Sample collection

A total of 36 loach specimens were collected from the Naka-ikemi Wetland (35°39'34.8” N, 136°05'18.6” E, altitude 50 m) in the River Shono system (Fig. 1, Table 1). An additional 18 specimens from the Ikenokochi Wetland (35°39'11.5” N, 136°08'00.6” E, altitude 300 m) located on the upper part of the same river system were collected to eliminate effects of phenotypic plasticity caused by environmental differences. Our preliminary genetic analysis using mtDNA cytochrome b (cytb) sequencing revealed that both populations belonged to the same local genetic group (R. Okada et al., unpublished data). Only specimens with a standard length (SL) of 70 mm and over were used to account for secondary sexual development. Fish were euthanised using benzoic acid, following which the right pelvic fin was cut and preserved in 99.5% ethanol for genetic species identification (Type I or Type II loach). The rest of the body was fixed in 10% formalin and preserved in 70% ethanol.

Table 1.

Counts and measurements of two loach species.

continued

Genetic species identification

Total DNA was extracted from ethanol-preserved fin tissues following the procedure described by Asahida et al. (1996). Species identification for all specimens was conducted via polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analyses of the cytb region in the mtDNA and the ribonuclease subunit H2 gene (RNaseh2) region in the nuclear gene according to the method previously reported by Okada et al. (2017). Additionally, RFLP analysis of nuclear recombination activating gene 1 (RAG1) was conducted for all specimens to identify particular species. PCR amplification was performed using the primer pair RAG-1F (5′-AGC TGT AGT CAG TAY CAC AAR ATG-3′) and RAG-9R (5′-GTG TAG AGC CAG TGR TGY TT-3′) (Quenouille et al. 2004). PCR conditions were 2 min at 95 °C, 30 cycles of 30 s at 95 °C, 30 s at 54 °C, 1 min at 72 °C and 5 min at 72 °C. The PCR products were then digested with the restriction enzyme PstI (TaKaRa, Kyoto, Japan) selected based on sequence information reported by Šlechtová et al. (2008). Digestion was performed following the manufacturer's instructions. The expected RFLP fragments for the PstI digestion in each species were ca. 511 and 474 bp for the Type I loach and 474, 316, and 195 bp for the Type II loach. The restriction fragments were subsequently separated by horizontal electrophoresis on 3% agarose gel.

Morphological analysis

Meristic methods followed Kottelat & Freyhof (2007), and morphometric methods generally followed Kottelat (1984), except for the following characteristics: dorsal-fin base and anal-fin base length were measured according to Kottelat & Freyhof (2007); the anal-fin length was measured as the distance between the base and the tip of the unbranched ray, even if any branched ray was longer; the pectoral-fin length was measured both as the length of the longest unbranched ray and as the 1^st branched ray; the caudal peduncle depth was measured both at the narrowest part of the caudal peduncle and the caudal-fin base. Vertebrae, including the four components associated with the Weberian apparatus, were counted from soft x-ray photos. The sex of each specimen was determined by the presence of the lamina circularis at the base of the 1^st unbranched ray of the pectoral fin that only male has. To observe the features of the lamina circularis, the right pectoral fin was placed in 4% KOH for 24 h to make it transparent and then stained with Alizarin Red S + 1% KOH for 24 h, according to the method described by Nakajima (2012). The description of the lamina circularis follows Nakajima & Uchiyama (2017) and Hata et al. (2020). In male specimens, the distance between the dorsal fin origin and the top of the raised ridge located posterior to the dorsal fin was measured as the position of the raised ridge (PRR). All measurements were made using a digital calliper and recorded to 0.1 mm. Differences in meristic and morphometric characteristics between the two species or the sexes were tested using Welch´s t-test. Differences in characteristics between the same species and sex specimens from two different sites were also tested using Welch´s t-test, except for Type I males, where only one individual was included from the Ikenokochi Wetland. Statistical analyses were conducted using R v. 3.5.2 (R Core Team 2018).

Fig. 2.

Fresh specimens of Type I and Type II loaches from Naka-ikemi Wetland. (a) Type I loach, KUN-P 59954, 92.7 mm standard length (SL), male; (b) Type I loach, KUN-P 59917, 104.6 mm SL, female; (c) Type II loach, KUN-P 59924, 96.3 mm SL, male; (d) Type II loach, KUN-P 59928, 86.8 mm SL, female.

Materials examined

The information on each specimen listed below is shown in Table S1.

Male Type I loach (n = 13). KUN-P59948-59950, 59952-59954, 64124, 64128-64130, 64826, 64827, 78.0-109.2 mm in standard length (SL), Naka-ikemi Wetland; KUN-P60028, 104.4 mm SL, Ikenokochi Wetland.

Female Type I loach (n = 16). KUN-P59917, 59947, 59951, 59955, 64126, 64127, 64131-64133, 64828, 64829, 78.7-130.3 mm SL, Naka-ikemi Wetland; KUN-P60023-60027, 118.7-126.8 mm SL, Ikenokochi Wetland.

Male Type II loach (n = 11). KUN-P59918, 59924-59927, 59929, 59930, 73.9-99.9 mm SL, Naka-ikemi Wetland; KUN-P60018-60021, 80.2-90.6 mm SL, Ikenokochi Wetland.

Female Type II loach (n = 14). KUN-P59919-59923, 59928, 86.8-113.0 mm SL, Naka-ikemi Wetland; KUN-P60010-60012, 60014-60017, 70.7-139.1 mm SL, Ikenokochi Wetland.

Table 2.

Results of analysis of Welch´s t-test for meristic and morphometric characters of mature two loach species.

continued.

Fig. 3.

Relationships between (a) body width (at anal origin), (b) body depth, (c) caudal-peduncle depth (at caudal-fin base), (d) dorsal-fin length, (e) dorsal-fin base length, (f) pectoral-fin length (at unbranched ray), (g) pectoral-fin length (at 1^st branched ray), (h) anal-fin length, (i) anal-fin base length, (j) pelvic-fin length, (k) caudal-fin length (as % of standard length), (l) rostral barbel length, (m) maxillary barbel length and (n) mandibular barbel length (as % of head length) to standard length and head length in Type I loach (male: red squares, female: red circles) and Type II loach (male: blue squares, female: blue circles).

Genetic species identification

All RFLP banding patterns of the mtDNA (cytb) and two nuclear gene regions (RNaseh2 and RAG1) concord with the expected results. All examined specimens were successively identified as Type I loaches (13 males and 16 females) and Type II loaches (11 males and 14 females) (Table 1). Only one Type I male was found among the specimens collected from the Ikenokochi Wetland. Hybrids of the two species were not detected.

Sexual dimorphisms of two loach species

Meristic and morphometric data of Type I and Type II loaches and the results of the statistical analyses are shown in Table 1. Of the 32 characters, significant (P < 0.05) differences were observed for 13 and 15 characters between the males and females in Type I and Type II loaches, respectively (Welch's t-test, Table 2). In both species, the dorsal and pectoral fin length (at the unbranched and 1^st branched rays), anal-fin length, anal-fin base length, pelvic- and caudal-fin length, and the rostral barbel length were longer in males than in females and commonly showed significant (P < 0.01) differences (Fig. 2). The caudal peduncle depth of the males (at the caudal-fin base) was significantly (P < 0.05) greater than that of females. Among these characteristics, significant (P < 0.05) differences in the dorsal- and pectoral-fin length (at the unbranched and 1^st branched rays) and anal-fin base length were observed between specimens of the same species from the two sites (Welch's t-test, Table 1, Table S1). In Type I loaches, the body width (at the anal origin), body depth, and dorsal-fin base length differed significantly (P < 0.01) between the sexes. In the Type II loaches, the maxillary and mandibular barbel length differed significantly (P < 0.01) between the sexes. Males of both species have raised ridges on both sides of the body, posterior to the dorsal fin. Among these characteristics, body width (at the anal origin) and depth showed significant (P < 0.05) differences between Type I loach specimens from two sites (Welch's t-test; Table 1, Table S1).

Morphological differences between two loach species

A variety of body colours and spots were observed among individuals of both species, and both Type I and Type II loaches commonly have a black spot on the caudal peduncle base (Fig. 3). Significant (P < 0.05) differences for 13 characters in male loaches and 15 characters in female loaches were observed between the two loach species (Welch's t-test, Table 2). Type I loaches have more pectoral-fin branched rays and total vertebrae than Type II loaches in both sexes (Tables 2-3). The caudal peduncle depth (at the caudal fin base), dorsal-fin length, dorsal-fin base length, pectoral-fin length (at the unbranched ray), and pelvic-fin length were significantly (P < 0.01) higher in Type II loaches (Fig. 2); body depth and anal-fin length were also significantly (P < 0.05) higher (Table 2). Among these characteristics, the dorsal-fin length, body depth, and pectoral fin length (at the unbranched ray) differed significantly (P < 0.05) between specimens of the same species from the two sites (Welch's t-test, Tables 1, Table S1). In contrast, the head length and depth, body width (at both the dorsal and anal origin), pectoral-fin length (at the 1^st branched ray), and rostral, maxillary, and mandibular barbel length differed significantly (P < 0.05) in only one sex. Among these characteristics, body width (at the dorsal and anal origin) in males and pectoral-fin length (at the 1^st branched ray) in females differed significantly (P < 0.05) between specimens of the same species from the two sites (Welch's t-test, Table 1, Table S1). The lamina circularis varies considerably in males, with intermediate shapes and sizes. It tends to be rice-scoop shaped in Type I loaches and pole-axe shaped in Type II loaches and is smaller in Type I than in Type II loaches (Fig. 4). All male Type I loaches had one pair of raised ridges on both body sides posterior to the dorsal fin (Fig. 5a). All mature Type II loaches had an additional keel-like ridge that extended along the body axis on both sides anterior to the dorsal fin (Fig. 5b). PRR was larger in Type I than in Type II loaches (Table 2, Fig. 6). Most adult Type II loach females (91.7%) had spawning scars above the pelvic fin (Fig. 3d), irrespective of the population. These scars were absent in Type I loach females (Fig. 3b).

Fig. 4.

Dorsal view of preserved specimens of male (a, b) Type I and (c, d) Type II loaches from Naka-ikemi Wetland. Arrows indicate raised ridges. (a) KUN-P 59954, 92.7 mm standard length (SL); (b) KUN-P 59952, 99.7 mm SL; (c) KUN-P 59918, 99.9mm SL; (d) KUN-P 59924, 96.3 mm SL.

Table 3.

Frequency of total vertebrae counts of two loach species.

Fig. 5.

Dorsal views of pectoral-fin of mature males stained with Alizarin Red S. (a) Type I loach, KUN-P 60028, 104.4 mm standard length (SL), Ikenokochi Wetland; (b) Type II loach, KUN-P 60020, 83.1 mm SL, Ikenokochi Wetland.

Fig. 6.

Relationships between the top of the raised ridge below the dorsal fin when the body side is the base and the dorsal fin origin (as % of standard length) to standard length in male Type I loach (red squares) and male Type II loach (blue squares).

Sexual dimorphism and morphological differences between the two loach species

Sexual dimorphism in the Japanese loach has been reported in previous studies before the two biological species (Type I and Type II loaches) were recognised (Okamoto 1921, Ikeda 1936, Tsukahara 1948, Kobayashi 1951, Kubota & Matsui 1955, Kubota 1961). This observation suggests that the sexual dimorphisms of the two species were recognised equivocally. Our comparison of the two genetically distinct species in a single river system revealed morphological differences and sexual dimorphisms between the two loach species.

Kubota & Matsui (1955) reported the sexually dimorphic morphological traits of the loach M. anguillicaudatus from Shimonoseki, Yamaguchi Prefecture, Japan. This population was identified as Type II loaches based on the distribution area (Morishima et al. 2008, Koizumi et al. 2009, Fujimoto et al. 2017). The males have longer pectoral and dorsal fins, dorsal fin bases, pelvic, anal and caudal fins, barbels (rostral, maxillary, and mandibular barbel), and greater caudal peduncle depth than females. Females have a wider body width than males, and the dorsal fin of males is located more anteriorly than females. In this study, Type II loaches showed sexual dimorphisms similar to those previously reported. Furthermore, the following characteristics were observed: Type II male loaches have a wider body width at the dorsal fin origin, longer caudal peduncle, and anterior anal fin; however, no sexual dimorphism in the dorsal-fin base length was observed in this study, consistent with the report by Kubota & Matsui (1955). These differences may be attributed to genetic, geographical, or environmental differences among the local populations of Type II loaches.

The elongation of the male pectoral fin is one of the most common sexual dimorphisms in Cobitidae (e.g. Vladykov 1935, Šlechtová et al. 2008), and noticeable sexual dimorphism was observed in this species, even when environmental factors were considered. Kubota & Matsui (1955) also reported that males of this species have elongated pectoral fins. The male pectoral-fin length (at the unbranched and 1^st branched rays) differed significantly among Type II loach specimens from the two sites despite genetic homogeneity and geographic proximity (Table 1) and without range overlap between males and females (% SL 11.5-13.2 in males and 6.8-9.8 in females), and is considered a sexually dimorphic trait in this species.

In Type I loaches, males showed longer pectoral, dorsal, pelvic, anal, and caudal fins and rostral barbel than females, similar to the Type II loaches in this study. Additionally, males had a longer dorsal fin base, greater body depth, and wider body width at the anal fin origin than females, and the male anus was located more posteriorly than females. Although only one individual was investigated from the Ikenokochi population, the measurement of these characteristics was within the range of the Naka-ikemi population. Among the sexually dimorphic characters in Type I loaches, body width (at the anal origin), body depth, pectoral-fin length (at the unbranched and 1^st branched rays), and anal-fin base length differed significantly in the female specimens from the two sites (Table 1). Of these, differences in body width and depth should not be regarded as sexual dimorphisms, as they generally vary depending on environmental factors (e.g. food availability, water temperature, and collection season). Pectoral fin length (at the unbranched and 1^st branched rays) showed significant differences without range overlap between males and females (% SL 9.2-12.5 in males and 5.9-8.8 in females), which is also thought to be a sexual dimorphism of this species.

Based on an examination of a population from Kyushu Island (Fukuoka Prefecture), Tsukahara (1948) reported that males had a well-developed lamina circularis at the base of the 1^st branched ray of the pectoral fin and two pairs of raised ridges on the body side and similar depressions. The specimen was identified as a Type II loach. In this study, although all Type II loach males had two pairs of raised ridges on the body side, all Type I loaches lacked anterior keel-like ridges and only had one raised ridge on the posterior side of the dorsal fin. The PRR was also distinctly different between the two loach species. In this study, the lamina circularis was more developed in Type II than in Type I loaches. Tsukahara (1948) also reported that most adult females have a small depression on the body of the upper part of the pelvic fin (referred to as spawning scars). Spawning scars were observed in the wild captured Type II loach females regardless of the season but not in the female Type I loach (Fig. S1). As the spawning scars remained in captivity for more than one year (Y. Yashima, unpublished data), the presence of spawning scars may be regarded as a lifelong mark in female Type II loaches that had spawned previously.

Tsukahara (1948) reported that Type II loach males squeeze eggs from the female body by winding their body around that of the female during spawning. The developed lamina circularis on the pectoral fin and two ridges on the body side fix and squeeze the female´s body during spawning. The lamina circularis of the male creates spawning scars on the female's body during this process. The smaller size and degenerated shape of the lamina circularis, lack of anterior keel-like ridges, longer PRR in males, and absence of spawning scars in females suggest that the spawning behaviour of the Type I loach may differ from that of the Type II loach, and may be among the factors responsible for the reproductive isolation between the two species in Naka-ikemi Wetland. Direct observation of Type I loach spawning behaviour is important to examine whether the sexual dimorphisms observed in the two loach species act as a reproductive barrier between them.

Morphological identification and taxonomic problems of the two loach species

Although a difference in the total number of vertebrae was reported between the two loach species in the Naka-ikemi Wetland (P < 0.01, Welch's t-test; Okada et al. 2017), no useful external characteristics were known for distinguishing them. In our observation, distinguishing the two species based on colouration was difficult because of the considerable variation in colour in both species. Other external morphological characteristics, regardless of site, were observed between the two species. In the morphological analysis, Type I loaches could be distinguished from Type II loaches by pectoral fin branched rays (8-10 (mode 9) vs. 7-9 (mode 8)), caudal-peduncle depth, and dorsal-fin, dorsal-fin base, pectoral-fin (at the unbranched rays), and pelvic-fin lengths, which were shorter than those of Type II loaches of the same sex. Some characteristics showed some overlap between Type I and Type II loaches. Therefore, identification of the sex of each individual is necessary for morphological species identification. Male Type II loaches have a keel-like ridge on both body sides anterior to the dorsal fin in addition to the ridge on the posterior side, and most female Type II loaches have spawning scars, which are absent in Type I loaches. Type II loaches have been reported to have hybridised and genetically introgressed with an unknown member of the genus Cobitis in the past (7-10 million years ago; Kitagawa et al. 2011, Okada et al. 2020a). The contribution of genomic introgression from hybrid species to these morphological differences must be addressed. Further comparisons of reproductive dimorphism with a wider range of species, including the genus Cobitis, are required.

In this study, common differences between the two species were observed between the specimens from the two sites at different altitudes, indicating that the traits are stable in this river system (Table 1). However, the morphological differences observed in this study may only be common to some populations of the two loach species across Japan. Nakajima & Uchiyama (2017) described the morphological characteristics of populations from Lake Tofutsu in Hokkaido with Type I mtDNA as Misgurnus sp. (Clade A) sensu Nakajima & Uchiyama (2017) and reported that the shape of the lamina circularis could be a diagnostic characteristic of Misgurnus sp. (Clade A) and Type II loaches. However, morphological differences in the lamina circularis between the two species were not as apparent as previously reported, and considerable variation was observed between the two species. Recently, Shedko & Vasil'eva (2022) described a new species of the genus Misgurnus loach with Type I mtDNA as M. chipisaniensis from Sakhalin Island without examining specimens from Japan, including Misgurnus sp. Type I sensu Okada et al. (2017) and Misgurnus sp. (Clade A) sensu Nakajima & Uchiyama (2017). Recent genetic studies have revealed that highly genetically diverged local populations are included within Type I loaches (R. Okada et al., unpublished data). Taxonomic examination of Type I loaches from a wider area, including M. chipisaniensis, Misgurnus sp. (Clade A) sensu Nakajima & Uchiyama (2017), is therefore necessary.

Type II loaches are also distributed in East Asia. It has been reported that Japanese Type II loach populations are monophyletic and morphologically distinguishable from the continental populations (Matsui & Nakajima 2020). Zhang et al. (2021) assigned scientific names to each of these two populations, but the taxonomic identity of the genus Misgurnus, including the type specimens from Japan, needs to be re-examined.