環境毒性学会誌 (Jpn. J. Environ. Toxicol.), 24, 12–25, 2021

Research Articles

Effect of ontogenetic changes of feeding habits on total mercury level in red , akajei

Teerapong DUANGDEE1, *, Wachirah JAINGAM2, Jun KOBAYASHI3 and Hiroaki TSUTSUMI3

1 Graduate School of Environmental and Symbiotic Sciences, Prefectural University of Kumamoto/3–1–100 Tsukide, Higashi-ku, Kumamoto 862–8502, 2 Faculty of Fisheries, Kasetsart University/50 Ngam Wong Wan Rd, Lad Yao, Chatuchak, Bangkok, 10900, Thailand 3 Faculty of Environmental and Symbiotic Sciences, Prefectural University of Kumamoto/3–1–100 Tsukide, Higashi-ku, Kumamoto 862–8502, Japan

ABSTRACT In Isahaya Bay, Kyushu, Japan, one of the edible ray , , Hemitrygon akajei tends to acceleratedly accumulate total mercury (THg) that was released from Mt. Unzen beside the bay as it grows. In this study, we collected 20 individuals of the female ray with wide variety of body sizes (16 to 65 cm in disc size, 140 to 9,540 gww in body wet weight), and aimed to clarify its mechanism. The stomach content anal- ysis revealed ontogenetic changes of feeding habits among the immature females, females in transitional maturity, and mature females in its long life of 10 to 15 years or more. Immature females favor to preying on epifaunal macro-benthic such as shrimps, which are classi ed as “Low THg content group” of the secondary consum- ers, while mature ones mainly feed on short-neck clam, Ruditapes philippinarum, and polychaete, Nectoneanthes ijimai, which are “High THg content group” of the primary consumers. Therefore, mature females with larger bodies tend to acceleratedly accu- mulate THg and reached a max. 1,370 ng g−1 dw, although the stable isotope analysis of carbon and nitrogen of the muscles of the mature females indicates the descent of the trophic position to the immature ones in the system.

Key words: biomagni cation, feeding habits, Hemitrygon akajei, mercury, ontogenetic changes, red stingray, Ruditapes phillipinarum

1. INTRODUCTION and has brought serious negative impacts on Bioaccumulation of hazardous substances the health of not only the wildlife located at the such as heavy metals, polychlorinated bi- higher trophic positions in the food chain3, 10), phenyls, dioxin, etc. has been reported from but also humans through dietary intake11, 12). various aquatic ecosystems in lakes1–3), coastal Among these harmful substances, our research waters4–7) and oceans8, 9) throughout the world, has focused on the bioaccumulation process-

*Corresponding author, Email: [email protected]; Tel: 096–321–6715; Fax: 096–384–6765 Received: 28 July 2020; Accepted: 28 December 2020

— 12 — Teerapong Duangdee et al.

es of mercury to the animals occurring in the in the sediment in several times higher levels coastal water, since 7,400 tons of mercury to the organic particles derived from the dead per year is still being emitted into the envi- bodies of phytoplankton, since the organic ronment throughout the world not only from particles deposited on the sea oor tends to anthropogenic sources but also natural ones be decomposed by bacteria, but the mercury including volcanos13), which are among the contained in the particles is apt to remain in basic geographical components in the Japa- the sediment. The THg concentrated in the nese Archipelago. In Japan, the majority of the sediment is transferred to infaunal macro-ben- mercury emitted to the environment is nally thic animals such as clams and deposit-feed- discharged to the coastal waters, where over ing polychaetes that burrow in the sediment 1,879,000 tons of sh and shell sh has been through their feeding activities directly or caught to obtain seafood14). Careful intake of indirectly, while epifaunal macro-benthic ani- naturally contaminated mercury in seafoods mals such as amphipods and mussel occurring is, therefore, needed for the people living in on the surface of the sediment are apt to be Japan because of its negative impact on human much less exposed to the mercury accumulat- health. ed in the sediment. Consequently, although In the aquatic ecosystem, the mercury con- they were located at the same trophic position taminated in water tends to be absorbed by in the food web system in the bay as primary phytoplankton or aquatic animals such as sh consumers, the former infaunal consumers (through their respiratory activities), which contained 101±23 ng g−1 of THg, which were is referred to as “bioconcentration”. It is apt referred to as “High THg content group”, while to be further biologically concentrated on the the THg contents of the latter epifaunal ones body tissues of animals in higher trophic lev- were about one fourth of “High THg content els linked together with a food chain by their group”, 25.2±8.5 ng g−1, which were referred to predation, which is referred to as “biomag- as “Low THg content group”16). In the benthic ni cation”15). In the previous studies on the ecosystem, these two dierent levels of THg bioaccumulation, a process in which a mercury contained in the muscles of primary consumers is absorbed in an organism by all routes of are transferred to the secondary consumers in exposure as occurs in the aquatic ecosystem, the prey and predator relationship. Therefore, which includes its bioconcentration process by “High” and “Low” THg content groups were phytoplankton and biomagni cation by preda- also found among carnivorous benthic animals. tory animals, most have described the discon- THg contents of crabs and sea stars which prey tinuous increase of mercury in the body tissues on infaunal macro-benthic animals were sever- of the organisms linked together with a food al times higher than those of carnivorous poly- chain in the pelagic water2, 5, 6, 8). chaete and shrimp which depended on diets to Jaingam et al.16) found another characteristic the epifaunal ones. bioaccumulation process of mercury through Jaingam et al.17) found a further curious bio- the benthic ecosystem in the eld surveys in a magni cation process of THg between the mac- bay, Isahaya Bay, located beside an active vol- ro-benthic animals and carnivorous bentho- cano, Mt. Unzen, in Kyushu, Japan. The dead pelagic sh, red stingray (Hemitrygon akajei), bodies of phytoplankton that absorbed mercury in Isahaya Bay. The results of stable isotope from the surrounding water easily deposits on analysis of carbon and nitrogen of the muscles the sea oor of the shallow water of less than of red stingray indicate that its juvenile was 10 m in depth, and the mercury accumulates classi ed as one of the intermediate consumers

— 13 — E ect of feeding habits on total mercury level in red stingray

(very close to the tertiary one), but its trophic position in the benthic food web system de- scended to the level of an almost secondary consumer as it grew to an adult. Nevertheless, the THg contents of the adult increased to ex- tremely high levels, 670 to 3,700 ng g−1, which far exceeded the THg contents of the tertiary consumers of shes (about 250 to 270 ng g−1) oc- curring in the pelagic system of the same bay. Fig. 1. Study area, Isahaya Bay located in the Our study has focused on this biomagni- inner parts of Ariake Bay, Kyushu, western cation process of THg to the muscles of H. Japan. akajei. It is an endemic species of stingray that is distributed in a wide area of Paci c Ocean and an average depth of 10 m (Fig. 1). An ac- between northwestern part, Japan, Russia, tive volcano, Mt. Unzen, is located in the center , Thailand, Malaysia, and Indonesia18, 19). of Shimabara Peninsula, it is one of the major In Japan, it has been used as a popular sea natural sources of discharging mercury into the food item for a long time, and 113 tons of its bay21, 22), since this peninsula is located beside catch as sea food was recorded in 2006 accord- the southern side of Isahaya Bay. According to ing the catch statistics in Ariake Bay20). We the most recent record of volcanic eruptions, need to know the detailed mechanisms of how Mt. Unzen erupted for 1928 days between 1990 this species accumulates to such extremely and 1995, and emitted 2.95 metric tons of mer- high levels of THg in its muscles, although it cury into the surrounding environment22, 23). is located at the trophic level close to the sec- ondary consumer in the food web system of the 2.2 Red stingray collection bay. In total, 22 red stingray individuals were In this study, we caught 22 red stingray in- captured from Isahaya Bay with a longline dividuals with a wide variety of body size in eet on 27 April 2019 and a gill net on 15 Isahaya. We examined the stomach contents of May 2019. The specimens were weighed with these specimens to identify its main food items, a digital balance, and their disc widths were determined their stable isotope ratios of carbon measured with a measuring tape. The gender and nitrogen to describe their trophic positions was determined by the presence or absence in the food web system, and determined the of claspers. According to Furumitsu et al.20), THg content in the muscle of the stingray. We the individuals whose disc widths were larger clarify the mechanisms of the magni cation of than 56 cm in females and 35 cm in males were extremely high levels of THg in the muscles of regarded as sexually mature ones, the females the adult, and discuss the characteristics of the with the disc width between 45 and 55 cm were bioaccumulation process of THg in the benthic treated as ones in transitional maturity, and system in the coastal shallow water. those with the disc width or the less than 44 cm were regarded as immature. The ventral part 2. MATERIALS AND METHODS of the body of each specimen was dissected to 2.1 Study area remove the stomach. Following the methods Isahaya Bay is located at the inner part of noted by Bowen24) and Weidner et al.25), the Ariake Bay, Kyushu, Western Japan. Isahaya stomach samples were put in plastic bottles Bay is an enclosed bay with total area of 65 km2 and xed with undiluted formalin solution

— 14 — Teerapong Duangdee et al.

to prevent decomposition, putrefaction, and 2.4 Chemical analysis autolysis of the stomachs for approximately 2.4.1 Stable isotope analysis one month, and the subsequent 70% ethanol Frozen specimens of the muscles of red sting- preservation is done to maintain the integri- ray were freeze-dried for 48 h, ground, and ty of xed stomachs specimens for long-term homogenized using a pestle and mortar. After storage. The muscles of each specimen were lipids were removed from these samples using sampled from the dorsal part using a surgical a chloroform–methanol mixture solution (2 : 1, knife, and kept in an icebox. At the laboratory, V/V), they were centrifuged at 10,000 rpm at they were kept in a freezer at −30°C prior to 4°C for 5 min twice, treated with 100% metha- the chemical analyses. nol, centrifuged at 10,000 rpm at 4°C for 5 min We captured one individual of immature again, and nally vacuum-dried for 48 h. They female red stingray in Isahaya Bay to observe were also treated with 2 N HCl, centrifuged at the teeth structure on 14 October 2020. 35,000 rpm at 4°C for 5 min twice, and vacu-

um-dried for 48 h to remove CaCO3. Prior to the 2.3 Stomach contents analysis analysis, the samples were ground and kept in The stomach specimens of the red stingray plastic vials. were weighed, and dissected. Their contents The stable isotope ratios of carbon and nitro- were observed under a stereoscopic microscope. gen of the samples were determined using an The remains of the preys in the stomach were elemental analyzer (Flash Elemental Analyzer identi ed, counted and weighed within each 1112 Series, Thermo Electron) and continuous identi able taxonomic group. The unidenti - ow isotope ratio mass spectrometer (Delta able or highly digested items were just weight. Plus, Thermo Electron). The data of stable The data on the identi able prey items were isotope ratios of carbon and nitrogen are ex- used to calculate “Percentage of Index of Rel- pressed with delta notations as parts per mill ative Importance, % IRI” with following three (in ‰) as follows: indices, Percentage by number (%N), Percent- 13 12 age by weight (%W), and Percentage of occur- 13 C /Csample δ C =×13 12 − 1 1000 (‰) rence (%O). C/ Cstandard

15N /N 14 Index of Relative Importance (IRI) = δ15N=×sample −1 1000 (‰) 15N /N 14 (%N+× %W) %O standard

26) n %IRIi=× 100 IRI ii IRI Pee Dee Belemnite (PDB) and atmospheric (  i=1 ) nitrogen were used as references for 13C and n: the total number of identi able taxo- 15N, respectively. Glycine was used as a work- nomic group, ing standard in this study. The overall analyti- i: a given identi able taxonomic group cal error was within ±0.2‰. 2.4.2 Total mercury content analysis %N= 100×(total count of a given prey Prior to the determination of total mercu- taxa)/(total count of all prey taxa)27) ry contents, the specimens of the muscles of %W= 100×(total mass (g) of a given prey red stingray were ground and homogenized. taxa)/(total mass of all prey taxa)27) Total mercury contents of these samples were %O=100 ×(total number of stomachs con- determined using the MA-3000 mercury ana- taining a given prey taxa)/(total num- lyzer (Thermal vaporization atomic absorption, ber of stomachs with any prey)24) Nippon Instruments). The detection was based

— 15 — E ect of feeding habits on total mercury level in red stingray

on cold-vapor atomic absorption spectroscopy at a wavelength of 253.7 nm. The NIMJ CRM 7402-a (cod sh tissue) was used as standard reference for test the accuracy of the method.

2.5 Statistical analysis The statistical signi cances of the correla- tions between disc width and body weight, between the values of δ 13C and δ15N of the mus- cle, and between disc width and THg content of Fig. 2. Relationship between disc width and whole red stingray were calculated with a nonpara- body weight of 22 individuals of red sting- metric test, Spearman’s correlation coecient ray, Hemitrygon akajei, captured in this by rank. Those of the mean values of δ13C study in Isahaya Bay. Open circles: imma- ture females, Gray circles: females in tran- and δ15N of the muscles were evaluated with sitional maturity, Closed circles: mature fe- Mann–Whitney’s U test among the immature males, Closed diamonds: mature males. The female, female in transitional maturity, and number of each plot indicates its specimen mature female. These statistics were conducted number noted in Appendix Table A1. with Statview ver. 4.5 (Abacus Corporation, U.S.) for Apple Macintosh, and used through- is almost equivalent to the specimen No. 02 out, using a signi cance level of p<0.05. (immature female with the disc width of 19 cm and the body weight of 240 g). Therefore, ma- 3. RESULTS ture females with the disc widths of 56 to 65 cm 3.1 Relationship between disc width and body seem to be regarded as individuals at least 10 weight to 15 years old, and further elder ones occurred Figure 2 indicates the relationship between in the study area, Isahaya Bay. the disc width and body weight of 22 red sting- The sex ratio of the specimens was extremely ray individuals captured in this study (The biased to female (20 individuals of the total 22 data are noted in Appendix Table A1). A signif- red stingrays). Since the period required to the icant exponential curve is found between them size at sexual maturity is signi cantly dierent (BW=0.03DW3.09, DW: Disc width, BW: Body between males and females in red stingrays20), weight) (Spearman’s correlation coecient by we compared the food items, trophic positions rank, z=4.517, p<0.0001, n=22). The growth in the food web system, and THg contents of rate and life span of red stingray are not clear the muscles among the specimens of immature in the previous studies. However, National females, ones in transitional maturity, and ma- Geographic reported that the red stingray ture ones in this study. reached max. 2 m in the whole body length and about 100 kg, and estimated that the max 3.2 Stomach content analysis longevity was 15 to 25 years28). We were able to The biological information of the 20 spec- collect a max. 9,540 g of the individual in this imens of female red stingray used for the study, but found a further larger one although analysis are noted in Appendix Table A1. The it was too heavy to capture with the longline disc widths of seven specimens of immature eet and gill net used in this study. Yokota29) female, three ones of female in transitional estimated that the individuals with the disc maturity, and ten ones of mature female were width of about 18 cm were four years old, which 25±5 cm (mean±S.D.), 50±5 cm, and 60±3 cm,

— 16 — Teerapong Duangdee et al.

Fig. 3. “Percentage of Index of Relative Importance (%IRI)” of the identi able stomach contents of female red stingray, Hemitrygon akajei, caught in Isahaya Bay. (A) Immature individuals (n=7), (B) Fe- males in transitional maturity individuals (n=3), (C) Mature individuals (n=10). respectively. The stomach contents of these diet preferences of the mature one drastically three dierent maturity level of females were changed to infaunal macro-benthic animals compared with an index, %IRI, (Fig. 3, Appen- burrowing the sediment such as short-neck dix Table A2). In this study, 71.7%, 53.5% and clam, R. philippinarum, and polychaete, N. 58.0% of the stomach contents in weight were ijimai. identi able to at least phylum in the specimens of immature female, female in transitional 3.3 Stable isotope ratios of carbon and nitrogen maturity, and mature female, respectively. We evaluated how the changes of the food In immature female, the identi able preys of items aected the female red stingray following the stomach contents were mainly made up of the growth on its trophic position in the ben- Arthropoda (75.4%) in %IRI, involving other thic food web system with the results of stable crabs (42.3%) and shrimp (33.1%), small shes isotope analysis of carbon and nitrogen of the as Chordata (16.5%), and Annelida (8.1%). The muscles of the specimens. Figure 4 compares %IRI value of Arthropoda decreased to 34.6% the relationship of the values of stable isotope in female in transitional maturity, and 8.2% in ratios of carbon and nitrogen of the muscles mature female, while those of and An- among the immature female, female in tran- nelida increased to 2.1% and 21.9% in female in sitional maturity, and mature female. The transitional maturity, and 45.2% and 21.6% in values of δ13C and δ15N of the muscles of the mature female, respectively. In particular, the immature female and female in transitional mature females exclusively favored to feed on maturity were −15.8±0.8‰ and 15.7±0.6‰ the short-neck clam, Ruditapes phillipinnarum (mean±S.D., n=7) and −16.5±0.9‰ and (45.0%) and a polychaete, Nectoneanthes ijimai 15.5±0.5‰ (mean±S.D., n=3), respectively (21.6%). (Appendix Table A1), and the value of δ15N of Thus, in the case of female red stingray, the these ten specimens signi cantly increased food items markedly changed as it grew up to in proportion to that of δ13C (r2=0.679, n=10) be sexually mature. The immature one tend- (Spearman’s correlation coecient by rank, ed to feed on mainly epifaunal macro-benthic z=2.534, p<0.011). In the mature female, all animals such as crab and shrimp, while the of the values of δ15N were located below the

— 17 — E ect of feeding habits on total mercury level in red stingray

13 15 Fig. 4. Relationships between δ C and δ N values Fig. 5. Relationships between Disc width and of the body tissues of females red stingray, THg content of the body tissues of mature Hemitrygon akajei, collected in Isahaya and immature females of red stingray, Bay. Open circles: immature females (n=7), Hemitrygon akajei, collected in Isahaya Gray circles: females in transitional matu- Bay. Open circles: immature females (n=7), rity (n=3), Closed circles: mature females Gray circles: females in transitional matu- (n=10). rity (n=3), Closed circles: mature females (n=10), Triangles: immature females from the previous study17) (n=2). regression line between the values of δ13C and δ15N of the immature female and female in transitional maturity. philippunarum: 9.7±0.4‰ (Jaingam et al.16)) The mean value of δ15N of the mature female (45.2% in %IRI) and polychaete, N. ijimai, was signi cantly lighter (−0.7‰) than that of (21.6% in %IRI) as primary consumer in the the immature one (Mann–Whitney’s U test, mature one (Fig. 3). U=13.50, p=0.035). The mean value of δ13C of the mature female was slightly heavier than 3.4 Biomagnification of THg those of the immature female (+0.4‰) and The THg contents of the muscles of 20 indi- female in transitional maturity (+1.1‰), but vidual of females red stingray with the three the dierence of δ13C values between all ten dierent maturity levels (immature, transition- individuals of the immature female and female al maturity, and mature) collected from Isaha- in transitional maturity and mature one was ya Bay ranged between 241 and 1,370 ng g−1 dw not statistically signi cant (Mann–Whitney’s (Appendix Table A1). Figure 5 indicates the U test, U=32.00, p=0.173). These results in- relationship between the disc width and dicate that all of the females in three dierent THg content of all specimens. A statistically sexually mature levels preyed on the diets signi cant relationship was found between nutritionally sustained by the same prima- the disc width and THg content of the mus- ry producers, but the trophic position of the cles (y=159.99e0.03x, r2=0.683, n=20, Spear- mature female was signi cantly lower than man’s correlation coecient by rank, z=3.479, that of the immature one. This signi cant de- p<0.001). However, there were no signi cantly crease of δ15N value of the mature female was increasing trends of THg content of the mus- supported by the ontogenetic changes of main cles to the growth of immature female (Spear- food items from shrimp (e.g., Metapenaeus man’s correlation coecient by rank, z=0.525, joyneri: 14.4±0.4‰ (Jaingam et al.16)) and crab p=0.600), although they seem to be made up as the secondary consumers (75.4% in %IRI) of the one with a variety of age at least for four in the immature female to short-neck clam (R. years, judging from the growth rate report-

— 18 — Teerapong Duangdee et al.

ed by Yokota29). These results indicate that a at the lower trophic position to the immature signi cant biomagni cation of THg was not individuals. recognized in the muscles of immature females. (We added two more data of the specimens 4. DISCUSSION of female immature individuals with the disc According to common knowledge of the bio- size of 33 cm (△ in Fig. 5, Appendix Table A1) magni cation of heavy metals to the organisms collected by the previous study17) in the same in the aquatic ecosystem, the contents of the study area. However, a statistically signi cant contaminants increase discontinuously in each relationship was not found between the disc step of the trophic levels in the food web sys- width and THg content of the muscles of the tem8, 9, 30, 31). Aquatic predatory animals with total nine individuals. Spearman’s correlation large body sizes such as ray, shark, and dol- coecient by rank, z=1.349, p=0.177). phin are often located at the highest position The biomagni cation of THg of the muscles of the food web system. Extremely high levels of female red stingray has a unique character- (mg g−1 dw levels) of heavy metals including istic. Figure 6 shows the relationship between mercury tend to accumulate in the muscles the δ15N value and THg content of the mus- and in some organs such as gills and livers of cles. The THg content of mature female was the adults with large body sizes of 10 to 100 kg 869±268 ng g−1 dw (mean±S.D.), which was sig- or more in dolphin and whale8), shark10, 32–37), ni cantly higher than that of the immature one ray38–40), skate41), ray and guitar sh42). How- (309±76 ng g−1 dw) (Mann–Whitney’s U test, ever, it is not easy to show the representative U=1.000, p<0.001), although the δ15N value of contents of these substances in the body of immature female (15.7±0.6‰) indicates that each species, nor is it easy to compare their it located at the higher trophic position than species-speci c values, since they have accu- mature one (δ15N: 15.0±0.6‰) in the food web mulated them in the body throughout their system (Mann–Whitney’s U test, U=13.500, long lives, having accelerated them along their p=0.036). Thus, in the case of female red sting- growth. ray, signi cantly higher levels of THg accu- In some cases of rays and sharks, ontogenet- mulated in the muscles, although they located ic changes of the feeding habits occur following the growth, and they bring a further elevation of the trophic position in the food web system, and increase of mercury contents in their bod- ies33, 34, 43, 44). In this study, the results of the stomach content analysis of red stingray (H. akajei) clearly showed the ontogenetic changes of the feeding habits between the immature and mature females, too (Fig. 3). The immature individual has undeveloped small teeth of less than 500 µm in diameter, sparsely distributed Fig. 6. Relationships between δ15N values and in the mouth (Fig. 7(A)), while the mature one THg content of the body tissues of females has much larger teeth (about 5 mm in length), red stingray, Hemitrygon akajei, collected overlapping each other (Fig. 7(B)) as reported in Isahaya Bay. Open circles: immature by Taniuchi and Shimizu45). Therefore, the im- females (n=7), Gray circles: females in transitional maturity (n=3), Closed circles: mature ones can only eat the animals with soft mature females (n=10). body structure such as shrimp, crab, small sh,

— 19 — E ect of feeding habits on total mercury level in red stingray

Fig. 7. Teeth of female red stingray, Hemitrygon akajei, caught in Isahaya Bay. (A) Immature individuals with the disc width of 12 cm, (B) Mature individuals with the disc width of 59 cm. etc., but the developed teeth of mature ones en- gam et al.16)). able them to feed on bivalves with hard shells. The mercury contained in the muscles of the The ontogenetic changes of the feeding infaunal macro-benthic animals classi ed as habits exceptionally led a descending order of members of “High THg content group” of the the trophic position in the food web system to primary consumers is derived from one emitted the mature females (Fig. 4 and Fig. 6) since from an active volcano via two steps of biocon- the immature ones tended to prefer epifaunal centration processes in the water and the sedi- macro-benthic animals as the secondary con- ment46). It can, therefore, be concluded that the sumers involving Arthropoda such as shrimps unique biomagni cation of THg to the muscles and crabs for diets, while the mature ones are of mature red stingray females observed in Isa- apt to prey on infaunal macro-benthic animals haya Bay has been proceeded by the combined as the primary consumers mainly consisting of eects of the emission of mercury during vol- short-neck clam, R. philippinarum, and poly- canic activity, deposition of high levels of THg chaete, N. ijimai (Fig. 3). Nevertheless, the ma- in the sediment in the bay by the two steps of ture females were followed by the accelerated bioconcentration processes by phytoplankton accumulation of THg content in their muscles in the water and the bottom sediment and (Fig. 5). transference of the THg deposited on the sed- The key factor that raised the THg contents iment to the infaunal macro-benthic animals in the muscles of the mature females of red burrowing the sediment, and its ontogenetic stingray occurring in Isahaya Bay is, therefore, changes of the feeding habits from the epifau- the ontogenetic changes of the feeding habits nal macro-benthic animals to the infaunal ones between the immature and mature females. between the immature and mature females. The most preferred prey items by the mature The ontogenetic changes of main food items females after the ontogenetic changes of the of red stingray are not clearly observed from feeding habits have occurred, infaunal mac- both males and females in the population ro-benthic animales, belong to “High THg con- occurring in Tokyo Bay45). It mainly favors tent group” of the primary consumers, while to prey on small sh, including the ones by the immature females, epifaunal shrimp, mantis shrimp, and crab irrespective macro-benthic animals, belong to “Low THg of sexual maturity and growth. The dierence content group” of the secondary consumers. of feeding habits between Isahaya Bay and They have signi cantly lower THg contents Tokyo Bay seems to be responsible for the to- than the infaunal primary consumers (Jain- pography of the shore. In Tokyo Bay, the sandy

— 20 — Teerapong Duangdee et al.

tidal ats that dense patches of clam have Kyushu, Japan, clearly showed ontogenetic often established have disappeared widely due changes of feeding habits as it grows. The to reclamation in the past ve decades, and the immature females favor to prey on epifaunal stock of clams has markedly declined in the macro-benthic animals as secondary consum- bay47). Nowadays, it is dicult for red stingrays ers, mainly shrimps and crabs, while mature to nd clams on the shore in Tokyo Bay even if ones prefer to feed on infaunal ones burrowing it favors to prey on it. in the sediment as primary consumers, such as In this study, all of the 22 specimens of red short-neck clam and polychaete. It is very like- stingray collected from Isahaya Bay contained ly that this changes lead the mature females THg between 241 and 1,370 ng g−1 dw, and had of red stingray to accelerate the accumulation an increasing tendency of THg content in the of THg to its muscles (max. 1,370 ng g−1 dw) muscles to the growth (Fig. 5). The provisional since the food items preferred by the mature regulation value of THg for sh and shell sh is ones contained higher levels of THg than those 400 ng g−1 ww in Japan48), which is equivalent to by the immature ones, and to descend the tro- 2,000 ng g−1 dw17). According to the relationship phic position to the immature females in the between the disc width and THg content of red food web system. The unique bioaccumulation stingray in this study (Fig. 5), the THg content of THg to the muscles of mature females red of the muscles of the individual that grows up stingray observed in Isahaya Bay seems to to a disc width of 84.2 cm reaches provisional have been proceeded by the combined eects regulation value. Its body weight is estimated of the continuous emission of mercury from an to 26,713 g ww from the equation between the active volcano, Mt. Unzen, two steps of biocon- disc width and body gained from the speci- centration processes of THg by phytoplankton mens of this study (BW=0.03DW3.09, BW: body in the water and the deposition of the organic weight, DW: disc width). Therefore, it seems to particles derived from the phytoplankton and be safe to eat the market size (4 to 5 kg ww, 50 their decomposition in the sediment, the trans- to 60 cm in disc width; less than 10 kg ww) of mission of the THg deposited in the sediment red stingray. to the infaunal macro-benthic animals, and The present study deals with the process of nally accumulation at high values in adult THg transference and its biomagni cation pro- females by their ontogenetic changes of feeding cesses between the bottom sediment and red habits to mainly feed on infaunal preys. stingray in an enclosed bay. As a contaminant in sea foods, another component of mercury, Acknowledgment methylmercury, has a much higher toxicity to This research was nancially supported by human health11, 12), and it is very likely that its Kumamoto Prefectural Government Oce and transference and biomagni cation processes Prefectural University of Kumamoto, Inter- in the benthic system in the bay do not neces- national Postgraduate program for Research sarily coincide with those of THg shown in this on Mercury. The authors are greatly grateful study. Now, we are examining in these issues to Dr. Tetsuro Agusa for his criticism to the with same samples and specimens. The results results of this study and Dr. Jerey Stewart will be reported elsewhere. Morrow for his critical reading and correction of the English text of the draft paper. 5. CONCLUSIONS The stomach contents of red stingray, References Hemitrygon akajei, occurring in Isahaya Bay, 1) Bartons, M., Joan, O. G., Jordi, C. (2012)

— 21 — E ect of feeding habits on total mercury level in red stingray

Food web bioaccumulation of organoha- web: Insights from stable nitrogen isotope logenated compounds in high mountain analysis. Canadian Journal of Fisheries lakes. Limnetica, 31, 155–164. and Aquatic Sciences, 55(5), 1114–1121. 2) Lavoie, R. A., Jardine, T. D., Chumchal, 9) Clayden, M. G., Arsenault, L. M., Kidd, K. M. M., Kidd, K. A., Campbell, L. M. (2013) A., O’Driscoll, N. J., Mallory, M. L. (2014) Biomagni cation of mercury in aquatic Mercury bioaccumulation and biomagni - food webs: A worldwide meta-analysis. En- cation in a small Arctic polynya ecosystem. vironmental Science & Technology, 47(23), Science of the Total Environment, 509–510, 13385–13394. 206–215. 3) Ward, D. M., Nislow, K. H., Folt, C. L. 10) Al-Reasi, H. A., Ababneh, F. A., Lean, D. (2010) Bioaccumulation syndrome: Identi- R. (2007) Evaluating mercury biomag- fying factors that make some stream food ni cation in sh from a tropical marine webs prone to elevated mercury bioaccu- environment using stable isotopes (δ13C mulation. Annals of the New York Acade- and δ15N). Environmental Toxicology and my of Sciences, 1195(1), 62–83. Chemistry, 26(8), 1572–1581. 4) Baeyens, W., Leermakers, M., Papina, T., 11) Harada, M. (1995) Minamata disease: Saprykin, A., Brion, N., Noyen, J., De Gi- Methylmercury poisoning in Japan caused eter, M., Elskens, M., Goeyens, L. (2003) by environmental pollution. Critical Re- Bioconcentration and biomagni cation of views in Toxicology, 25(1), 1–24. mercury and methylmercury in north sea 12) Yokoyama, H. (2018) Mercury Pollution in and scheldt sh. Archives of Envi- Minamata. Springer Open, 74 pp. ronmental Contamination and Toxicology, 13) UN Environment (2019) Global Mercury 45(4), 498–508. Assessment 2018. UN Environment Pro- 5) Cardoso, P. G., Pereira, E., Duarte, A. C., gramme, Chemicals and Health Branch Azeiteiro, U. M. (2014) Temporal char- Geneva, Switzerland. acterization of mercury accumulation at 14) Ministry of Agriculture, Forestry and Fish- dierent trophic levels and implications for ery (2018) FY2018 Trends in Fisheries, metal biomagni cation along a coastal food FY2019 Fisheries Policy, White Paper on web. Marine Pollution Bulletin, 87(1–2), Fisheries: Summary. Chapter III Trends in 39–47. Japan’s Fisheries Since FY2017. 6) Harding, G., Dalziel, J., Vass, P. (2018) 15) Arnot, J. A., Gobas, F. A. P. C. (2006) A Bioaccumulation of methylmercury within review of bioconcentration factor (BCF) the marine food web of the outer Bay of and bioaccumulation factor (BAF) assess- Fundy, Gulf of Maine. PLoS One, 13(7), ments for organic chemicals in aquatic e0197220. organisms. Environmental Reviews, 14(4), 7) Kim, E., Kim, H., Shin, K. H., Kim, M. S., 257–297. Kundu, S. R., Lee, B. G., Han, S. (2012) 16) Jaingam, W., Komorita, T., Ishimatsu, S., Biomagni cation of mercury through the Takenaka, R., Umehara, A., Kobayashi, benthic food webs of a temperate estuary: J., Yamamoto, M., Arizono, K., Tsutsumi, Masan Bay, . Environmental Toxicol- H. (2018) Biomagni cation process of total ogy and Chemistry, 31(6), 1254–1263. mercury in the macro-benthic animals in 8) Atwell, L., Hobson, K. A., Welch, H. E. an enclosed bay, Isahaya Bay, Kyushu, (1998) Biomagni cation and bioaccumu- Japan. Jpn J Environ Toxicol, 21, 9–20. lation of mercury in an arctic marine food 17) Jaingam, W., Komorita, T., Ishimatsu, S.,

— 22 — Teerapong Duangdee et al.

Kobayashi, J., Yamamoto, M., Arizono, K., ytrygon violacea (Bonaparte, 1832) diet Tsutsumi, H. (2018) Inuence of benthic from the western North Atlantic Ocean. biomagni cation process on the total mer- Journal of Applied Ichthyology, 33(3), cury content of sh and mega-benthos 386–394. in an enclosed bay. Journal of the Japa- 26) Cortés, E. (1997) A critical review of nese Society of Shiranuikai & Kumagawa methods of studying sh feeding based on Regional Studies, 12, 3–14. analysis of stomach contents: Application 18) Yamaguchi, A., Aonuma, Y., Yagishita, N., to elasmobranch shes. Canadian Journal Yoshino, T. (2013) Dasyatidae. In: Nakabo of Fisheries and Aquatic Sciences, 54(3), T ed Fishes of Japan with Pictorial Keys to 726–738. The Species, Third edition: Tokai Universi- 27) Hyslop, E. J. (1980) Stomach contents ty Press, 220–226. analysis: A review of methods and their 19) Weigmann, S. (2011) Contribution to the application. Journal of Fish Biology, 17(4), and distribution of eight ray 411–429. species (, Batoidea) from 28) National Geographic Home Page, Red coastal waters of Thailand. Proc Soc Nat Stingray. The Illustrated Book of Ani- Sci Hamburg, 46, 249–312. mals (Doubutsu Daizukan), Red Sting- 20) Furumitsu, K., Wyels, J. T., Yamaguchi, ray (akaei). (In Japanese) https://natgeo. A. (2019) Reproduction and embryonic de- nikkeibp.co.jp/nng/article/20141218/429037/. velopment of the red stingray Hemitrygon 29) Yokota, T. (1952) Studies on the stocks of akajei from Ariake Bay, Japan. Ichthyolog- sharks and rays. 1. A method of age esti- ical Research, 66(4), 419–436. mation. Nippon Suisan Gakkaishi, 17(10), 21) Nriagu, J. O. (1989) A global assessment of 321–325. natural sources of atmospheric trace met- 30) Braune, B., Chételat, J., Amyot, M., als. Nature, 338(6210), 47–49. Brown, T., Clayden, M., Evans, M., Fisk, 22) Nriagu, J., Becker, C. (2003) Volcanic A., Gaden, A., Girard, C., Hare, A., Kirk, emissions of mercury to the atmosphere: J., Lehnherr, I., Lether, R., Loseto, L., Global and regional inventories. Science of Macdonald, R., Mann, E., McMeans, B., the Total Environment, 304(1–3), 3–12. Muir, D., O’Driscoll, N., Poulain, A., Re- 23) Unzen Restoration Project Oce (2007) imer, K., Stern, G. (2015) Mercury in the Unzen-Fugendake eruption executive sum- marine environment of the Canadian Arc- mary 1990–1995. Unzen Restoration Proj- tic: Review of recent ndings. Science of ect Oce, Kyushu Regional Construction the Total Environment, 509–510, 67–90. Bureau, Ministry of Land, Infrastructure 31) Lavoie, R. A., Hebert, C. E., Rail, J. F., and Transport, 22 pp. Braune, B. M., Yumvihoze, E., Hill, L. G., 24) Bowen, S. H. (1996) Quantitative de- Lean, D. R. S. (2010) Trophic structure scription of the diet. In: Murphy BR and and mercury distribution in a Gulf of St. Willis DW eds. Fisheries Techniques, 2nd, Lawrence (Canada) food web using stable Bethesda, MD: American Fisheries Soci- isotope analysis. Science of the Total Envi- ety. 513–532. ronment, 408(22), 5529–5539. 25) Weidner, T. A., Hirons, A. C., Leavitt, 32) Suk, S. H., Smith, S. E., Ramon, D. A. A., Kerstetter, D. W. (2017) Combined (2009) Bioaccumulation of mercury in gut-content and stable isotope trophic pelagic sharks from the northeast Paci c analysis of the pelagic stingray Pteropla- ocean. Bioaccumulation in Pelagic Sharks

— 23 — E ect of feeding habits on total mercury level in red stingray

CalCOFI Rep, 50, 172–177. in muscle and liver of the golden cownose 33) Endo, T., Hisamichi, Y., Haraguchi, K., ray, Rhinoptera steindachneri, Evermann Kato, Y., Ohta, C., Koga, N. (2008) Hg, Zn and Jenkins, 1891, from the upper Gulf of and Cu levels in the muscle and liver of California, Mexico. Bulletin of Environmen- tiger sharks (Galeocerdo cuvier) from the tal Contamination and Toxicology, 83(2), coast of Ishigaki Island, Japan: Relation- 230–234. ship between metal concentrations and 39) Horvat, M., Degenek, N., Lipej, L., Trat- body length. Marine Pollution Bulletin, nik, J. S., Faganeli, J. (2014) Trophic 56(10), 1774–1780. transfer and accumulation of mercury in 34) Endo, T., Hisamichi, Y., Kimura, O., Kota- ray species in coastal waters aected by ki, Y., Kato, Y., Ohta, C., Koga, N., Hara- historic mercury mining (Gulf of Trieste, guchi, K. (2009) Contamination levels of northern Adriatic Sea). Environmental mercury in the muscle of female and male Science and Pollution Research, 21(6), spiny dog shes (Squalus acanthias) caught 4163–4176. o the coast of Japan. Chemosphere, 40) Murillo-Cisneros, D. A., O’Hara, T. M., 77(10), 1333–1337. Castellini, J. M., Sánchez-González, A., 35) Endo, T., Hisamichi, Y., Kimura, O., Oga- Elorriaga-Verplancken, F. R., Marmole- sawara, H., Ohta, C., Koga, N., Kato, Y., jo-Rodríguez, A. J., Marín-Enríquez, E., Haraguchi, K. (2013) Levels of mercury Galván-Magaña, F. (2018) Mercury con- in muscle and liver of star-spotted dog sh centrations in three ray species from the (Mustelus manazo) from the northern re- Paci c coast of Baja California Sur, Mexico: gion of Japan: A comparison with spiny Variations by tissue type, sex and length. dog sh (Squalus acanthias). Archives of Marine Pollution Bulletin, 126, 77–85. Environmental Contamination and Toxi- 41) Taylor, D. L., Kutil, N. J., Malek, A. J., cology, 64(3), 467–474. Collie, J. S. (2014) Mercury bioaccumula- 36) Endo, T., Kimura, O., Ogasawara, H., tion in cartilaginous shes from Southern Ohta, C., Koga, N., Kato, Y., Haraguchi, K. New England coastal waters: Contami- (2015) Mercury, cadmium, zinc and copper nation from a trophic ecology and human concentrations and stable isotope ratios health perspective. Marine Environmental of carbon and nitrogen in tiger sharks Research, 99, 20–33. (Galeocerdo cuvier) culled o Ishigaki Is- 42) Murillo-Cisneros, D. A., O’Hara, T. M., land, Japan. Ecological Indicators, 55, 86– Elorriaga-Verplancken, F. R., Sánchez- 93. González, A., Marín-Enríquez, E., Marmo- 37) Biton-Porsmoguer, S., Bǎnaru, D., Boudour- lejo-Rodríguez, A. J., Galván-Magaña, F. esque, C. F., Dekeyser, I., Bouchoucha, M., (2019) Trophic structure and biomagni ca- Marco-Miralles, F., Lebreton, B., Guillou, tion of total mercury in ray species within a G., Harmelin-Vivien, M. (2018) Mercury in benthic food web. Archives of Environmen- blue shark (Prionace glauca) and short n tal Contamination and Toxicology, 77(3), mako (Isurus oxyrinchus) from north-east- 321–329. ern Atlantic: Implication for shery man- 43) Dale, J. J., Wallsgrove, N. J., Popp, B. N., agement. Marine Pollution Bulletin, 127, Holland, K. N. (2011) Nursery use 131–138. and foraging ecology of the brown stingray 38) Gutiérrez-Mejía, E., Latres, M. L., Sosa- lata determined from stomach Nishizaki, O. (2009) Mercury and arsenic contents, bulk and amino acid stable iso-

— 24 — Teerapong Duangdee et al.

topes. Marine Ecology Progress Series, 433, 46) Tsutsumi, H., Jaingam, W., Duangdee, T., 221–236. Arizono, K. (2019) Characteristics of the 44) Lyons, K., Carlisle, A. B., Lowe, C. G. bioaccumulation process of mercury in the (2017) Inuence of ontogeny and environ- benthic ecosystem in the enclosed coastal mental exposure on mercury accumulation seas. J Environ Saf, 10, 99–108. in muscle and liver of male Round Sting- 47) Toba, M. (2004) The decline of manila clam rays. Marine Environmental Research, stock in Tokyo Bay. Bull Fish Res Agen, 130, 30–37. 1(Supplement), 13–18. 45) Taniuchi, T., Shimizu, M. (1993) Dental 48) National Institute for Minamata Disease, sexual dimorphisim and food habits in the Ministry of Environment (2013) Minama- stingray Dasyatis akajei from Tokyo bay, ta Disease Archives. Mercury and health. Japan. Nippon Suisan Gakkaishi, 59(1), http://www.nimd.go.jp/english/kenkyu/ 53–60. docs/Mercury_and_health.pdf

— 25 —