Barbus rajanorum Luciobarbus mystaceus Barbus kersin Luciobarbus esocinus Luciobarbus longiceps (Valenciennes, 1842)Luciobarbus schejch

Freyhof, Jörg, Segherloo, Iraj Hashemzadeh, Vatandoust, Saber, Abdollahi-Mousavi, Seyed Eisa, Normandeau, Eric, Geiger, Matthias F. & Yoğurtçuoğlu, Baran, 2025, Resolving a 182 - year-old taxonomic puzzle: Luciobarbus in the Persian Gulf basin (Teleostei: Cyprinidae), Zootaxa 5620 (1), pp. 29-71 : 38-65

Barbus rajanorum .

Heckel (1843) described B. rajanorum from the Qweiq at Aleppo. He counted 65 total lateral-line scales in this species, the same as those represented on the figure by Heckel (1843: Fig. 1 on Plate 14, reproduced here as Fig. 5 ), and in the type of the species NMW 54494 ( Fig. 6 ) examined by JF. Karaman (1971) listed L. rajanorum as a valid species without examining the type, but Almaça (1991:64) identified B. rajanorum as a hybrid. Also, Coad (1991) identified it as a hybrid between Luciobarbus pectoralis and Capoeta damascina , providing F. Krupp (in litt. 1986) as a reference for this statement. Nevertheless, Fareed Krupp (pers. comm. 2022) has since clarified that this statement was likely made in a personal communication, and not in any published work, suggesting that Coad (1991) may have confused this with Almaça (1991) and Coad (1995), reproducing the classification of the species from his 1991 publication. Later authors, including Freyhof et al. (2021), followed Coad (1991) without critical evaluation.

As L. pectoralis is endemic to the Eastern Mediterranean basin and likely never occurred in the Qweiq (see discussion below in the chapter about L. kersin ), it certainly cannot be one of the parental species. Therefore, Barbus rajanorum is identified as the hybrid between L. schejch and C. damascina . Hybrids between Luciobarbus and Capoeta ( Fig. 7 ) are frequently found in nature, but cannot be easily identified. The head and mouth shapes, as well as the length of the barbels, are the key characters that distinguish Luciobarbus from Capoeta . These are quite variable in Capoeta , and hybrids bridge the gap between both genera. Hybrids are identified by their unequal number of chromosomes, as all Luciobarbus are tetraploid and Capoeta are hexaploid ( Yang et al. 2015). This character cannot be examined in the field or on preserved materials. Compared to Capoeta , Luciobarbus x Capoeta hybrids possess relatively longer heads, and two pairs of long barbels (vs. one pair of short barbels in C. damascina ), and the lower jaw is much less or not keratinised. There is also a thin but usually fully developed lower lip in the hybrids (vs. interrupted in Capoeta ). Compared to Luciobarbus , the most evident characters in the hybrids are the stouter and deeper body; thinner last unbranched dorsal-fin ray, blunt head, and shorter snout. Also, the lower lip is often interrupted and thin, never fleshy, or with a median lobe, as in many Luciobarbus . Following Coad (2021), L. barbulus (= L. schejch ) has 46–59 total lateral line scales, and following Alwan (2010), C. damascina has 75–90 total lateral line scales, based on fishes analysed from the Euphrates and Tigris. The type specimen of B. rajanorum , having a short snout, thin lips and no keratinised lower jaw, has 65 scales, fitting perfectly between L. schejch and C. damascina .

Berg (1949) included L. mystaceus and L. schejch as synonyms of B. rajanorum , and Ladiges (1960), and Karaman (1971) followed this opinion. However, there is no reason to reject the hypothesis by Almaça (1991) (that L. rajanorum is a hybrid), as we also identify L. rajanorum as a hybrid. Heckel (1843) had already pointed out the high similarity of B. rajanorum to C. damascina , but he described it as a species of Barbus .As a hybrid, B. rajanorum is an unavailable name for the Luciobarbus species of the Euphrates and Tigris, as there are no indications that these hybrids might have formed their own reproductive unit independent from their parental species.

Luciobarbus mystaceus .

From the description by Pallas (1814:293), it is evident that L. mystaceus is a synonym of L. mursa , a species endemic to the Caspian Sea basin, and its type locality is in the Kura River. Already, Heckel (1843:252) indicated that Pallas’s L. mystaceus is identical to L. mursa . Heckel (1843) is not the author of L. mystaceus , as he only identified some of his material as this species. Potentially, it was Karaman (1971), followed by Almaça (1986), who suspected that Heckel (1843) was the first author of L. mystaceus , and that the species has its type locality in the Tigris in Mosul. Subsequent authors followed this opinion, leading to widespread confusion (see Tsigenopoulos et al. 2003, Kaya et al. 2016, Valiallahi & Coad 2017a). Berg (1949) pointed out that L. mystaceus is based partly on L. mursa and L. capito , while Bogutskaya et al. (2003) considered it as most probably a synonym of L. capito . Neither L. capito nor L. mursa have been confirmed from Mosul or other places in the Persian Gulf basin. The fish identified as L. mystaceus ( Fig. 8–10 ) by Heckel (1843) are housed at NMW and are not conspecific with L. mursa , but with L. schejch .

Barbus barbulus .

Luciobarbus barbulus was described by Heckel (1849) based on syntypes from the Mond in the Iranian Gulf basin, and from the Qweiq in Syria. Only the syntype ( NMW 6596, Fig. 11 ) from Iran is available at NMW. The molecular analysis in Khaefi et al. (2017), and the morphological analysis in Valiallahi & Coad (2017a, b), and results of the current study, suggest that L. barbulus is not endemic to the Mond but also occurs widely in other rivers of the Persian Gulf, including the Tigris. As we find no characters to distinguish NMW 6596 from L. schejch , L. barbulus is treated as a junior synonym of L. schejch . This issue was not resolved earlier, as types of L. schejch have thin, not fleshy lips, and so it was identified as L. pectoralis , while fish with thick lips were identified as L. barbulus (see discussion below).

Luciobarbus schejch .

Heckel (1843) distinguished his materials identified as L. mystaceus from sympatric L. schejch ( Fig. 8 ) by the width of the fleshy lips with a central lobe in L. mystaceus (vs. thin, without a lobe in L. schejch ). However, what Heckel considered a “central lobe” is a mental pad following the nomenclature of Kottelat & Freyhof (2007), which can become pendulous as the fish grows. This character has also been widely used to distinguish fish identified as L. pectoralis (with thin lips) from those identified as L. barbulus (with thick, fleshy lips) in the Persian Gulf basin ( Coad 2010). The length and strength of the last unbranched dorsal-fin ray are often mentioned as additional characters that distinguish these species. Günther (1868) had already mentioned that the presence or absence of a median lobe/pad in the lower lip is of no taxonomic value as its development is subject to great variation. Also, Krupp (1985), who studied fish identified as L. pectoralis and L. barbulus , mentioned that the size of the mental pad in the lower lip, thickness of the lips, as well as the strength and length of the last unbranched dorsal-fin ray, are highly variable characters not suitable to distinguish these species. Recently, Khaefi et al. (2017:838) analysed fish identified as L. barbulus by molecular characters and again found the median pad/lobe quite variable: i.e., “with or without a median lower lip lobe”. The thickness of the lips, and the presence or absence of the median pad have been demonstrated to be very variable in other Luciobarbus species, especially in North Africa ( Brahimi et al. 2017, 2018), and questioned as a taxonomic character in African Labeobarbus by Vreven et al. (2016), where different shapes of lips are widespread within populations.

There is no published evidence to consider the fish identified as L. mystaceus by Heckel (1843) as a species distinct from L. schejch . Indeed, some syntypes of L. schejch ( Figs. 12–13 ) have a surprisingly short snout compared to Heckel’s L. mystaceus ( Fig. 8–10 ) and we understand, that earlier authors hesitated, to identify them as the same species. Heckel’s L. schejch and L. mystaceus could easily be differentiated by this character if only these individuals were available. But L. schejch is very variable in its body shape and snout length (see below, Fig. 12–16 , 20–21 , 31–36 ), and the individuals identified as L. mystaceus by Heckel (1843) are here recognised as L. schejch as we find no difference to distinguish them from others. Heckel (1843) already noted that the local fishers in Mosul (“die Araber [the arabs]”) did not distinguish between L. schejch and L. mystaceus .

Fish identified as L. schejch and L. mystaceus by Heckel (1843) correspond to those identified by Coad (2010) as L. barbulus , L. pectoralis and L. xanthopterus in the Gulf basin. As L. xanthopterus is an unavailable name (see below), L. pectoralis is endemic to the Mediterranean basin, and L. schejch has priority over L. barbulus , L. schejch is the name available for this widespread species. Luciobarbus schejch is diagnosed by having 11–24 total gill rakers on the first arch, 50–62 total lateral-line scales, snout length 1.1–1.9 times postorbital length, and an inferior or slightly subterminal mouth. Many individuals (but not all), especially juveniles, have a long and strong last unbranched dorsal-fin ray ( Fig. 13–14 , 32–34 ), a character state not seen in other Luciobarbus in the Persian Gulf basin, and certainly not in L. pectoralis from the adjacent Mediterranean basin. Coad (2010) also reported relatively low numbers of total lateral-line scales in L. schejch (43), but on our materials, the lowest number counted was 52, from an individual collected in Iran.

The variability of head and body shapes, as well as the colouration of live individuals in L. schejch , is indeed surprising ( Fig. 12–16 , 20–21 , 31–36 ) and is likely the reason why it was not previously understood whether these fish are indeed conspecific. We fully appreciate authors such as Heckel (1843) and Valiallahi & Coad (2017a), who were confident of identifying more than one species in what we call L. schejch in our study. If just character extremes are compared, such as silvery stout or blackish fishes with thin lips, and slender, yellowish individuals with thick lips, we would also have suggested that two species are involved. Especially in wetlands and large rivers, the body shape ranges from stout, deep-bodied individuals ( Fig. 31–36 ) to very slender ones, with short or long heads, and from silvery individuals to blackish or yellowish ones. Indeed, our genomic data indicate that most individuals morphologically identified as L. schejch showed signatures of hybridisation with L. esocinus . In addition to morphological variation caused by diverse ecological conditions throughout their distribution range, hybridisation with L. esocinus is likely one of the important reasons for the high variability of L. schejch in body shape and colouration (see more detailed discussion below).

Barbus kersin .

Like B. rajanorum , this species has been described from the Qweiq in Aleppo. The six syntypes of L. kersin examined have 43, 52, 53, 56, 56, and 59 total lateral line scales largely within the range of L. schejch (50–62). Only NMW 54213/1 (138 mm SL) has 43 scales, remarkably fewer than the others, suggesting it may be an aberrant individual. Luciobarbus kersin occurred in sympatry with L. schejch ; two individuals from Aleppo ( NMW 54385, Fig. 11 ) were originally identified as L. mystaceus (= L. schejch ) by Heckel (1843). Heckel (1843) gives 55–56 lateral line scales for L. kersin , indicating that not all fish identified as syntypes in NMW might have been considered by him. All individuals (except NMW 54213/1) in NMW 54212, NMW 54213, NMW 54215, and ZMB 3237 ( Fig. 16–17 ) examined for this study are identical to L. schejch in every meristic counts.

Since Heckel’s times, the Qweiq has dried out in Syria, and no Luciobarbus has been found in its remaining small headwaters in TÜrkiye ( Dağlı & Erdemli 2009; JF personal explorations). The population is therefore likely extirpated, and no fresh materials have been collected since Heckel (1843). Hamidan et al. (2014) discussed the biogeographic relationships of the fishes of the Qweiq, which has most species in common with the Euphrates located adjacent to the east, except a few species otherwise found only in the Orontes located adjacent to the west. There are no endemic fish species known from the Qweiq. Luciobarbus in the Qweiq were likely conspecific with Luciobarbus found in adjacent Euphrates. Despite this, we consider the possibility that the fish of the Qweiq might have been L. pectoralis ( Fig. 19 ) from the Orontes, or even hybrids between L. schejch and L. pectoralis . The two individuals from the Qweiq identified as L. mystaceus by Heckel (1843) are very likely conspecific with those identified as L. kersin by Heckel (1843), as we found no differences, and it is very hard to imagine that two such similar species might occur in syntopy. It might be worthwhile attempt to extract DNA from the type series to be analysed using advanced museomic methods, as discussed by Raxworthy & Smith (2021). We made a similar attempt on ZMB 3237, but were unsuccessful.

Luciobarbus pectoralis and L. schejch are closely related, and we are unaware of any study presenting morphological characters to distinguish these species. Their mean K2P genetic distance in COI is only 1.77% ( Table 3 ), which falls within the intraspecific variation of many freshwater fish species ( Geiger et al. 2014, Freyhof et al. 2022). Luciobarbus pectoralis has 47–59 total lateral line scales, overlapping with that of L. kersin and L. schejch , and 14–20 gill rakers (vs. 16, 18, 18, 19, 19, 23 in types of L. kersin , 11–24 in L. schejch ). Both these meristic counts do not allow identification of syntypes of L. kersin as L. schejch or L. pectoralis . Coad (2010) distinguished L. kersin from L. pectoralis by having the dorsal-fin origin situated behind the vertical of the pelvic-fin origin, a character state we confirm in the types of L. kersin . We also support the statement of Coad (2010), as most (but not all) individuals of L. pectoralis we examined have the dorsal-fin origin situated above, or in front of the pelvic-fin origin. But the situation is more variable in L. schejch , as in this species, the dorsal-fin origin is usually situated behind the vertical of the pelvic-fin origin (as in L. kersin ). However, in several specimens we examined, it is situated either above or in front of the origin of the pelvic fin (as in L. pectoralis )—a character which suggests that L. kersin is more likely to be conspecific with L. schejch .

The issue needs more research in the future, especially as L. kersin has a general appearance quite different from many (but not all) individuals of L. schejch (very short head and deep body), making it look more similar to L. pectoralis . Also, both L. pectoralis and L. kersin lack an enlarged last unbranched dorsal-fin ray, which is commonly (but not always) found in small and medium-sized individuals of L. schejch . Our molecular data restrict L. pectoralis to the Mediterranean basin, but we cannot exclude the fact that it was also found in the Qweiq, and could now be extirpated.

We do not treat L. kersin as a valid species, since we see no clear character state distinguishing it from L. schejch and L. pectoralis . As the types of L. kersin fall within the variability of L. schejch , we treat the former as a synonym of the latter (See Fig. 20 for an individual of L. schejch superficially similar to the types of L. kersin ). If L. kersin is indeed a valid species, then it would be extinct globally, and we hesitate to conclude this from the few individuals available. Also, it would be the only species endemic to the Qweiq, which is relatively unlikely as the Qweiq was a tributary of the Euphrates until the early Holocene ( Hamidan et al. 2014). Luciobarbus kersin could be conspecific with L. pectoralis as both have a relatively stout body and short head (vs. slender body and usually long and pointed head in L. schejch ). In that case, L. kersin would become a synonym of L. pectoralis . Alternatively, it could also be conspecific with L. schejch because of the similar position of the dorsal-fin origin. As we give here priority to L. schejch over L. kersin (see below), it becomes a synonym of L. schejch . Alternatively, L. kersin could also be a hybrid between L. pectoralis and L. schejch , and we believe that this hypothesis can only be tested using high-resolution museomic methods, as both potential parental species are closely-related.

The short and round head of L. kersin could also indicate that the species is a hybrid between Luciobarbus and Capoeta . When Heckel (1943) described L. kersin , he already noted that this species is similar to L. rajanorum (“Die zweite stumpfnasige Barbe gleicht sehr dem vorhergehenden B. rajanorum ”— translated as “the second blunt-nosed barb is very similar to the previous B. rajanorum ”]), a species based on a hybrid between Luciobarbus and Capoeta . It is to be noted that NMW 54215 and ZMB 3237 are superficially similar to hybrids between Luciobarbus and Capoeta ( Fig. 5–7 ). Having 59 total lateral line scales, it cannot be fully excluded that NMW 54215 is indeed such a hybrid, as the head shape is similar to such hybrids. But as Capoeta damascina has greater number of scales than L. schejch , their hybrids should also have more scales than L. schejch , or their scale counts should be in the upper part of the counts in L. schejch . ZMB 3237 has 56 total lateral line scales (and 16 gill rakers), a count representing the middle of the range of L. schejch , rejecting the hypothesis that L. kersin is a hybrid between Luciobarbus and Capoeta . Also, NMW 54215 has too few scales to qualify as a hybrid (see also the discussion under B. rajanorum above). From a cytogenetic point of view, Luciobarbus is tetraploid (2n=100), and Capoeta is hexaploid (2n=150), and their hybrids would theoretically have 2n=125 an odd chromosome number, which could result in meiotic irregularities and potential infertility.

It is difficult to draw a conclusion based on the limited materials available. As L. kersin cannot be identified as a hybrid between Luciobarbus and Capoeta , it is very unlikely an extinct species endemic to the Qweiq, and it must either be a synonym of L. schejch or L. pectoralis , or a hybrid between the two. As types of L. kersin have their position of dorsal and pelvic fins similar to L. schejch , and that the sympatric NMW 54385 also share the general appearance with L. schejch , we preliminarily treat L. kersin as a synonym of L. schejch , rather than as a synonym of L. pectoralis . This case might be revisited when more information about character states differentiating L. pectoralis and L. schejch becomes available.

Heckel (1843: Plate XIV) showed NMW 54215 (reproduced as Fig. 17 , specimen in Fig. 18 ) as the largest and best-preserved syntype at NMW. Coad (2010) published a drawing of this syntype as a reference to identify the species. NMW 54215, as well as the well-preserved syntype ZMB 3237 ( Fig. 18 ), have different appearances from the fish identified as L. kersin by Khaefi et al. (2018) and Jouladeh-Roudbar et al. (2020) ( Fig. 21 ). The fish identified by these authors as L. kersin have very deep bodies, and we cannot fully exclude that they are malformed. Jawad et al. (2015) had also reported an individual of L. schejch (as L. xanthopterus ) with vertebral coalescence strongly resembling the fish identified as L. kersin by Khaefi et al. (2018) and Jouladeh-Roudbar et al. (2020).

Khaefi et al. (2018) distinguished L. kersin from L. barbulus (= L. schejch ) by having a higher body depth, measured in front of the dorsal-fin origin (111–130% HL in L. kersin vs. 79–107% HL in L. barbulus ). In the five syntypes of L. kersin measured for this character, the body depth is 94–122% HL, overlapping with the measurements taken in both “species” by Khaefi et al. (2018). Body depth does not seem to be a character suitable for distinguishing L. schejch from L. kersin . Most likely, Khaefi et al. (2018) selected individuals with a very deep body and identified these as L. kersin . However, these preselected individuals cannot be positively identified as L. kersin . Khaefi et al. (2018) also made one sequence of their L. kersin available in NCBI GenBank ( MF599072 ), which Parmaksız et al. (2022) used as a reference to uncritically identify fish from the Turkish Euphrates as L. kersin . This particular sequence ( MF599072 ) is placed in L. schejch COI clade B in our analysis (see discussion below).

It also remains to be seen how Khaefi et al. (2018) distinguished their postulated hybrid of L. barbulus (= L. schejch ) and L. kersin from the parental species. Their alleged hybrid between L. barbulus and L. kersin (MC1814, GenBank accession number MF599071 ) is placed within L. barbulus (= L. schejch ) in their tree, and since they did not provide evidence on how to distinguish it from L. barbulus and L. kersin , we identify this postulated hybrid as an individual of L. schejch .

First Reviser actions to stabilise the use of Luciobarbus schejch . The First Reviser action is a principle which is used in conflicts between simultaneously published names, where the first subsequent author can decide which has precedence (ICZN 1999: Article 24.2). Heckel (1843) described L. esocinus , L. pectoralis , L. xanthopterus , L. schejch , L. kersin , L. perniciosus , and L. rajanorum . Of these, Luciobarbus esocinus , L. pectoralis , and L. schejch are accepted here as valid species. Luciobarbus pernicious is treated as a synonym of L. pectoralis by Karaman (1971), who gave priority to L. pectoralis over L. pernicious . As discussed in this study, L. rajanorum is based on a natural hybrid and is not an available name. None of the studies referred to in the introduction, and in Table 1 of this paper, ever treated L. schejch and L. kersin as conspecific. For example, Karaman (1971) treat L. kersin as a subspecies of L. capito and L. schejch as a subspecies of L. rajanorum . Neither Karaman (1971), Almaça (1983, 1991), Krupp (1985), Coad (1991, 2010, 2021), nor any subsequent authors accepted L. schejch as a synonym of L. kersin , or vice-versa. They were either treated as valid taxa, as synonyms of L. pectoralis , or L. kersin as a valid species and L. schejch as a synonym of L. pectoralis . To stabilise the names in Mesopotamian Luciobarbus , we act as First Revisers to give priority to L. schejch over L. kersin and treat L. kersin as a subjective synonym of L. schejch , maybe until the case is reviewed again by museomics approaches.

Luciobarbus xanthopterus .

The difference in general appearance between L. xanthopterus and L. esocinus is evident from Heckel’s plate IV (reproduced here as Fig. 22 ). While there is no doubt that L. esocinus represent a valid species (see also below), the identity of L. xanthopterus is confusing. NMW has three syntypes of L. xanthopterus ( NMW 91215, NMW 54786, NMW 54841/1, Fig. 23 ) and an additional jar with nine juveniles ( NMW 54841). These might be syntypes also, as Heckel (1843:64) mentioned ‘small fish’ in his description. The nine juveniles in NMW 54841 are identified as L. esocinus , having 60–70 scales along the lateral line and 9–11 gill rakers. The individual NMW 91215 is recognised as L. schejch , having a low snout/postorbital length ratio (1.7), and only 59 total lateral line scales. No gill rakers could be counted because the specimen is stuffed and dry. NMW 54786 is the specimen illustrated by Heckel (1843, Plate IV), and the one the description is based on. NMW 54786 and NMW 54841/1 ( Fig. 23 ) have a long, slightly barbel-shaped head with an almost terminal mouth, 60 and 62 total lateral line scales, and 13 total gill rakers. In NMW 54786, the snout length is 1.9 and in NMW 54841/1, it is 1.8 times the postorbital length. NMW 54786 and NMW 54841/1 have their numbers of total lateral line scales within the range overlapping between L. esocinus and L. schejch (59–71 in L. esocinus vs. 50–62 in L. schejch ), more gill rakers than L. esocinus (8–11 in L. esocinus , 11–24 in L. schejch ), and the snout/postorbital length ratio is at the uppermost range of L. schejch (2.0– 2.9 in L. esocinus vs. 1.1–1.9). These intermediate character states indicate that NMW 54786 and NMW 54841/1 are likely hybrids between L. esocinus and L. schejch . Due to the similarity of the two syntypes of L. xanthopterus with L. esocinus , it is not surprising that Karaman (1971) and Coad (1991) mentioned that L. esocinus and L. xanthopterus might be conspecific.

There are few individuals in our dataset having character states similar to the types of L. xanthopterus . For example, VPFC AhvazBazar 2023.05 ( Fig. 25 ) has 61 total lateral line sales and the ratio between the snout and the postorbital length is 2.4 (indicating it might be L. esocinus ). Additionally, they possess 14 gill rakers, and its mouth is inferior (suggesting it might be L. schejch ). This individual and the other two in Fig. 25 are identified as potential hybrids between both species. Three other individuals have 11 total gill rakers, but a snout/postorbital index below 2.0, and having 53, 54, and 55 total lateral line scales. They are identified as L. schejch . We received images ( Fig. 26 ) of three fishes having an almost terminal mouth and 60 or fewer scales. From these, no gill-raker numbers or molecular data are available. They also superficially agree with the types of L. xanthopterus , and their lateral line scales are intermediate between L. esocinus and L. schejch . Only for one of these morphologically intermediate individuals were we able to develop genomic data: VPFC AhvazBazar 2023.05 (FSJF-DNA 3772, Fig. 27 ). Our genomic data place this individual close to, but not within L. esocinus . It has mtDNA in COI clade A, 11 gill rakers, 58 total lateral line scales, and a snout length 1.7 times postorbital length.

Most fish morphologically identified as L. schejch are hybrids between this species and L. esocinus (see below). The syntypes of L. xanthopterus are also likely such hybrids. Back-crossing with L. esocinus could explain their morphological similarity to the syntypes of L. esocinus . Indeed, there are not enough such intermediate individuals studied for their genomic data to provide strong support for this hypothesis, and more research is required to understand the situation better.

From a taxonomic perspective, it is important to understand if L. xanthopterus is a distinct species, or a hybrid between the other two. Fish that morphologically correspond to the types of L. xanthopterus are rare in our material, and show no unique but intermediate character states between the other two species. Our COI sequence data are not of any use in resolving this puzzle, and the few individuals sequenced are all placed in the COI clade A (see below). Only one individual analysed for genomic data is placed very close to, but outside of, L. esocinus ( VPFC AhvazBazar 2023.05 FSJF-DNA 3772, Fig. 27 ).

Luciobarbus xanthopterus is a part of the ‘ L. esocinus x L. schejch hybrid complex’ discussed below. Luciobarbus schejch forms many largely or fully reproductively isolated populations independent from L. esocinus . While L. schejch might be a species of hybrid origin (see discussion below), this is not the case in those hybrids corresponding to the syntypes of L. xanthopterus . Such fish are rare and were not found in geographic isolation, but always occurred together with L. esocinus and L. schejch . They seem unlikely to represent their own reproductive unit independent from the other two species. Therefore, we reject the hypothesis that L. xanthopterus is a species of its own. As L. schejch and L. xanthopterus are part of the L. esocinus x L. schejch hybrid complex, they are best treated as conspecific. Günther (1874) treated L. xanthopterus as a synonym of L. schejch and by this, had already given priority to L. schejch over L. xanthopterus .

We designate NMW 54786 as a lectotype of L. xanthopterus , as two species (or two different hybrids) are included in the type series: NMW 91215 might be an individual of L. schejch, NMW 54841 is identified as L. esocinus, NMW 54786, and NMW 54841/1 are identified as hybrids between L. esocinus and L. schejch . The lectotype is the specimen on which Heckel’s description is based on, and the one illustrated in his Fig. 1 of Plate IV (reproduced here as Fig. 22 ). It is identified as a hybrid between L. esocinus and L. schejch due to its character states being intermediate between both species (see above). This excludes L. xanthopterus from the list of available names, until it can be demonstrated that fish corresponding morphologically to the lectotype form their stable reproductive unit in nature, potentially as a species of hybrid origin. Until now, there have been no indications for this.

Luciobarbus esocinus .

Luciobarbus esocinus ( Fig. 28–30 ) and L. subquincunciatus are widely accepted and well-diagnosable species ( Coad 2021; discussion and key above for details). However, the position of L. esocinus in our COI analysis demonstrates the need to discuss this species in more detail. We examined 43 L. esocinus for morphological characters, including four syntypes. These are well-distinguished from L. schejch and its hybrids by having more scales (59–71 total lateral line scales vs. 50–62 in our material of L. schejch , n=149), fewer gill rakers (8–11 vs. 11–14), snout length 2.0–2.9 times the postorbital length in individuals larger than 150 mm SL (vs. 1.1–1.9), and juveniles often with many small, dark-brown spots on the flank (vs. absent). In L. esocinus , the snout is short (24–33% HL vs. 30–42 in L. schejch ), and the postorbital distance is long (57–70% HL vs. 45–60). DÜŞÜkcan et al. (2022) counted 60–66 total lateral line scales and 11–15 gill rakers in their materials of L. esocinus (n=32). They also reviewed the published character states of this species, and observed 60–73 total lateral line scales and 10–22 gill rakers in L. esocinus . The fishes with a high gill raker count> 11 might be identified as hybrids of L. esocinus with L. schejch (see discussion above on L. xanthopterus ). Indeed, this case should be revisited. It remains strange that DÜŞÜkcan et al. (2022) identified fish with such high gill raker numbers as L. esocinus . Hybrids might blur the situation in the Özlüce Reservoir, making it impossible to draw a clear line between hybrids and pure L. esocinus . From all the fishes examined during this study, including fresh juveniles and syntypes, we would expect individuals with more than 60 scales, less than 12 total gill rakers, a snout length 2.0 or more times in the postorbital length, a long, pike-like head, and a terminal or almost terminal mouth, so as to be identified as L. esocinus .

We agree with previous authors that L. esocinus is a valid species. However, our mtDNA analysis does not resolve it as a COI- based molecular clade independent from L. schejch . This situation has already been described by Khaefi et al. (2018) and Parmaksız et al. (2022). It should be noted that by purpose, we sequenced only ten individuals of L. esocinus , as their morphological identification is quite clear. Alternatively, we searched more intensively for L. schejch and included 111 individuals of this species in our molecular analysis. There is no doubt that L. schejch represents a species on its own that is also well-supported by morphological and nuclear genomic characters. Indeed, L. esocinus must have extensively hybridised with L. schejch , as this species is largely introgressed by L. esocinus .

The Luciobarbus esocinus x Luciobarbus schejch hybrid complex

COI clades A and B of L. schejch are conspecific. We tried to find characters to distinguish fish identified by morphological characters as L. schejch , nesting in both the COI clades. Fish in COI clade A ( Figs. 14 , 20 , 32 , 34 , 35 ) have 11–19 total gill rakers on the first arch, and 50–61 total lateral line scales (n=46) (vs. 11–19 and 54–60, in L. schejch in COI clade B, n=12, Figs. 31 , 33 , 36 ). We also tried to find differences between genomically “pure” (based on SNP data) and introgressed individuals of L. schejch .

Following our genomic data, the four “pure” L. schejch individuals possess 12, 15, 15, and 15 total gill rakers and 55, 56, 57, and 57 total lateral line scales. Three of these “pure” individuals are shown in Fig. 31 . There are no morphological differences between the four genetically “pure” L. schejch , and those with hybridisation in the past (in COI clade A Figs. 14 , 20 , 32 , 34 , 35 ). This situation reflects the challenges of applying rigid species concepts to groups where ongoing and historical gene flow has blurred genetic and morphological boundaries. In the case of L. schejch , the lack of morphological differences between “pure” and introgressed L. schejch might be caused by backcrossing with the paternal lineage (in this case L. schejch ), which had been documented in other cyprinids as well ( Hashemzadeh Segherloo et al. 2021b). While L. esocinus is morphologically well-distinguished from L. schejch and most of its hybrids, genomically “pure” fish and those belonging to both COI clades of L. schejch cannot be distinguished. Therefore, we treat all these individuals as conspecific.

Most, if not all, L. schejch have been subject to introgressive hybridisation. From 17 fish morphologically identified as L. schejch that have been included in our nuclear genomic analysis, seven (41%) were likely introgressed to various degrees by L. esocinus . This ratio might be even higher, as we selected fish from COI clade B for our nuclear genomic analysis. Together, 102 (84%) fish morphologically identified as L. schejch host the mtDNA of L. esocinus (COI clade A). All six (100%) L. schejch from COI clade A selected for our nuclear genomic analysis show signs of hybridisation. In our COI clade B, there are 21 individuals, nine of which were included in our genomic analysis, within which only four (44%) might be genetically “pure” L. schejch . That means that potentially more than 50% of individuals in COI clade B and 100% of the fish morphologically designated as L. schejch in COI clade A have experienced hybridisation. As most individuals are placed in the COI clade A, more than 90% of fish identified as L. schejch show signs of hybridisation events from their evolutionary past. The widespread introgression observed in L. schejch is consistent with recent evidence from freshwater fishes, particularly cyprinids and Leuciscids, that hybridisation can be both common and evolutionarily consequential ( Sousa-Santos et al. 2014, Dubut et al. 2010). Hybridisation widens morphological and ecological traits, and provides novel raw material on which natural selection, genetic drift, or local adaptation can act ( Hayden et al. 2010, Seehausen 2004). “Pure” L. schejch are all found in the COI clade B, in contrast to only 44% of fish in clade B. Anyhow, these ratios are likely to vary geographically. As each genetically “pure” L. schejch was collected at a different place, together with introgressed individuals, it is not expected that “pure” L. schejch occur as distinct, reproductively isolated populations in nature. Also, our four individuals are likely not genetically 100 % “pure”, but just those with the least amount of genomic DNA from L. esocinus . However, clarification of reproductive relationships between L. esocinus and L. schejch awaits detailed genomic analyses using larger sample sizes from different parts of their geographic range.

This pattern also implies strong unidirectional hybridisation, in which L. esocinus is the maternal, and L. schejch is the paternal lineage. The frequent hybridisation found in our materials also indicates that hybrids are fertile. Therefore, backcrosses and genomic mixtures between both species should ideally occur in nature, an argument supported by our genomic data ( Fig. 3 ), where hybrids span a large distance between parental species. This explains the high variability in head and body shape in fish identified here as L. schejch —some with short and blunt, and others with relatively long and pointed heads. Also, the high variability of the length of the last unbranched dorsal-fin ray could be related to hybridisation. In many headwater populations of L. schejch , allopatric from L. esocinus , this fin-ray is very long and strong, especially, but not exclusively, in juveniles. In L. esocinus of all age groups, this ray is relatively short. In reservoirs, marshes and large rivers, where both species co-occur, there is high variability in the length of the last unbranched dorsal-fin ray in L. schejch . The occurrence of hybrids between L. esocinus and L. schejch implies that there must be individuals in nature that cannot be identified as one of the parental species, a prediction confirmed by this study. This situation remains unsatisfying for fisheries managers and other taxonomy users trying to identify fish in the field or laboratory.

Luciobarbus schejch in COI clades A+B have no distinct distribution range. Fish in COI clade B were always found in sympatry with fish in COI clade A. Most of the individuals in COI clade B have been found in Iran, primarily in the Shadegan wetlands and the lower Karun, and lower Karkheh rivers, where our collecting efforts have been most intensive. But we found fish from the same COI clade also in Türkiye, the middle Euphrates in Syria, the lower Euphrates in Iraq, and the Greater and Lesser Zab in Iraq, indicating it is widespread in the Euphrates and Tigris drainages.

Hybridisation between both species has been ongoing for a very long time, potentially since both came in contact after their independent invasion of the Persian Gulf basin from the Mediterranean drainages. Luciobarbus pectoralis , L. esocinus , L. longiceps , and L. schejch might have had a common ancestor that likely invaded Mesopotamia twice, resulting in the COI clades A ( L. esocinus ) and B ( L. schejch ) ( Fig. 1 ). The 1.77% K2P distance in the COI sequence data between L. schejch and L. pectoralis are here hypothesised to indicate a Pliocene (<2 million years) invasion from the Mediterranean. The occurrence of L. schejch with mtDNA of L. esocinus in the Iranian Mond and Helleh drainages (where L. esocinus is absent) indicates a pre-Holocene hybridisation between both species. Luciobarbus schejch introgressed by L. esocinus must have colonised these rivers, when sea levels have been much lower in the Gulf basin during glaciations, connecting the Mond and Helleh to the Tigris. Interestingly, a phylogeographic structure within L. schejch clade A indicates an isolated evolutionary history of some populations of L. schejch with mtDNA of L. esocinus , also an indication that their hybridisation has been ongoing for very long.

Both species can avoid forming one hybrid swarm. Luciobarbus schejch is a species strongly impacted by hybridisation, and most, if not all, samples studied originate from specimens that have hybridised with L. esocinus in their evolutionary past. The fact that they can be recognised as two morphological species demonstrates, that they represent two mostly reproductively isolated lineages that occur in sympatry or allopatry in parts of their range. Indeed, if there were no reproductive barrier between both, L. esocinus would have disappeared and would no longer be a distinct diagnosable unit. That L. esocinus is a distinct diagnosable unit is demonstrated by our morphological and genomic data. Luciobarbus esocinus is reproductively isolated and can protect its gene pool from introgressive hybridisation from L. schejch . But this must logically not be the case in L. schejch . Naturally, as we included only 10 individuals of L. esocinus in our analysis, additional samples might challenge this finding. Hybridisation between L. esocinus and L. schejch seems largely unidirectional (female L. esocinus and male L. schejch ) as most L. schejch have the mtDNA of L. esocinus . Both species might have different spawning places or spawning times, but data from the field is lacking. It can only be speculated that male L. esocinus could exclude most male L. schejch and male hybrids from spawning with its females. By this, they might avoid the collapse of L. esocinus into a hybrid swarm. Female F1 hybrids might be less attractive to male L. esocinus but very attractive to male L. schejch , and such females mostly reproduce then with other hybrids and L. schejch , successfully increasing the introgression of the gene pool of L. schejch with DNA from L. esocinus . Along this pathway or via different processes, L. esocinus can retain its reproductive isolation against L. schejch . But vice versa, this might not be the case.

Luciobarbus schejch is resilient against introgressive hybridisation from L. esocinus . It is unclear if the hybridisations with L. esocinus have changed the evolutionary fate of the ancestral lineage of L. schejch . The high morphological similarity of L. schejch , including its hybrids, with Mediterranean L. pectoralis and L. longiceps , is striking. It might be speculated that there was an evolutionary character displacement after both came into contact following their invasion of Mesopotamia. That character displacement seems to have happened more on the side of L. esocinus than on the side of L. schejch . Indeed, many thousands of years of (unidirectional) hybridisation and introgression should have diluted the gene pool of L. schejch to such a degree that both species are no longer distinguishable. As this is not the case, it indicates a strong resilience of L. schejch against introgression.

Indeed, the term “hybrid” is too general when discussing the Luciobarbus species studied here. Generally speaking, F1 hybrids between two species should be morphologically intermediate between parental species. Such individuals are very rare in our materials. This indicates that F1 hybridisation might be a rare event. The strong morphological similarities between L. pectoralis , L. longiceps , genomically “pure” individuals of L. schejch , and individuals showing mtDNA introgression and/or genomic introgression are striking. This suggests that backcrosses (F2–Fx) occur mostly with L. schejch , not L. esocinus . By this process, hybrids become more and more similar to L. schejch ; genomically “pure” individuals taken as references. Therefore, we do not simplistically identify all or most individuals of L. schejch as hybrids, but rather as members of L. schejch with introgressions from L. esocinus .

This case is a complicated example in which unidirectional hybridisation has shaped the genome of a species and diversified its morphology and likely ecology. We support the hypothesis that both L. esocinus and L. schejch , including their hybrids, form two largely reproductively isolated units similar to other species, in which introgressive hybridisation occasionally occurs.

Luciobarbus esocinus and L. schejch are two species. From a taxonomic point of view, it is relevant to discuss whether L. schejch might be based on syntypes that are hybrids. The geographically isolated populations of L. schejch in COI clade A indicate that hybridisation is not a recent result of dams or other environmental modifications. Therefore, we expect that the syntypes of L. schejch (and L. kersin ) might be individuals introgressed by L. esocinus , as is the case for most L. schejch today. We have no genomic data from the syntypes, as we could not extract DNA from their tissues. However, the syntypes cannot be distinguished from genomically “pure” L. schejch by morphological characters. In contrast to the syntypes of L. xanthopterus , none of the syntypes of L. schejch shows character states that are intermediate between L. schejch and L. esocinus . Therefore, the syntypes, as most other individuals we identified, are recognised as L. schejch (very likely introgressed by L. esocinus ). That means the name L. schejch is not based on hybrids (F1), but on individuals of a species different from L. esocinus that has experienced some introgressive hybridisation in its evolutionary past. Luciobarbus schejch is the valid name available for this group. The fish identified morphologically as L. schejch are definitely L. schejch , most if not all, being introgressed by L. esocinus at various degrees.

Indeed, L. schejch could be a species of hybrid origin, as it is largely isolated from L. esocinus , but carries a long history of unidirectional hybridisation and, by this, the introgression of genes. If we treat L. schejch as a species of hybrid origin, we should consider that hybridisation must have happened not only once but many times, and is still ongoing. Therefore, it is unlikely that L. schejch —if treated as a species of hybrid origin—has a monophyletic evolutionary history. However, monophyly in a species’ early stages is neither required, nor is a logical situation in speciation ( Mayden 1999). A polyphyletic evolutionary history could also result in a monophyletic species after it reaches reproductive isolation following the initial polyphyletic steps of speciation. We do not treat L. schejch as a species of hybrid origin. The large similarity of L. schejch and Mediterranean L. pectoralis and L. longiceps indicates that it is not independent of one of its parental species (“pure” L. schejch ). Further, L. schejch forms many largely or fully reproductively isolated populations independent from L. esocinus . In conclusion, we accept L. schejch as a valid species, considering that there was introgressive hybridisation from L. esocinus and that many (84% in our study) individuals of L. schejch are carrying the mitochondrial genome of L. esocinus . Recognising L. schejch as a valid species suggests that widespread introgression does not necessarily eliminate distinct evolutionary trajectories. Instead, it may have the potential to reinforce them by allowing populations to explore a wider part of the adaptive mosaic. In this way, it promotes resilience and long-term persistence in the context of ongoing ecological and climatic challenges.

Conclusion. There are two major conclusions of this study: 1) Even in the year 2025, there are largely unresolved taxonomic puzzles in vertebrates, in well-researched areas, and in fish groups which received significant interest in the past. In such cases, historic type material remains the key to resolving taxonomic challenges, and an integrative analysis of morphological as well as nuclear and mitochondrial evidence is essential to understanding the case, and 2) After centuries of relying solely on morphology, and two decades of largely mtDNA studies, we increasingly enter a new era where nuclear molecular data are applied to understand taxonomic challenges awaiting our attention. Nuclear DNA data are essential in cases where there seem to be mismatches between results from morphological and mtDNA studies.

Hybridisation between fishes is widespread but not well-studied. It is a focal research area, especially (but not exclusively) in East African cichlids (see Peñalba et al. 2024 and citations therein), but its importance in many other geographical areas should not be underestimated. Widespread hybridisation is already known from Mesopotamian freshwater fishes. Freyhof et al. (2018) found all Alburnus caeruleus ( Leuciscidae ) in the Tigris carrying the mtDNA of sympatric A. sellal , and all Cyprinion kais ( Cyprinidae ) mtDNA of C. macrostomum (JF own unpublished results). Alwan (2010) highlighted the occurrence of sympatric individuals of Capoeta ( Cyprinidae ) with low ( C. damascina ) and high ( C. umbla ) lateral-line scale counts, all sharing the same mtDNA. Khaefi et al. (2016) published the finding of a widespread mismatch between morphological and mtDNA identification in Squalius berak and S. lepidus ( Leuciscidae ), and we (JF & IHS own unpublished results) found all Squalius verepi studied from the Tigris to have the mtDNA of S. berak or S. lepidus . Based on these observations (and many others), we curiously look forward to the era of nuclear DNA studies to better understand the diversity of fishes.

Material examined

Barbus rajanorum .

NMW 54494 , holotype, 190 mm SL; of Syria: Aleppo [about 36.212, 37.150] .

Luciobarbus esocinus .

NMW 54088 , 2 , syntypes, 58–61 mm SL ; NMW 54091 , 1 , syntype, 372 mm SL ; NMW 54092 , 1 , syntype, 321 mm SL ; NMW 54841 /2–10; 9, 49–64 mm SL; Iraq: Mosul [about 36.355, 43.126] .— FSJF 3819 , 10 , 190–249 mm SL; Iraq: Ducan reservoir, 36.1000 44.9000 .— VPFC Shadegan 1393.11, 18, 248– 322 mm SL; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381) .— VPFC Horalazim 1392.11, 2, 203– 228 mm SL; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market, 31.7183, 47.9769) .

Luciobarbus longiceps .

FSJF 2721 , 5 , 44–150 mm SL: Syria: Yarmuk in Wadi Jallayn , 32.7392, 35.9822 .

Luciobarbus pectoralis .

FSJF 2329 ; 2, 115– 160 mm SL; TÜrkiye: Adana prov.: Çakıt south of SalbaŞ, lower part of Pozantı , 37.0961, 35.1170 .— FSJF 2424 ; 1, 82 mm SL; TÜrkiye: Adana prov.: ÜçÜrge at Karakuyu , 37.1385, 35.1422 .— FSJF 2450 ; 1, 154 mm SL; Türkiye: Adana prov.: Körkün at Karakuyu , 37.1529, 35.1606 .— FSJF 2697 ; 14, 119– 192 mm SL; Syria: Lake Qattinah south-east of Homs, 34.6619, 36.6183 .— FSJF 2854 ; 5, 88–185 mm SL; Türkiye: Osmaniye prov.: Savrun at Kadirli , 37.3731, 36.0925 .— FSJF 2988 ; 2, 172– 395 mm SL; Türkiye: Adana prov.: fish marked at Ceyhan north of Sakarcalık , 37.1933, 36.0828 .— FSJF 3078 ; 10, 55–145 mm SL; TÜrkiye: Mersin prov.: Göksu at Hamam, two km west of Mut , 36.6313, 33.3674 .

Luciobarbus schejch :

NMW 50399 , 1 , syntype of L. schejch ; 138 mm SL ; NMW 54520 , 2 , syntypes of L. schejch ; 269–275 mm S; Iraq: Mosul [about 36.355, 43.126] .— NMW 54212 , 1 , syntype of B. kersin ; 145 mm SL ; NMW 54213 , 4 , syntypes of B. kersin ; 87–138 mm SL ; NMW 54215 , 1 , syntype of B. kersin ; 163 mm SL ; ZMB 3237 , 1 , syntype of B. kersin ; 143 mm SL; Syria: Aleppo [about 36.212, 37.150] .— NMW 54841 , 1 , lectotype of L. xanthopterus , 216 mm SL ; NMW 54786 , 1, 293 mm SL ; NMW 91215 , 1 , about 430 mm SL; Iraq: Mosul [about 36.355, 43.126] .— NMW 6596 , 1 , syntype of L. barbulus , about 125 mm SL; Iran: Qarah Aqaj .— NMW 54385 , 2 , 142–157 mm SL, identified as L. mystaceus ; Syria: Aleppo [about 36.212, 37.150] .— NMW 16472 , 1, 420 mm SL, identified as L. mystaceus ; NMW 50394 , 2 , 104 – 210 mm SL, identified as L. mystaceus ; NMW 54384 , 2 , 233 – 260 mm SL, Iraq: Mosul [about 36.355, 43.126] .— FSJF 2861 , 1, 195 mm SL. Türkiye: Batman prov.: Tigris 5 km west of Hasankeyf , 37.7238 41.3605 .— FSJF 2931 , 1, 121 mm SL; Türkiye: Gaziantep prov.: stream Merziman at Bağtepe , 37.3248 37.6445 .— FSJF 2998 , 2 , 122–125 mm SL; TÜrkiye, Gaziantep prov.: stream Merziman south of Yavuzeli , 37.2924 37.7231 .— FSJF 3382 , 9 , 61–104 mm SL; Iraq: Lesser Zab at Altun Kopri , 35.7169 44.1167 .— FSJF 3387 , 3 , 108–143 mm SL; Iraq: Tabin west of Zarbi , 35.8017 44.9797 .— FSJF 3388 , 1, 185 mm SL; Iraq, Shatt al Arab at Basra, 30.539517 47.831181 .— FSJF 3682 , 1, 116 mm SL; TÜrkiye: Diyarbakır prov.: stream Ambar at Ambar , 38.2647 40.4602 .— FSJF 3818 , 2 , 245–290 mm SL; Iraq: Ducan reservoir, 36.1000 44.9000 .— FSJF 3953 , 1 , 67 mm SL; Iraq: Great Zab about 2 km upriver of confluence with Ravanduz , 36.7484 44.2997 .— VPFC Nilzan 1392.9, 33, 218– 284 mm SL; Iran: Ilam prov.: Karkheh at Karkheh reservoir (from market, about 32.503, 48.059) .— VPFC Horalazim 1392.11, 17, 149– 258 mm SL; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market, 31.7183, 47.9769).—88 - SKU, 1, 259 mm SL; Iran: Chaharmahal va Bakhtiary Prov.: Karun (31.632069, 50.641702).—90 - SKU, 1, 243 mm SL; Iran: Chaharmahal va Bakhtiary Prov.: Karun (31.632069, 50.641702).—111 - SKU; Iran: Khuzestan prov.: Shadegan Wetland, 30.854413, 48.683324. ( GenBank accession number: PQ860901) .— VPFC Shadegan 1392.11, 22, 253– 342 mm SL; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381) .— VPFC Karun 4 Dam 1392.11, 17, 209– 384 mm SL ; VPFC Asiaban 1392.09, 1, 299 mm SL; Iran: Khuzestan prov.: Karun at Karun 4 reservoir (from market, about 31.604, 50.477) .— VPFC Shadegan 1393.11, 16, 182– 244 mm SL; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381) .— VPFC Dezful 2023.05., 10, 331– 396 mm SL; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029) .— VPFC Dez 7tape 2023.05., 4, 268– 300 mm SL; Iran: Khuzestan prov.: Dez at Haft Tape (from market, about 32.078, 48.394) .— VPFC DezBazar 2023.05., 5, 285– 310 mm SL; Iran: Khuzestan prov.: Dez (from market) .— VPFC DezShoaybiye 2023.05., 7, 268– 300 mm SL; Iran: Khuzestan prov.: Dez at Shoaybiye (from market) .— VPFC Gotvand 2023.05., 4, 197– 292 mm SL; Iran: Khuzestan prov.: Karun at Gotvand reservoir (from market, about 32.26, 48.941) .— VPFC Sosangerd-HorAlazim 2023.05., 9, 285– 384 mm SL; Iran: Khuzestan prov.: Horalazim wetland, Sosangerd (from market, 31.7183, 47.9769) .— VPFC AhvazBazar 2023.05., 9, 159– 322 mm SL; Khuzestan prov.: Iran: Karun at Karun 4 reservoir (from market, about 31.604, 50.477) .— VPFC Shadegan 2023.05., 9, 187– 306 mm SL; Khuzestan prov.: Iran: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381) .

Luciobarbus schejch x Capoeta damascina .

FSJF 2929, 135 mm SL; Türkiye: Merziman , Euphrates drainage, 37.3248 37.6445 .

Luciobarbus subquincunciatus .

VPFC Gotvand 1393.11, 1, 360 mm SL; Khuzestan prov.: Karun at Gotvand reservoir, 32.3036, 48.9900 .

New material used in molecular genetic analysis

Barbus cyri :

3801; Iran: Sefid River. (GenBank accession number: PQ860909)— 3869; Iran: Urmia Lake basin. ( Genebank accession number: PQ860910)— 3785b; Iran: Aras River . ( GenBank accession number: PQ860911) .

Barbus karunensis :

3846; Iran: Jarrahi River. (GenBank accession number: PQ860912)

.

Barbus lacerta :

3822; Iran: Karkheh River. (GenBank accession number: PQ860913)

.

Luciobarbus esocinus : FSJF-DNA 2627; Iraq: Dukan reservoir, 36.10, 44.90. (GenBank accession numbers: PV053993, PV053991, PV054018, PV054002, PV054014)—FSJF-DNA 1334; Türkiye: Tigris 5 km west of Hasankeyf, 37.7238 41.3605. (GenBank accession numbers: PV054020, PV054027).

Luciobarbus schejch : FSJF-DNA 2247; Iraq: Shatt al Arab at Basra, 30.5395 47.8312. (GenBank accession numbers: PV053998, PV054013). —FSJF-DNA 2426; Iraq: Shatt al Arab at Basra, 30.5395 47.8312. (GenBank accession number: PV053996). —FSJF-DNA 2428; Iraq: Shatt al Arab at Basra, 30.5395 47.8312. (GenBank accession number: PV054034). —FSJF-DNA 2511; Iraq: Shatt al Arab at Basra, 30.5395 47.8312. (GenBank accession number: PV054015) .—FSJF-DNA 2512; Iraq: Shatt al Arab at Basra, 30.5395 47.8312. (GenBank accession number: PV054010). — FSJF DNA-2516; Iraq: Abul Khaseeb, 30.4664, 47.9451. (GenBank accession number: PV054003). —FSJF-DNA 1436; Turkey: Tigris 5 km west of Hasankeyf, 37.7238, 41.3605. (GenBank accession number: PV053992). —FSJF-DNA 1438; Türkiye: stream Merziman south of Yavuzeli, 37.2924 37.7231. (GenBank accession numbers: PV054004, PV054028, PV053997). —FSJF-DNA 2229; Iraq: Lesser Zab at Altun Kopri, 35.7282 44.1230. (GenBank accession number: PV054030). —FSJF-DNA 2237; Iraq: Tabin west of Zarbi, 35.8017 44.9797. (GenBank accession number: PV054007). —FSJF-DNA 2243; Iraq: Dukan reservoir, 36.1000 44.9000. (GenBank accession number: PV054030). —FSJF-DNA 2627; Iraq: Dukan reservoir, 36.10, 44.90. (GenBank accession number: PV053994). —FSJF-DNA 2635; Iraq: Dukan reservoir, 36.1000 44.9000. (GenBank accession number: PV053999). —FSJF-DNA 2649; Iraq: Greater Zab about 2 km upriver of confluence with Ravanduz, 36.7484, 44.2997. (GenBank accession number: PV054025). —FSJF-DNA 2974; Iran: Shadgan wetland, 30.6622 48.5206. (GenBank accession number: PQ860893) —FSJF-DNA 2989; Iran: Karun 3 reservoir south of Shalu Acu bridge, 31.7696 50.1211. (GenBank accession number: PQ860865). —FSJF-DNA 3058; Iran: Sirvan reservoir, 35.1618 46.3367. (GenBank accession number: PQ860846) —FSJF-DNA 3726; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860871). —FSJF-DNA 3727; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860892). — FSJF-DNA 3728; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860875). —FSJF-DNA 3729; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860877). —FSJF-DNA 3730; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860847). —FSJF-DNA 3731; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860853) .—FSJF-DNA 3732; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860878). — FSJF-DNA 3733; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860890). —FSJF-DNA 3735; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860872). —FSJF-DNA 3736; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860882). —FSJF-DNA 3737; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860891). —FSJF-DNA 3738; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860869). — FSJF-DNA 3739; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860885). —FSJF-DNA 3740; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860887). —FSJF-DNA 3741; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860884). —FSJF-DNA 3742; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860889). —FSJF-DNA 3743; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860866). — FSJF-DNA 3744; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860870). —FSJF-DNA 3745; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860895). —FSJF-DNA 3746; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860879). —FSJF-DNA 3747; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860876). —FSJF-DNA 3748; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860849). — FSJF-DNA 3749; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860855) .—FSJF-DNA 3750; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860883). —FSJF-DNA 3751; Iran: Khuzestan prov.: Dez at Dezful (from market, 32.3784, 48.4029). (GenBank accession number: PQ860845). —FSJF-DNA 3752; Iran: prov.: Gotvand in Karun drainage, 32.2165, 48.8225. (GenBank accession number: PQ860852). —FSJF-DNA 3753; Iran: prov.: Gotvand in Karun drainage, 32.2165, 48.8225. (GenBank accession number: PQ860854). —FSJF-DNA 3754; Iran: prov.: Gotvand in Karun drainage, 32.2165, 48.8225. (GenBank accession number: PQ860874) .—FSJF-DNA 3755; Iran: prov.: Gotvand in Karun drainage, 32.2165, 48.8225. (GenBank accession number: PQ860898). —FSJF-DNA 3756; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860903). —FSJF-DNA 3758; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860860). —FSJF-DNA 3759; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860886). —FSJF-DNA 3760; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860908). —FSJF-DNA 3761; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860857). —FSJF-DNA 3762; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860899). —FSJF-DNA 3763; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860858). —FSJF-DNA 3764; Iran: Khuzestan prov.: Horalazim wetland at Bostan (from market) 31.7183, 47.9769. (GenBank accession number: PQ860851). —FSJF-DNA 3765; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860900). —FSJF-DNA 3766; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860907). —FSJF-DNA 3767; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860856). —FSJF-DNA 3768; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860896). —FSJF-DNA 3769; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860868). —FSJF-DNA 3770; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860862). —FSJF-DNA 3771; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860881). —FSJF-DNA 3772; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860867). —FSJF-DNA 3773; Iran: Khuzestan prov.: Bazar Mahi Ahvaz (from market) 31.2371, 48.6537. (GenBank accession number: PQ860848). —FSJF-DNA 3774; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number:). —FSJF-DNA 3775; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860904). —FSJF-DNA 3776; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860902). —FSJF-DNA 3777; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860850). —FSJF-DNA 3778; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860873). —FSJF-DNA 3779; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860888). —FSJF-DNA 3780; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860859). —FSJF-DNA 3781; Iran: Khuzestan prov.: Shadegan wetland at Shadegan (from market, 30.6841, 48.5381. (GenBank accession number: PQ860880). — FSJF-DNA 3883; Iran: Khersaan at Atishgah, 31.2444, 50.9898. (GenBank accession number: PQ860861). —109- SKU; Iran: Khuzestan prov.: Shadegan Wetland, 30.854413, 48.683324. (GenBank accession number: PQ860894). —110- SKU; Iran: Khuzestan prov.: Shadegan Wetland, 30.854413, 48.683324. (GenBank accession number: PQ860897). —111- SKU; Iran: Khuzestan prov.: Shadegan Wetland, 30.854413, 48.683324. (GenBank accession number: PQ860901). —113- SKU; Iran: Khuzestan prov.: Shadegan Wetland, 30.854413, 48.683324. (GenBank accession number: PQ860906). —SMF (not catalogued); Syria: Euphrates. (GenBank accession numbers: PV054022, PV054009) —SMF (not catalogued); Iran: Kheir Abad, 32.5199, 51.5077. (GenBank accession number: PV054021). —ZM-CBSU Lb1015F; Iran: Fars prov.: Rudbal River, Estahban, 28.9713, 54.3905. (GenBank accession number: PV054012). —ZM-CBSU Lb1016F; Iran: Fars prov.: Rudbal River, Estahban, 28.9713, 54.3905. (GenBank accession number: PV053988). —ZM-CBSU Lb1017F; Iran: Fars prov.: Rudbal River, Estahban, 28.9713, 54.3905. (GenBank accession number: PV054031). —ZM-CBSU Lb1018F; Iran: Fars prov.: Rudbal River, Estahban, 28.9713, 54.3905. (GenBank accession number: PV054017). —ZM-CBSU Lb915F; Iran: Kohmareh sorkhi in Helleh drainage, 29.3943, 52.1615. (GenBank accession number: PV054006). —ZM-CBSU Lb916F; Iran: Kohmareh sorkhi in Helleh drainage, 29.3943, 52.1615. (GenBank accession number: PV054005). — ZM-CBSU Lb917F; Iran: Kohmareh sorkhi in Helleh drainage, 29.3943, 52.1615. (GenBank accession number: PV054001). —ZM-CBSU Lb918F; Iran: Kohmareh sorkhi in Helleh drainage, 29.3943, 52.1615. (GenBank accession number: PV054026). —ZM-CBSU Lb949F; Iran: Fars prov.: Kavar in Mond drainage, 29.1820, 52.6924. (GenBank accession number: PV053990). —ZM-CBSU Lb950F; Iran: Fars prov.: Kavar in Mond drainage, 29.1820, 52.6924. (GenBank accession number: PV053989). —ZM-CBSU Lb951F; Iran: Fars prov.: Kavar in Mond drainage, 29.1820, 52.6924. (GenBank accession number: PV054033). —ZM-CBSU Lb952F; Iran: Khozestan prov.: Ali Kalleh in Dez drainage, 32.4032, 48.4072. (GenBank accession number: PV054019). —ZM-CBSU Lb953F; Iran: Khozestan prov.: Ali Kalleh in Dez drainage, 32.4032, 48.4072. (GenBank accession number: PV054008). —ZM-CBSU Lb954F; Iran: Sirvan drainage, 35.1119, 46.2566. (GenBank accession number: PV054011). —ZM-CBSU Lb956F; Iran: Seimare at Simre bridge checkpoint, 33.6805, 47.0560. (GenBank accession number: PV054023). — ZM-CBSU Lb957F; Iran: Seimare at Simre bridge checkpoint, 33.6805, 47.0560. (GenBank accession number: PV053995). —ZM-CBSU Lb958F; Iran: Seimare at Simre bridge checkpoint, 33.6805, 47.0560.(GenBank accession number: PV054035). —ZM-CBSU Lb969F; Iran: Shadegan wetland, 30.6835, 48.5415. (GenBank accession number: PV054032) —ZM-CBSU Lb982F; Iran: Shadegan wetland, 30.6835, 48.5415. (GenBank accession number: PV054029). —ZM-CBSU Lb983F; Iran: Shadegan wetland, 30.6835, 48.5415. (GenBank accession number: PV054016).

Figures

FIGURE 1. Maximum Likelihood (ML) phylogram reconstructed for a common 503 bp segment of COI. Numbers along the branches are bootstrap support values for the ML tree (before slash) and Bayesian posterior probabilities (after slash). Only bootstrap values over 50 and posterior probabilities over 0.95 for branches which we discuss are presented. Red arrows denote the branches to which the bootstrap values and posterior probabilities correspond.Absolute support values of 100/1 are denoted with a star. Red circles represent individuals that align with L. esocinus morphologically. Brown (A1) and grey (A2) vertical bars in Clade A identify two sub-clades with 70% and 53% support values. Each individual is represented by species, drainage/basin, country, and GenBank accession number/BOLD sequence ID (denoted with black circles).

FIGURE 3. Distribution of L. esocinus and L. schejch individuals along the first four principal component axes representing 54.7% of SNP variation: (a) PCI-II, (b) PCI-III, and (c) PCI-IV.The PCA graphs are combined with admixture results. Pie graphs corresponding to panel d (admixture graph) show genomic clusters of each individual. Red pie graphs denote L. esocinus; the red pie graph with a black outline is individual VPFC AhvazBazar 2023.05 / FSJF-DNA 3772 (Fig. 27), which is morphologically in the overlap of meristic counts between L. esocinus and L. schejch (11 total gill rakers, 58 total lateral-line scales) but shares a similar genome with L. esocinus; blue pie graphs represent L. schejch. A and B denote the mtDNA clade of each individual.

FIGURE 5. Barbus rajanorum, NMW 54494, holotype, 190 mm SL; Syria: Aleppo. Modified from Heckel (1943).

FIGURE 6. Barbus rajanorum, NMW 54494, holotype, 190 mm SL; Syria: Aleppo.

FIGURE 7. FSJF 2929, 135 mm SL; hybrid between Luciobarbus schejch (male) and Capoeta damascina (female), with mtDNA of C. damascina; Türkiye: Euphrates drainage.

FIGURE 8. Luciobarbus schejch, syntype, and an individual identified as L. mystaceus; Iraq: Mosul. Modified from Heckel (1943).

FIGURE 9. Luciobarbus schejch, identified as L. mystaceus by Heckel (1843), NMW 54384/1, 233 mm; 54382/2, 260 mm SL; Iraq: Mosul.

FIGURE 10. Luciobarbus schejch, identified as L. mystaceus by Heckel (1843), NMW 54385/1, 157 mm; 54385/2, 142 mm SL; Syria: Aleppo.

FIGURE 11. NMW 6596, Barbus barbulus, syntype, about 125 mm SL; Iran: Qarah Aqaj, Mond drainage.

FIGURE 12. Luciobarbus schejch, syntype, NMW 50399, 138 mm SL; Iraq: Mosul.

FIGURE 13. Luciobarbus schejch, syntypes, NMW 54520, 269 mm SL; NMW 54520, 275 mm SL; Iraq: Mosul.

FIGURE 14. Luciobarbus schejch, from top: FSJF 3682, 116 mm SL; Türkiye: stream Ambar; FSJF 2931, individual with mtDNA within COI clade A, 121 mm SL; Türkiye: stream Merziman; FSJF 3387, 143 mm SL, Iraq: Tabin; FSJF 2861, 195 mm, individual with mtDNA in COI clade A; Türkiye: Tigris at Hasankeyf; not preserved; 196 mm SL; Türkiye: Murat, 38.8681, 40.7926. No molecular data is available if not mentioned.

FIGURE 15. Luciobarbus schejch, VPFC Horalazim 1396.11; 331 mm SL; 229 mm SL; Iran: Hor Alazim wetland.

FIGURE 16. Luciobarbus schejch, from top, left column: AJRPC Kut 1397.8; 259 mm SL; AJRPC Kut 1397.8; 201 mm; Iraq: Tigris at Kut; VPFC Shoshtar 1396.11; 310 mm SL, Iran: Karun. Right column: AJRPC Kut 1397.8; 178 mm SL, Iraq: Tigris at Kut; VPFC Asiaban 1396.11; 298 mm SL; Iran: Karkheh; VPFC Keshavarz 1396.11; about 400 mm SL; Iran: Dez. No molecular data is available for these fish.

FIGURE 17. Barbus kersin, NMW 54215, syntype, 163 mm SL; Syria: Aleppo. Modified from Heckel (1943).

FIGURE 18. Luciobarbus kersin, from top: ZMB 3237, syntype, 143 mm SL; NMW 54215, syntype, 163 mm SL; Syria: Aleppo.

FIGURE 19. Luciobarbus pectoralis, not preserved, about 100 mm SL; Syria: Orontes at Mashr’a al Bouz, 35.9508 36.3958; FSJF 2329; 160 mm SL; Türkiye: Seyhan drainage; FSJF 2697, 190 mm SL; Syria: Qattina Lake.

FIGURE 20. Luciobarbus schejch, VPFC Sosangerd-Horalazim 2023.05 (FSJF-DNA 3769, COI clade A), 226 mm SL; Individual superficially similar to L. kersin; Iran: Fish market in Ahwaz, lower Karun.

FIGURE 21. Luciobarbus schejch, VPFC Horalazim 1392.11, 252 mm SL; individual similar to fish identified as L. kersin by Khaefi et al. (2018); Iran: Horalazim wetland. No molecular data is available.

FIGURE 22. Luciobarbus xanthopterus, NMW 54786, 340 mm SL; individual with snout length 1.9 times in postorbital length; Luciobarbus esocinus, NMW 54092, 321 mm SL; Iraq: Mosul. Modified from Heckel (1943: Plate IV).

FIGURE 23. Luciobarbus xanthopterus, NMW 54841/1, 216 mm SL; snout length 1.8 times in postorbital length; NMW 54786, lectotype, 340 mm SL; snout length 1.9 times in postorbital length Iraq: Mosul.

FIGURE 25. Luciobarbus schejch with meristic counts in the overlapping range of L. esocinus and L. schejch. VPFC Dez7tape 2023.05; 357 mmm SL; 12 gill rakers, 59 total lateral-line scales, snout/postorbital 2.1; Iran: Karun; VPFC AhvazBazar 2023.05; 250 mmm SL; 14 gill rakers, 60 total lateral-line scales, snout/postorbital 1.8; Iran: Karun; VPFC AhvazBazar 2023.05¸ 211 mmm SL; 14 gill rakers, 61 total lateral-line scales, snout/postorbital ratio 2.4; Iran: Karun.

FIGURE 26. Potential hybrids between L. esocinus and L. schejch, not preserved and not sequenced. From top: about 600 mm SL, 60 total lateral-line scales; Iraq: Dukan reservoir; about 500 mm SL, 55 total lateral-line scales; Iraq: Bagdad; photo by Omar F. Al-Sheikhly; about 450 mm SL, 60 total lateral-line scales; Iraq: Shatt-al Arab at Basra; photo by Atheer Ali.

FIGURE 27. Hybrid between L. esocinus and L. schejch, VPFC AhvazBazar 2023.05 (FSJF-DNA 3772), 257 mm SL; individual with mtDNA in COI clade A; Iran: Karun. SNP data place this individual close to, but not within, L. esocinus (compare with Fig. 28). Individual with 11 gill rakers, 58 total lateral-line scales, and snout length 1.7 times in postorbital length; Iran: Fish market in Ahwaz, lower Karun.

FIGURE 28. Luciobarbus esocinus, FSJF 3819, juvenile, 220 mm SL; Iraq: Ducan Reservoir.

FIGURE 29. Luciobarbus esocinus, VPFC Shadegan 1392.11, 285 mm SL; Iran: Shadegan wetland. Individual with 59 total lateral-line scales and 9 gill rakers. No molecular data.

FIGURE 30. Luciobarbus esocinus, VPFC AhvazBazar 2023.05 (FSJF-DNA 3773, COI clade A), 207 mm SL; individual with 9 gill rakers, 65 total lateral-line scales, and snout length 2.0 times in postorbital length; Iran: Fish market in Ahwaz, lower Karun.

FIGURE 31. Luciobarbus schejch, individuals with mtDNA in COI clade B and “pure” genome-wide SNP data of L. schejch, from top, VPFC Sosangerd-Horalazim 2023.05 (FSJF-DNA 3756), 375 mm SL; 15 total gill rakers and 57 total lateral-line scales; Iran: Karkheh; VPFC Gotvand 2023.05 (FSJF-DNA 3755), 326 mm SL; 12 total gill rakers and 57 total lateral-line scales; Iran: Karun; VPFC DezShoaybiye 2023.05 (FSJF-DNA 3745), 426 mm SL; 15 total gill rakers and 55 total lateral-line scales; Iran: Karun.

FIGURE 32. Luciobarbus schejch, individuals with mtDNA in COI clade A, from top, left column: VPFC Shadegan 2023.05 (FSJF-DNA 3780), 257 mm SL; Iran: Jarrahi; VPFC Sosangerd-Horalazim 2023.05 (FSJF-DNA 3761), 301 mm SL; Iran: Karkheh. Right column: VPFC AhvazBazar 2023.05 (FSJF-DNA 3750), 384 mm SL; Iran: Karun; VPFC Dez7tape 2023.05 (FSJF-DNA 3736), 365 mm SL; Iran: Karun.

FIGURE 33. Luciobarbus schejch, individuals with mtDNA in COI clade B, from top, left column:VPFC Sosangerd-Horalazim 2023.05 (FSJF-DNA 3760), 285 mm SL; Iran: Karkheh; VPFC Shadegan 2023.05 (FSJF-DNA 3774), 351 mm SL; Iran: Jarrahi. Right column: VPFC AhvazBazar 2023.05 (FSJF-DNA 3768), 174 mm SL; Iran: Karun; VPFC Shadegan 2023.05 (FSJF-DNA 3776), 217 mm SL; Iran: Jarrahi.

FIGURE 34. Luciobarbus schejch, juvenile individuals with mtDNA in COI clade A. FSJF 3953, 67 mm SL; Iraq: Great Zab; FSJF 3382, 104 mm SL; Iraq: Lesser Zab.

FIGURE 35. Luciobarbus schejch, voucher of FSJF-DNA 2243, not preserved, individual with mtDNA in COI clade A; about 400 mm SL; Iraq: Ducan reservoir.

FIGURE 36. Luciobarbus schejch, FSJF 3818, 290 mm SL; individual with mtDNA in COI clade B; Iraq: Ducan reservoir.