Introduction

The Muraenidae (moray eels: Anguilliformes) includes about 200 species in 16 genera distributed throughout temperate and tropical waters of the Indo-Pacific and Atlantic oceans (Nelson et al. 2016). Muraenid species typically have elongated bodies, lack pelvic and pectoral fins, and inhabit cavities in rocky and coral reefs. Most species have large mouths with long, sharp canine-like teeth (Randall et al. 1990). Moray eels are resident top predators in coastal rocky or coral reefs (Hixon and Beets 1993), and although many species exist, knowledge of their reproductive biology is generally limited because of their cryptic habits (Matić-Skoko et al. 2011).

Spawning season and maturation of Gymnothorax equatorialis (Lucano-Ramírez et al. 2008), Gymnothorax javanicus (Brock 1972), Gymnothorax minor (Oomori et al. 2022b) and Muraena helena (Matić-Skoko et al. 2011; Sallami and Ben Ibrahim 2021) have been studied by monthly sampling and histological observations on gonads. Spawning behaviors of Gymnothorax pictus and Gymnothorax thyrsoideus described from aquarium studies reveal that the male grips the female’s snout with its jaw, and they both then swim toward the surface to spawn (Loh and Chen 2018). The information on age and growth of moray eels is limited to M. helena, which is estimated to live for 12 years (y) (Matić-Skoko et al. 2011) or 10 y (Sallami et al. 2016).

The Kidako moray Gymnothorax kidako occurs in the temperate and subtropical waters of Japan, the southern part of the Korean Peninsula, and Taiwan (Hatooka 2013). In Japan, records of this species being eaten date to the 18th century (Toyama et al. 2003), and in local cities of southern Japan it continues to be consumed as a local dish (Yoshimoto 2017; Torii and Kato 2019). Despite this, other than its spawning behavior (Moyer and Zaiser 1982), the ecology of this species is poorly known. Accordingly, we report on the reproductive ecology, age structure, and growth rates of G. kidako in Tateyama, Chiba, Japan.

Materials and methods

Sample collection. Specimens were sampled from Banda, Tateyama Bay, Chiba, Japan (34°98' N, 139°77' E), where the Kuroshio Current flows offshore. Mean monthly water temperature, recorded at Tateyama Station, Tokyo University of Marine Science and Technology at Banda, ranged 13.7 °C (February) to 26.1 °C (August) (Fig. 1a).

Fig. 1
figure 1

Water temperatures (°C) (a), and average gonadosomatic index (GSI) for Gymnothorax kidako males (b) and females (c) from January to December 2015. Vertical bars standard deviation

Full size image

Specimens were caught in the second half of each month from January to December 2015 using tube pot fishing (plastic pipes of 75-cm length, 13-cm diameter, with one end closed) placed on the seabed in shallow (2 m) water at 1800 hours. Traps were left for 12–24 h. Using this technique, 198 Gymnothorax kidako were caught (Table 1). Specimens were then transported to the laboratory at Tateyama Station and euthanized in ice seawater; their total lengths (TL) were measured to the nearest 0.1 cm and body weights (BW) to 0.1 g. Gonad tissues and the right sagittal otolith of each specimen were then removed. Sexes were initially differentiated based on reproductive tissue appearance: male gonads are smooth surfaced, and female gonads have complex folds (Fig. 2). Subsequent histological observations confirmed the accuracy of this method.

Table 1 Numbers of male and female and frequency of mature female (%) in which tertiary vitellogenic oocytes were detected in Gymnothorax kidako collected in 2015. Numbers in parentheses indicate numbers of specimens from which otoliths were successfully analyzed
Full size table
Fig. 2
figure 2

Testis (a) and ovary (b) tissues of Gymnothorax kidako. Scale bar 1 cm

Full size image

All specimens caught in 2015 exceeded 47 cm TL. An additional 12 specimens caught by hand net on 9 and 10 May 2022 at Banda were smaller than 47 cm TL. They were sexed by the above method.

Gonad histology. Gonad weight (GW) was determined to the nearest 0.1 g, and a gonadosomatic index (GSI) was calculated by GSI = GW × 100/BW. Extracted gonads were sequentially fixed in Bouin’s solution for 24 h, preserved in 70% ethanol, dehydrated, embedded in paraffin, sectioned at 5-µm, stained with hematoxylin and eosin, and observed by optical microscope (BX51, Olympus, Tokyo). Terminology for developmental stages of spermatogenesis and oogenesis follows Brown-Peterson et al. (2011). If an ovary contained tertiary vitellogenic oocytes, we regarded it as being mature.

Collecting pairs. At depths of 14–25 m off Banda in 2015, 2016 and 2018, eight pairs of moray eels were collected from cave habitat (Table 2). For six of these pairs (#1–6), sex was determined by the methods as mentioned above. Two pairs (#7 and 8) were deposited in collections of the Kanagawa Prefectural Museum of Natural History (KPM-NI). Gonad condition was assessed by naked eye following dissection of two specimens with swollen abdomens, and by histology for any other.

Table 2 Sizes, sexes, and collection dates in the pairs of Gymnothorax kidako
Full size table

Age determination and growth estimation. Because increment counts have been validated for Muraena helena (Matić-Skoko et al. 2011; Sallami et al. 2016), we estimated age by counting otolith increments. Following the methods of Shimizu et al. (2022), otoliths were sequentially washed in distilled water, dried, embedded in epoxy resin, sectioned transversely by grinder (ML-1100NT, Maruto Instrument Co., Tokyo), and sections were then mounted on glass slides and coated with commercial nail enamel. Because opaque and translucent zones alternated (Fig. 3), we considered opaque zones to represent growth bands, and counted them at × 100 using an optical microscope. Of the 198 specimens collected in 2015, bands were successfully counted on the otoliths of 53 males and 76 females (Table 1), and on 5 males and 7 females collected in 2022.

Fig. 3
figure 3

a Transverse section of Gymnothorax kidako otolith from 34-year-old individual; b overall view of the same otolith. Arrows opaque zones, solid circle core, scale bars 500 µm

Full size image

To determine when opaque zones formed, we calculated the marginal growth rate (MGR) using:

$$\mathrm{MGR }= (R - {r}_{max}) / ({r}_{max} - {r}_{max-1}),$$

where R represents the distance from the otolith core to its edge, rmax the distance from the core to the distal edge of the last completed opaque zone, and rmax–1 the distance from the core to the distal edge of the second outermost opaque zone (Fig. 3). The von Bertalanffy growth formula was fitted to estimate the relationship between TL and age in G. kidako using the non-linear least-squares method (Gorie 2001).

$${L}_{t}=L\infty \left\{1 -\mathrm{ exp }\left[-k \left(t-{t}_{0}\right)\right]\right\},$$

where Lt is TL at age t, and L∞, k, and t0 are the asymptotic length, growth coefficient, and hypothetical age at TL = 0, respectively. Parameters were estimated using solver programs in MS-Excel (Gorie 2001).

Results

Sex ratio and total length. Of the 198 specimens collected during monthly sampling in 2015, 88 were male and 110 were female (Table 1); 12 additional specimens collected in 2022 comprised 5 males and 7 females. In total, 93 males and 117 females were collected. There were no significant differences in sex ratio (binomial test, P > 0.05), and between the sizes of males (mean ± SD = 76.2 ± 16.6 cm TL, range = 26.7–102.5 cm TL) and females (75.7 ± 13.0 cm TL, 32.1–97.5 cm TL) (Welch’s t-test, t = 0.24, d.f. = 172, P > 0.05).

GSI and sexual maturation. There were no major changes in male GSI throughout the year, although there was a slight increase in June (Fig. 1b). In contrast, female GSI was low from January to May, slowly increased through June, rapidly increased and peaked in July, then decreased through September and remained low to December (Fig. 1c).

All year around, spermatogonia, spermatocytes, and sperm were detected, indicating the occurrence of spermatogenesis (Fig. 4a–c). Ovaries from January to May were full of primary growth and cortical alveolar oocytes (Fig. 4d); in June, most oocytes were at a primary vitellogenic stage (Fig. 4e); in July, in all females, tertiary vitellogenic oocytes appeared, while primary growth oocytes were also detected (Fig. 4f; Table 1). In August, tertiary vitellogenic oocytes were detected in 70% of the females, while oocytes remained at primary vitellogenic or cortical alveolar stages in 30% (Table 1). In September, primary growth and cortical alveolar oocytes were detected; from October to December, only primary growth oocytes were found in the ovary.

Fig. 4
figure 4

Structure of Gymnothorax kidako testes in January (a), August (b) and December (c), and ovaries in February (d), June (e), and July (f). CA cortical alveolar, PG primary growth, S sperm, SC spermatocyte, SG spermatogonium, Vtg1 primary vitellogenic oocytes, Vtg3 tertiary vitellogenic oocytes. Scale bars 100 µm

Full size image

The smallest mature male (47.4 cm TL) and female (62.1 cm TL) (age could not be determined from extracted otoliths), were collected in June and July 2015, respectively.

Pairs. All moray eels collected in pairs comprised one male and one female, with the male being the larger in all eight pairs (Table 2). In pairs #7 and 8, smaller individuals with swollen abdomens were fully mature females, with ovulated oocytes spilling through the abdominal incision (Fig. 5); sperm ducts of males in these pairs were also filled with sperm.

Fig. 5
figure 5

Female Gymnothorax kidako (90.3 cm TL: KPM-NI 66088) with swollen abdomen after dissection. Photo (KPM-NR 228772B) by N. Oomori. Arrows ovulated oocytes from the abdomen

Full size image

Age and growth. There was no rapid change in MGR from January to July, but MGR decreased steeply from July to August before gradually increasing to December (Fig. 6). The results indicated that a single annulus had been formed in August. Both male and female ages ranged 3–34 years. Male, female and sex-combined von Bertalanffy growth formulae are expressed as:

Fig. 6
figure 6

Monthly fluctuation in average marginal growth rate (MGR) in otolith of Gymnothorax kidako. Vertical bars standard deviation

Full size image
$$\mathrm{Male}: {L}_{t}= 93.4 \left\{1 -\mathrm{ exp }\left[-0.15\left(t-0.62\right)\right]\right\}.$$
$$\mathrm{Female}: {L }_{t}= 83.5 \{1 -\mathrm{ exp }\left[-0.21\left(t-1.35\right)\right]\}.$$
$$\mathrm{Sex combined}: {L}_{t} = 86.7 \{1 -\mathrm{ exp }\left[- 0.19\left(t-1.05\right)\right]\}.$$

Males and females grew rapidly over the first 10 years, after which their growth rate slowed, with that of males then being slightly greater than that of females (Fig. 7). Growth curves of males and females differed significantly (F-test, F = 5.78, P < 0.01).

Fig. 7
figure 7

Individual length at age and estimated von Bertalanffy growth curves for Gymnothorax kidako males, females, and sex combined

Full size image

Discussion

Throughout the year, male GSI varied little (Fig. 1b), and sperm was detected (Fig. 4a–c). Conversely, female GSI was especially high in July and August (Fig. 1c), and tertiary vitellogenic oocytes were detected in gonads in July and August, indicating being mature (Fig. 4f; Table 1). Spawning was likely imminent in females with ovulated oocytes in pairs #7 and 8 collected in August 2018 (Table 2). Therefore, it appears that the spawning season of Gymnothorax kidako at Tateyama extends from July to August, coinciding with increased water temperatures (Fig. 1). Spawning in G. javanicus at Johnston Atoll, Hawaii, runs from July to August (Brock 1972). There may be two spawning peaks (June and November) in G. equatorialis from the western coast of Mexico (Lucano-Ramírez et al. 2008). In the Mediterranean Sea, Muraena helena spawns in summer (Matić-Skoko et al. 2011; Sallami and Ben Ibrahim 2021). Spawning of G. minor is predicted to run from June to September in Atami, Shizuoka, Japan (Oomori et al. 2022b). Females in male–female pairs of both Enchelycore pardalis and Enchelycore lichenosa from Tateyama collected in August 2018 had abdomens full of ovulated oocytes, indicating that spawning was imminent (Oomori et al. 2022a). These results indicate that moray eel spawning generally occurs in summer, but further comparative study is needed to confirm this.

In August, tertiary vitellogenic oocytes were not detected in 30% of the females, suggesting that these females had already spawned, because all females matured in July (Table 1). Primary vitellogenic or cortical alveolar oocytes in these females may develop to tertiary vitellogenic stage, and the females may spawn again.

Formation of a single annulus in M. helena coincided with spawning season (Matić-Skoko et al. 2011). In G. kidako, MGR decreased from July to August (Fig. 6). Because spawning should occur from July to August, investment into pairing by male and ovary development by female may affect growth, leading to formation of an annual ring. Fig. 7 suggests that G. kidako grows rapidly until 70 cm TL over the first 10 years, and then grows more slowly until reaching a larger size. By 34 years of age, males appear to reach 93.4 cm TL, and females 83.5 cm TL.

Generally, the number of individuals declines by aging in various fishes (e.g., Solomon et al. 1987; Kurita et al. 1991; Shimose and Tachihara 2005; Kanak and Tachihara 2006; Akita et al. 2017; Akita and Tachihara 2019). However, in G. kidako, many aged individuals were collected in both males and females (Fig. 7). This may be due to low predation and fishing pressures on adult fish of this species in Tateyama.

While species of Anguilla (Anguillidae) inhabit rivers and Conger (Congridae) coastal areas, both of them migrate to spawn in offshore open-ocean waters (Miller et al. 2011; Tsukamoto et al. 2011; Kurogi et al. 2012; Miller and Tsukamoto 2020). However, fully mature female E. pardalis, E. lichenosa (Oomori et al. 2022a), and G. kidako (this study) indicate that spawning likely occurs in coastal Tateyama Bay waters. Female Mediterranean moray eel M. helena with mature oocytes have not been collected in coastal waters of the Adriatic Sea, likely because they move into deeper waters to spawn (Matić-Skoko et al. 2011). These studies suggest that muraenid species do not migrate far offshore unlike those of Anguilla and Conger.

Gymnothorax kidako is a pair spawner (Moyer and Zaiser 1982). All of our pairs comprised a male and female (Table 2). Because pairs were established in June and spawning should commence in July, pairs might be maintained for longer than one month. In each pair, the male was larger than the female (Table 2). The greater growth rate of males may reflect this pairing pattern. The smallest mature female was 62.1 cm TL, and the smallest male in which sperm was detected was 47.4 cm TL. Because females established pairs with larger males, males < 62.1 cm TL are unlikely to establish pairs. However, further research is needed to clarify at what size individuals can participate in spawning.

Muraenid leptocephalus larvae have been identified using mitochondrial DNA, enabling a detailed description of their taxonomic characters (Tawa et al. 2012, 2013). However, the leptocephalus larva of G. kidako is unknown, and the smallest specimen collected in this study was 26.7 cm TL. Accordingly, we have no information on the early life history of this species. While we deduce that spawning extends from July to August, the mating system of G. kidako is unknown, except that spawning occurs in pairs (Moyer and Zaiser 1982). Further research is required to better understand the early life history, mating system, and ecology of this commercially fished moray eel for its sustainable management.