky, without black patch. According to Losse (1968) they would certainly appear to be two distinct species. Herklotsichthys lossei is distinguished from H. quadrimaculatus, another congeneric species of the Persian Gulf and Oman Sea by presence of a blackish spot on the distal part of dorsal fin (vs. absent in H. quadrimaculatus), presence of a row of dark spots on upper sides (vs. absence in H. quadrimaculatus) and lower gill rakers counts of 31–34 by Wongratana (1983) and also Randall (1995) and 30–33 here (vs. 30–36 in H. quadrimaculatus). H. lossei is further distinguished from H. quadrimaculatus by lacking the JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 TAXONOMIC STATUS AND MOLECULAR SYSTEMATICS 381 (a) (b) (c) (d) Fig. 2. Lateral and dorsal view of holotype (a, b) and paratype (c, d) of Herklotsichthys lossei preserved in the Natural History Museum of London (BMNH). wing–shaped predorsal scales hidden beneath the normal ones, having prominent ridges of teeth on palatines and pterygoids, no intestinal loop, less developed denticulations on hind edge of scales (Wongratana, 1983). H. lossei is distinguished from H. spilurus, by having a series of dark spots along the flanks (vs. absence in H. spilurus), presence of a row of dark spots on upper JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 sides (vs. absence in H. spilurus), and a higher number of pyloric caeca 36–43 (36–42 here) (vs. 27–35 in H. spilurus). Molecular systematics. Both the ML and BI phylogenetic trees were mostly similar in their topology, hence, here only the BI tree including the posterior probability values from the Maximum Likelihood phylogram is presented (Fig. 4). Table 3 contains the 382 L. PURRAFEE DIZAJ et al. not monophyletic as reported by others (e.g., Li and Ortí, 2007; Lavoue et al., 2014). The PTP model–based species delimitation approaches using the ML topology delivered two different estimates for the total species number present in the data: PTP detected 33 entities (p = 0.001, Null– model score: 230.563482, best score for single coalescent rate: 146.622980) and mPTP9 entities representing 21 putative species (Null–model score: 230.472620, equal to the best score for multi coalescent rate). Our results showed a K2P nearest neighbor distance of 10% between H. lossei and H. punctatus (Table 3). DISCUSSION Fig. 3. Key scale of Herklotsichthys lossei from 3th row beneath dorsal fin. average estimates of the evolutionary distance in the 653 base pairs long mtDNA COI barcode region and Table 4 lists the diagnostic nucleotide substitutions. Based on the results, the genus Herklotsichthys is not monophyletic. H. lossei and H. punctatus make a sister group to H. dispilonotus while three other species, H. quadrimaculatus, H.lippa and H. spilurus are not nested within other members of the genus, and are sister to genus Sardinella. Moreover, family Clupeidae is Table 3. Average estimates of the evolutionary distance in the 653 base pairs long mtDNA COI barcode region Species Distance, % 1 H. lippa H. lippa H. spilurus H. lippa H. spilurus H. dispilonotus H. lippa H. spilurus H. dispilonotus H. quadrimaculatus H. lippa H. spilurus H. dispilonotus H. quadrimaculatus H. punctatus 2 H. spilurus H. dispilonotus H. dispilonotus H. quadrimaculatus H. quadrimaculatus H. quadrimaculatus H. punctatus H. punctatus H. punctatus H. punctatus H. lossei H. lossei H. lossei H. lossei H. lossei 20 20 20 20 0 20 20 30 20 30 20 20 20 20 10 The family Clupeidae, the most diverse group of clupeiformes (Fricke et al., 2019), is poorly understood systematically (Lavoue, 2017). Uncertainty in the taxonomic status of taxa is highly problematic makes it difficult to investigate the phylogeny of subfamilies, genera and species. The genus Herklotsichthys belongs to this taxonomically complicated group (Whitehead, 1965). The type species is Harengula dispilonotus Bleeker, 1852; being a replacement name, for Herklotsella Fowler, 1934; which was preoccupied by Herklotsella Herre, 1933 and hence was transferred to Herklotsichthys Whitley, 1951. According to Whitehead (1963), Herklotsichthys is morphologically closely related to the genera Sardinella and Harengula. Based on the available molecular data (mtCOI) for six species of Herklotsichthys, it is clear that the genus is not monophyletic. The Western Pacific (Malaysia east to Philippines, north to Vietnam) species Herklotsichthys dispilonotus, the Red Sea species H. punctatus and the Persian Gulf species H. lossei make a sister lineage to another species group including the widely distributed Indo-West Pacific H. quadrimaculatus and the East Africa, Madagascar, Reunion species H. spilurus, however, H. lippa replaced between Sardinella species not near to any of congeneric species. This situation makes the taxonomic status of Herklotsichthys more complicated. As the type species of the genus Herklotsichthys, H. dispilonotus, is nested with H. lossei and H. punctaus, hence these 3 species remain in the same genus and three others, H. quadrimaculatus, H. lippa and H. spilurus should be transferred to another genus or genera. Based on both morphological and molecular data, H. lossei is closely related to H. punctaus confirming the previous studies (Whitehead, 1963; Randall, 1995). It seems that ecologically replacement has occurred due to endemism and an endemic clupeid fish of the Persian Gulf H. lossei has been replaced by the Red Sea species H. punctaus. Previous studies revealed that regional endemism is important at the species and genus levels in the clupeoid fishes (Lavou et al., 2013). Herklotsichthys lippa from Western CenJOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 TAXONOMIC STATUS AND MOLECULAR SYSTEMATICS 100/100 91.2/33 100/65 86.1/33 100/100 383 Escualosa thoracata EF607377.1 Alosa aestivalis KX459323.1 Alosa agotie KJ552733.1 Alosa fallax KM286449.1 100/99 Alosa alosa KC500192.1 Alosa immaculata KJ552592.1 Alosa braschnikowi MT371660 Alosa braschnikowi MT371662 Alosa braschnikowi MT371661 Alosa sapidissima KX459321.1 Clupeonella cultriventris KJ552938.1 Clupea harengus KJ205365.1 Herklotsichthys quadrimaculatus JF493638.1 Herklotsichthys quadrimaculatus JF493639.1 Herklotsichthys quadrimaculatus JF493640.1 Herklotsichthys quadrimaculatus KF489612.1 Herklotsichthys quadrimaculatus JF493637.1 100/100 Herklotsichthys spilurus JQ350053.1 Sardinella marquesensis MG816723.1 100/100 Anodontostoma chacunda AP011614.1 Anodontostoma chacunda MT371665 Nematalosa erebi KJ669557.1 Nematalosaflyensis KX274200.1 100/98 Nematalosa nasus JX983399.1 Nematalosa japonica KU942898.1 Amblygaster clupeoides KU942898.1 Herklotsichthys lippa HQ956377.1 Sardinella jussieu HQ231360.1 99.9/95 Sardinella sindensis KU942889.1 Sardinella gibbosa JF494406.1 Sardinella maderensis KY176607.1 Sardinella marquesensis KJ968221.1 Tenualosa ilisha JX260883.1 Harengula clupeola GU224871.1 100/100 55.73/36 Harengula jaguana JQ842515.1 Sardinella hualiensis KU942876.1 Sardinella albella MT371663 Sardinella albella MT371664 100/100 Sardinella albella KJ566766.1 Sardinella fimbriata HQ231356.1 Sardinella melanura KX223945.1 Sardinella zunasi MG835699.1 Sardinella aurita KM538516.1 100/100 Sardinella janeiro AM911174.1 Sardinella lemuru HQ231367.1 Sardinella longiceps KX988263.1 Herklotsichthys dispilonotus КХ223910.1 Herklotsichthys punctatus KM538358.1 Herklotsichthys punctatus KM538359.1 100/99 Herklotsichthys punctatus KR861530.1 Herklotsichthys punctatus KM538361.1 Herklotsichthys punctatus KM538360.1 100/99 Herklotsichthys punctatus KM538357.1 Herklotsichthys lossei MT371666 Herklotsichthys lossei MT371667 100/100 Herklotsichthys lossei MT371668 Herklotsichthys lossei MT371669 Herklotsichthys lossei KU170602.1 Herklotsichthys lossei KU508431.1 Dussumieria acuta EU014223.1 Dussumieria elopsoides KC500632.1 Anchoa_mitchilli MH570230.1 0.1 mPTP PTP Fig. 4. Maximum Likelihood and Bayesian phylogeny reconstructed based on 653 bp of COI 5' end. The values beside the branches before and after a slash are BI posterior probability and ML bootstrap values, respectively. Solid bars right to the specimen labels indicate species delimitation results from mPTP followed by the results of the PTP approach as dashed bars. tral Pacific is morphologically similar to H. punctatus and H. lossei in having the series of small round black spots on the flank but it has a dark spot behind gill cover and elongate wing–like scales underneath normal paired pre-dorsal scales that the last one has made it similar to H. quadrimaculatus. Herklotsichthys dispilonotus is separate from other congeners by the presence of two dark saddle–like blotches on the back, at the hind part of the dorsal fin base and a short distance behind this. This species overlaps range of H. quadrimaculatus, which lacks the black saddles and has elongate wing–like scales underneath the normal paired pre-dorsal scales. Unresolved taxonomic positions have been also reported for other genera (e.g. Lavoué et al., 2007; JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 2013; Li and Ortí, 2007; Bloom and Lovejoy, 2014). The molecular systematic position of two closely related species H. lossei and H. punctatus to other members of the genus and with Sardinella complex group is interesting (Fig. 4) showing that Clupeidae is not monophyletic. In the inferred phylogenetic reconstructions (Lavoué et al., 2017), the Clupeidae and the Dussumieriidae were never recovered as monophyletic. Position of Sardinella on the COI tree topology presented here, which has a broad tropical distribution, including the Indo–West Pacific region, and the East and West Atlantic regions supports Lavoué et al. (2017) idea about unresolved phylogeny of clupeids. Here, Sardinella aurita (the type species of Sardinella), is nested with S. lemuru, S. longiceps and 384 L. PURRAFEE DIZAJ et al. Table 4. Diagnostic nucleotide substitutions found in the COI barcode region of studied Herklotsichthys species Nucleotide position relative to Oryzias latipes complete mitochondrial genome (AP004421.1) Species H. lippa H. dispilonotus H. punctatus H. quadrimaculatus H. spilurus H. lossei H. lippa H. dispilonotus H. punctatus H. quadrimaculatus H. spilurus H. lossei H. lippa H. dispilonotus H. punctatus H. quadrimaculatus H. spilurus H. lossei H. lippa H. dispilonotus H. punctatus H. quadrimaculatus H. spilurus H. lossei H. lippa H. dispilonotus H. punctatus H. quadrimaculatus H. spilurus H. lossei 5657 5663 5669 5681 5684 5687 5690 5699 5702 5708 5711 5714 T C A T T G 5723 G T T A A T 5801 A A G A A A 5897 T G A C C A 5969 A G G G G G T C C C C C 5729 T C C C C C 5805 A A A A G A 5900 A A A A A G 5987 A T A A A A C C T C C C 5738 G A A A T A 5825 A C T T T T 5903 T C A C C A 5990 C T C C C C G A C A A T 5741 G C T G G T 5837 T C T T T T 5915 T A G A A A 5993 G A T G G C C T C C C C 5744 C A C C C T 5840 C T C C C C 5924 T T C T T T 5999 C C C C C T C T T T T T 5768 C T A A A A 5852 C C T C C C 5930 T G C C C C A T G T T G 5771 A T T T T T 5852 T C C C C C 5933 T C G A A A A A G A A A 5777 T A A G G A 5864 A T C T T C 5936 C A A A A A T A C C C C 5789 C A T T T T 5867 A A G A A A 5939 A G C C C C C A C G G C 5792 C A G C C G 5873 A T C A A C 5963 C C C C C A A G A C C A 5795 C T C C C C 5885 A A C G G C 5966 G G A G G G C A A T T A 5796 C T T T T T 5888 T A C T T C 5967 C C C C C T S. janeiro and the other members of this genus are nested within different subfamilies. at least 13 endemic fish elements (Owfi, 2015) including H. lossei. Due to the high–latitude geographical position, the relative shallowness, the high evaporation rates, temperature fluctuation between winter and summer seasons (15–36°C), salinity excessing 43 psu (70–80 psu in tidal pools and lagoons) (Price et al., 1993), specific water current pattern and its geomorphological history (Ghanbarifardi et al., 2018), the Persian Gulf supports Our results show that the genus Herklotsichthys presents a new taxonomic challenge in clupeid studies. Comparison with other genera in different subfamilies and families of clupeiformes using several combined molecular markers, associated with more morphological and anatomical evidence, will be an effective tool in delimiting genera. Adding such a taxonomically JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 TAXONOMIC STATUS AND MOLECULAR SYSTEMATICS important endemic taxon as H. lossei could help better elucidate the phylogenetic tree of the Clupeiformes which is currently only partially resolved (Li and Ortí, 2007; Lavoué et al., 2013, 2017; Bloom and Lovejoy, 2014) and will increase our knowledge about the biogeography and evolution of these fishes with a complex pattern of worldwide distribution. ACKNOWLEDGMENTS We thank Jame Maclaine, Senior Curator of Natural History Museum of London and Kevin Webb, photographer, for providing holotype and paratype photos of H. lossei, Reza Sadeghi (Shiraz University) and Vahid Mazarei (an engineer from MAPNA company), for helping in fish collecting. We are thankful to B.W. Coad (Canadian Museum of Nature) for editing the manuscript. FUNDING The research was funded by Shiraz University and was approved by the Ethics Committee of the Biology Department (SU–9430246). COMPLIANCE WITH ETHICAL STANDARDS Conflict of interests. The authors declare that they have no conflict of interest. Statement on the welfare of animals. All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. REFERENCES Baldwin, W., Mounts, J., Smith, D., and Weigt, L., Genetic identification and color descriptions of early life–history stages of Belizean Phaeoptyx and Astrapogon (Teleostei: Apogonidae) with Comments on identification of adult Phaeoptyx, Zootaxa, 2008, pp. 1–22. https://doi.org/10.11646/zootaxa.3133.1.1 Bloom, D. and Lovejoy, N., The evolutionary origins of diadromy inferred from a time-calibrated phylogeny for Clupeiformes (herring and allies), Proc. Biol. Sci., 2014, vol. 281. https://doi.org/10.1098/rspb.2013.2081 Bruford, M., Hanotte, O., Brookfield, J., and Bruke, T., Single-locus and multilocus DNA fingerprinting, in Molecular Genetic Analysis of Population, Hoelzel, A.R., Ed., Oxford: Oxford Univ. Press, 1992, pp. 225–269. Fischer, W. and Bianchi, G., FAO Species Identification Sheets for Fishery Purposes Western Indian Ocean (Fishing Area 51), Rome: FAO Fish. Dep., 1984. http://www.fao.org/3/ad468e/ad468e00.htm. Fricke, R., Eschmeyer, W.N., and Fong, J., Catalog of fishes: species by family/subfamily, version 05/2019, 2019. http://archive.calacademy.org/research/ichthyology/catalog/Species. Hall, T., BioEdit: a user–friendly biological sequence alignment editor and analysis program for Windows JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020 385 95/98/NT, Nucleic Acids Symp. Ser., 1999, vol. 41, pp. 95– 98. Higgins, D. and Sharp, P., CLUSTAL: a package for performing multiple sequence alignment on a microcomputer, Gene, 1988, vol. 73, pp. 237–244. https://doi.org/10.1016/0378-1119(88)90330-7 Kapli, P., Lutteropp, S, Zhang, J., Kobert, K., Pavlidis, P., Stamatakis, A., and Flouri, T., Multi-rate Poisson tree processes for single–locus species delimitation under maximum likelihood and Markov chain Monte Carlo, Bioinformatics, 2017, vol. 33, no 11, pp. 1630–1638. https://doi.org/10.1093/bioinformatics/btx025 Keskin, E. and Atar, H., DNA barcoding commercially important fish species of Turkey, Mol. Ecol. Resour., 2013, vol. 13, no 5, pp. 788–797. https://doi.org/10.1111/1755–0998.12120 Lavoue, S., Miya, M., Musikasinthorn, P., Chen, W., and Nishida, M., Mitogenomic evidence for an Indo-West Pacific origin of the Clupeoidei (Teleostei: Clupeiformes), PLoS One, 2013, vol. 8, no 2, art. ID e56485. https://doi.org/10.1371/journal.pone.0056485 Lavoue, S., Konstantinidis, P., and Chen, W., Progress in Clupeiform Systematics, in Biology and Ecology of Sardines and Anchovies, Ganias, K., Ed., Boca Raton, FL: CRC Press, 2014. https://doi.org/10.1201/b16682-3 Lavoue, S., Bertranda, J., Chena, W., Motomurae, H., Sadof, T., and Miya, M., Phylogenetic position of the rainbow sardine Dussumieria (Dussumieriidae) and its bearing on the early evolution of the Clupeoidei, Gene, 2017, vol. 623, pp. 41–47. https://doi.org/10.1016/j.gene.2017.04.032 Li, C. and Ortí, G., Molecular phylogeny of Clupeiformes (Actinopterygii) inferred from nuclear and mitochondrial DNA sequences, Mol. Phylogenet. Evol., 2007, vol. 44, pp. 386–398. https://doi.org/10.1016/j.ympev.2006.10.030 Losse, F., The Elopoid and Clupeoid Fishes of East Africa Coastal Waters, Rome: UN Food Agric. Org., 1968, pp. 77– 115. Mousavi-Sabet, H., Vatandoust, S., Geiger, M., and Freyhof, J., Paracobitis abrishamchiani, a new crested loach from the southern Caspian Sea basin (Teleostei: Nemacheilidae), Zootaxa, 2019, vol. 4545, no 3, pp. 375–388. https://doi.org/10.11646/zootaxa.4545.3.3 Nelson, J., Grande, T., and Wilson, M., Fishes of the World, New York: Wiley, 2016. Owfi, F. A review on systematic and taxonomic of the Persian Gulf fish species based on geographical distribution pattern and habitat diversity, using by GIS, PhD Thesis, Tehran: Islamic Azad Univ., 2015. Plough, L., Ogburn, M., Fitzgerald, C., Geranio, R., Marafino, G., and Richie, K. Environmental DNA analysis of river herring in Chesapeake Bay: a powerful tool for monitoring threatened keystone species, PLoS One, 2018, vol. 13, no 11, art. ID e0205578. https://doi.org/10.1371/journal.pone.0205578 Posada, D. and Crandall, K., Modeltest: testing the model of DNA substitution. Bioinformatics, 1998, vol. 14, no 9, pp. 817–818. https://doi.org/10.1093/bioinformatics/14.9.817 386 L. PURRAFEE DIZAJ et al. Price, A., Sheppard, C., and Roberts, C., The gulf: its biological setting, Mar. Pollut. Bull., 1993, vol. 27, pp. 9–15. https://doi.org/10.1016/0025–326x(93)90004–4 Randall, J., Coastal Fishes of Oman, Honolulu: Univ. of Hawaii Press, 1995. Randall, J. and Di Battista, J., Etrumeus makiawa, a new species of round herring (Clupeidae: Dussumierinae) from the Hawaiian Islands, Pac. Sci., 2012, vol. 66., pp. 97–111. https://doi.org/10.2984/66.1.6 Ronquist, F. and Huelsenbeck, J.P., MrBayes 3: Bayesian phylogenetic inference under mixed models, Bioinformatics, 2003, vol. 19, pp. 1572–1574. https://doi.org/10.1093/bioinformatics/btg180 Sayyadzadeh, G., Teimori, A., and Esmaeili H.R., Paraschistura kermanensis, a new stone loach species from southeastern Iran (Teleostei: Nemacheilidae), Zootaxa, 2019, vol. 18, no 4638, no. 4, pp. 571–83. https://doi.org/10.11646/zootaxa.4638.4.7 Stamatakis, A., RAxML-VI-HPC: maximum likelihoodbased phylogenetic analyses with thousands of taxa and mixed models, Bioinformatics, 2006, vol. 22, no. 21, pp. 2688–2690. https://doi.org/10.1093/bioinformatics/btl446 Tamura, K., Stecher, G., Peterson, D., Filipski, A., and Kumar, S., MEGA6: Molecular Evolutionary Genetics Analyses version 6.0, Mol. Biol. Evol., 2013, vol. 30, pp. 2725–2729. https://doi.org/10.1093/molbev/mst197 Whitehead, P., A contribution to the classification of clupeoid fishes, Ann. Mag. Nat. Hist., Ser. 12, 1963, vol. 5, pp. 737–750. Whitehead, P., A review of the elopid and clupeid fishes of the Red Sea and adjacent regions, Bull. Br. Mus. (Nat. Hist.), Zool., 1965, vol. 12, no 7, pp. 225–281. Wongratana, T., Diagnoses of 24 new species and proposal of a new name for a species of Indo-Pacific Clupeoid fishes, Jpn. J. Ichthyol., 1983, vol. 29, no 4, pp. 385–407. Zhang, J., Kapli, P., Pavlidis, P., and Stamatakis, A., A general species delimitation method with applications to phylogenetic placements, Bioinformatics, 2013, vol. 29, no. 22, pp. 2869–2876. https://doi.org/10.1093/bioinformatics/btt499 JOURNAL OF ICHTHYOLOGY Vol. 60 No. 3 2020