Detailed Observation of Spatial Abundance of Clam Larva Ruditapes philippinarum in Tokyo Bay, Central Japan

Journal of Oceanography, Vol. 60, pp. 631 to 636, 2004 Short Contribution Detailed Observation of Spatial Abundance of Clam Larva Ruditapes philippinarum in Tokyo Bay, Central Japan TOMOYUKI KASUYA 1, MASAMI HAMAGUCHI 2 and KEITA FURUKAWA 3 * 1 Corporation for Advanced Transport & Technology, Uchisaiwai-cho, Chiyoda-ku, Tokyo 100-0011, Japan 2 National Research Institute of Fisheries and Environment of Inland Sea, Maruishi, Ohno-cho, Saeki-gun, Hiroshima 739-0452, Japan 3 National Institute for Land and Infrastructure Management, Nagase, Yokosuka-shi, Kanagawa 239-0826, Japan (Received 2 September 2002; in revised form 26 June 2003; accepted 26 June 2003) The spatial distribution of the larval abundance of the clam Ruditapes philippinarum has been investigated at 65 stations throughout Tokyo Bay on August 2, 2001. The large number of small D-shaped larvae that were found shortly after hatching in the waters around the Banzu, Futtu, and Sanmaizu-Haneda areas indicates that spawning populations in these areas probably contribute greatly to the larval supply in the bay. Small larvae also occurred abundantly around the Yokohama and Ichihara port areas, suggesting that these port regions play a role in the transport of larvae into Tokyo Bay. Keywords: Bivalves, Ruditapes philippinarum, plankton, larva, reproduction, Tokyo Bay. 1. Introduction The clam Ruditapes philippinarum is one of the most commercially important food bivalves for the Japanese. It is abundant on the sand-mud sediments of tidal flats and shallows in regions stretching from Hokkaido to Kyushu. Recently, however, clam catches have decreased in Tokyo Bay as well as in other coastal and inlet waters areas in Japan and this decrease may be due to the destruction of habitats as coastal areas become further developed (Kakino, 1992). As most of the tidal flats and shallows in Tokyo Bay have disappeared, restoration of a habitat for R. philippinarum in the bay is being investigated. This could include the construction of artificial shallows or tidal flats, and protecting the clam s larval supply areas, but to do this successfully it is necessary to understand R. philippinarum larval advection during the planktonic stage, i.e. where do they come from and where * Corresponding author. E-mail: furukawa-k92y2@ysk.nilim. go.jp Present address: National Institute for Land and Infrastructure Management, Nagase, Yokosuka, Kanagawa 239-0826, Japan. Copyright The Oceanographic Society of Japan. do they go? Because the identification of larval bivalves is rather difficult, however, little is known about the larval abundance and advection of the clam (Matsumura et al., 2001). As a first step in clarifying the ecology of the early life stages of R. philippinarum in Tokyo Bay, this paper briefly describes the spatial distribution and size composition of the larvae, and then discusses the spawning areas of R. philippinarum in the bay. We also discuss the validity of estimating larval abundance using a 100- µm mesh net. 2. Materials and Methods 2.1 Sampling Tokyo Bay is located along the Pacific coast of Honshu, central Japan. The bay is divided into inner and outer areas by a boundary line connecting Futtu with Kannonzaki (Fig. 1). In this paper, Tokyo Bay means the inner part of the bay, covering a 960-km 2 surface area with a 15-m mean depth. In Tokyo Bay, R. philippinarum spawns twice a year, with peak periods from April to June and August to October (Toba et al., 1993). Samples were collected at 65 stations spaced at intervals of ca. 3.5 km in Tokyo Bay on August 2, 2001 (Fig. 1). We collected the samples within 631

AST-500) and water density (σt) was calculated from those data. Fig. 1. Tokyo Bay and locations of sampling stations (solid circles), where stations surrounded by a dashed line were investigated with one boat. The main habitats of benthic Ruditapes philippinarum in the bay are shaded and the dotted line at the mouth of Tokyo Bay defines the inner part of the bay. a five-hour period during daylight using eight boats. The stations were grouped into eight groups, and each group was covered by one boat. As hypoxia/anoxia occurs widely in the bottom layer of Tokyo Bay during summer (Furota, 1988), no sampling was conducted at depths below 12 m. We were able to collect the samples evenly at specific depth layers (0 4 m, 4 8 m, and 8 12 m) by repeatedly raising and lowering a weighted flexible hose attached to a water pump. We collected planktonic larvae of R. philippinarum by pumping up and filtering 200 liters of seawater from the three depth layers, first through a 100-µm mesh net and then through a 50-µm one to collect the smaller larvae. The number of depth layers sampled varied depending on the depth of each station (Table 1). After all the sampling had been completed, the samples collected using the 100-µm net were immediately cooled and divided into four sub-samples on land using a plankton splitter. Two of the four sub-samples were preserved in 5% buffered formalin, and the others were cooled before storage in a freezer (< 45 C) in the laboratory. The 50-µm net samples were also cooled immediately and stored in the same freezer. Temperatures and salinities were determined at each station using an STD recorder (Alec Electronics Co. Ltd., 2.2 Identification of larval Ruditapes philippinarum Most bivalves develop into a trochophore and then a veliger larva with a shell after hatching. Veliger larvae are classified into D-shaped larvae with a straight hinge line, umbo larva, or fully grown larva according to the stage of development. A fully grown larva grows into a juvenile clam and settles into a habitat. In this paper we identified R. philippinarum planktonic larvae using the veliger larvae because veliger larvae with shells are firmer than trochophore larvae and seem to maintain their shape after filtering. We used the monoclonal antibody method (cf. Hamaguchi et al., 1997) to identify R. philippinarum planktonic larvae in the frozen 50-µm and 100-µm net sub-samples. After exposing the thawed specimens to antibodies, they were observed using a fluorescence microscope. We counted the larvae, recorded their developmental stage (i.e., D-shaped or umbo larvae) and measured the shell length (SL) of up to a maximum of 100 specimens using an eyepiece micrometer accurate to the nearest 10 µm. Since fully grown larvae are morphologically similar to umbo larvae, we counted them as umbo larvae. The monoclonal antibody method achieved 95% accuracy in distinguishing larval R. philippinarum from other larvae in the Seto Inland Sea (Hamaguchi, pers. obs.). The clam Paphia undulata, which is closely related to R. philippinarum, was found in Tokyo Bay (Kuwabara, 1990) and the antibodies also crossreact with this clam (Hamaguchi, pers. obs.). In the present study, larvae that were difficult to identify were finally identified using the fluorescence microscope as R. philippinarum larva by conducting a polymerase chain reaction (PCR) test for D-shaped larvae and by morphologic observations according to Tanaka (1982) for umbo larvae. We then calculated the density of the R. philippinarum planktonic larvae (ind. m 3 ). 3. Results and Discussion 3.1 Abundance and size distribution of Ruditapes philippinarum planktonic larvae We obtained a large number of R. philippinarum larvae. Of all the bivalve larvae collected, 0.1 to 24.2% belonged to R. philippinarum. The SL of the collected planktonic larvae of R. philippinarum was in the range of 90 130 µm for D-shaped larvae and 130 230 µm for umbo larvae. Larvae with a SL of 90 210 µm, mainly 100 120 µm, were found in the 50-µm net sample and those with an SL of 130 230 µm, mainly 150 180 µm, were in the 100-µm net sub-sample (Fig. 2). The size frequency distribution of R. philippinarum larvae of the 50- and the 632 T. Kasuya et al.

Table 1. Densities of D-shaped (bold numbers) and umbo larvae (numbers in parentheses) of Ruditapes philippinarum calculated from 50-µm net samples and 100-µm net sub-samples. The symbol ( ) means no sampling. Stn. Depth (m) Density (ind. m 3 ) Stn. Depth (m) Density (ind. m 3 ) 0 4 m 4 8 m 8 12 m 0 4 m 4 8 m 8 12 m 1 4 320 (150) 34 23 360 (225) 310 (170) 200 (85) 2 6 290 (225) 35 10 200 (105) 515 (35) 3 7 60 (0) 36 5 955 (890) 4 6 20 (185) 37 18 430 (315) 290 (280) 550 (115) 5 13 420 (115) 255 (115) 175 (35) 38 25 260 (165) 380 (205) 750 (300) 6 11 145 (185) 410 (70) 39 28 5 (45) 15 (0) 210 (5) 7 9 55 (360) 145 (270) 40 23 225 (105) 160 (60) 670 (335) 8 5 50 (100) 41 18 155 (85) 240 (45) 160 (125) 9 9 65 (120) 230 (300) 42 30 145 (25) 130 (40) 250 (75) 10 11 250 (20) 345 (75) 43 25 235 (50) 200 (30) 260 (20) 11 13 160 (160) 115 (165) 390 (5) 44 16 825 (55) 390 (45) 575 (120) 12 14 510 (325) 845 (215) 190 (0) 45 14 1725 (365) 2510 (1100) 13 13 285 (210) 325 (70) 55 (70) 46 6 105 (350) 14 9 165 (65) 195 (0) 47 10 170 (260) 295 (110) 15 17 30 (40) 620 (435) 835 (60) 48 20 210 (125) 225 (65) 530 (65) 16 16 165 (80) 285 (110) 545 (365) 49 30 90 (25) 295 (5) 510 (140) 17 14 230 (25) 465 (105) 940 (110) 50 22 165 (45) 165 (45) 360 (270) 18 11 10 (80) 185 (40) 51 27 570 (40) 930 (5) 560 (30) 19 6 1710 (70) 52 25 55 (40) 220 (70) 215 (300) 20 11 2180 (35) 1085 (170) 53 20 90 (25) 80 (0) 105 (25) 21 14 330 (110) 775 (385) 375 (90) 54 10 710 (90) 260 (65) 22 17 95 (120) 140 (180) 680 (275) 55 11 805 (435) 830 (60) 23 19 195 (15) 460 (40) 110 (40) 56 4 920 (325) 24 18 70 (245) 190 (335) 210 (775) 57 15 715 (200) 1030 (110) 1010 (20) 25 14 85 (45) 510 (110) 285 (520) 58 15 165 (0) 115 (20) 235 (45) 26 15 245 (340) 785 (1525) 455 (125) 59 34 770 (85) 320 (25) 200 (150) 27 14 150 (775) 385 (465) 90 (275) 60 43 125 (20) 125 (60) 265 (55) 28 21 120 (20) 210 (145) 230 (225) 61 43 80 (60) 100 (25) 110 (25) 29 21 95 (20) 60 (20) 60 (40) 62 17 375 (240) 225 (90) 940 (115) 30 17 275 (5) 215 (180) 285 (310) 63 4 370 (565) 31 13 1165 (595) 510 (275) 260 (30) 64 28 190 (75) 335 (75) 350 (80) 32 21 345 (285) 570 (200) 385 (390) 65 19 390 (50) 315 (50) 390 (35) 33 24 75 (65) 170 (70) 265 (285) 100-µm-net specimens indicated that <140-µm-SL larvae might pass through a 100-µm mesh, i.e. 70% of a SL. D- shaped larvae of R. philippinarum generally exceed 100- µm SL (Toba, 1992), indicating that we could accurately estimate the abundance of D-shaped and umbo larvae by using a 50-µm mesh net after water sampling. R. philippinarum planktonic larvae were found at almost all stations (Fig. 3), indicating that they disperse widely during the planktonic stage. D-shaped larvae were abundant in the Sanmaizu-Haneda, Banzu, and Futtu areas with a maximum density of 2510 ind. m 3, and umbo larvae were abundant in the Ichihara, Banzu, and Futtu areas with a maximum density of 1525 ind. m 3 (Table 1, Fig. 3). R. philippinarum planktonic larvae were fairly evenly distributed vertically at 0 to 12 m in Tokyo Bay, and no relationship was found between the horizontal distribution of the larvae and water temperature, salinity, and σt at three depths (Figs. 3 and 4). R. philippinarum develops from a fertilized egg to a 100-µm SL larva within two days at 20 C in the laboratory (Toba, 1992), and settles within about 2 3 weeks after hatching at an SL of around 200 µm (cf. Toba, 1987). Because the D-shaped larvae, mainly in the 100 120-µm SL size class, appear to have been collected shortly after hatching, their distribution indicates a R. philippinarum spawning area in Tokyo Bay. In contrast, the umbo larvae collected, mainly in the 150 180-µm SL size class, probably drifted for several days after hatching and remained in a planktonic stage for a while. As R. philippinarum planktonic larvae certainly disperse widely during the planktonic stage in Tokyo Bay (see above), it remains unclear whether larvae in the 150 180-µm SL size class were spawned in Distribution of Ruditapes philippinarum Planktonic Larvae 633

Fig. 2. Size frequency distributions of Ruditapes philippinarum larvae collected using a 50-µm and 100-µm nets on August 2, 2001. The data are pooled from all stations, where n is the number of measured larvae. the Ichihara, Banzu, and Futtu areas or were transported there from other spawning areas. In Mikawa Bay, Japan, R. philippinarum planktonic larvae are mainly found at a 3-m depth (Suzuki et al., 2002). In the Seto Inland Sea, Japan, R. philippinarum planktonic larvae were concentrated at a depth layer of 30 of salinity in the water column when the salinity at the surface fell due to heavy rain (Hamaguchi, pers. obs.). In Tokyo Bay, as pycnoclines mostly existed at depths ranging from 8 to 15 m in August 2, 2001, R. philippinarum planktonic larvae may possibly be transported passively in a mixing layer in Tokyo Bay. To add to the data gathered from these field observations, the larval transport processes of R. philippinarum should be analyzed using a hydrodynamic numeric model in a future study. 3.2 Spawning area of Ruditapes philippinarum The horizontal distribution of D-shaped larvae with a 100-µm SL, obtained by multiplying the density (Table 1) by the size frequency of the respective samples, showed that densities of >500 ind. m 3 occurred around the Sanmaizu-Haneda and Banzu areas where there are habitats of benthic R. philippinarum in Tokyo Bay (Fig. 5). Larvae with a 100-µm SL also occurred at a density of 200 300 ind. m 3 around the Yokohama, Futtu, and Ichihara areas. As R. philippinarum develops from a fertilized egg to a 100-µm SL larva within two days at 20 C in the laboratory (Toba, 1992), larvae with a 100-µm SL may remain within the area (Banzu, Futtu, Sanmaizu- Haneda, Ichihara, and Yokohama) where they are spawned. The Banzu and Futtu areas are shallows and tidal flats, in an almost natural condition, and R. philippinarum occurs densely in these areas (Kakino, 1992). Their reproduction might lead to a high abundance of planktonic larva in these areas. In the Sanmaizu-Haneda area, some Fig. 3. Horizontal density distributions of Ruditapes philippinarum D-shaped and umbo larvae from three depth layers on August 2, 2001. Densities are proportional to the area of the circle. natural tidal flats remain, and some artificial shallows have been constructed. R. philippinarum occurs abundantly around Sanmaizu (Furota, pers. comm.) and around the mouth of the Tamagawa River (Kuwabara, 1990). Although there have been few quantitative studies of the abundance of R. philippinarum in the Sanmaizu-Haneda area, the density of small larvae in this area is almost equal to that in the Banzu area. The spawning populations in the Sanmaizu-Haneda area, as well as in the Banzu area, probably contribute greatly to the larval supply in Tokyo Bay. The Yokohama and Ichihara areas are ports protected almost completely by a vertical sea wall. In the Ichihara area, R. philippinarum appears in the sandy bottom adjacent to the sea wall (Toba, pers. comm.). R. philippinarum has also been collected from the port region around Yokohama (Kuwabara, 1990); these port regions might also play a role in the larval transport of R. philippinarum into Tokyo Bay. 4. Conclusion In this study we have shown the validity of our method of estimating the abundance of R. philippinarum 634 T. Kasuya et al.

Fig. 4. Horizontal distributions of water temperature, salinity, and σt at 0-, 4-, and 8-m depths. larvae using a 50-µm mesh net to filter samples taken at a range of depths. Our observations of the horizontal distribution of small D-shaped R. philippinarum larvae have revealed some spawning areas of the clam, including a port region in Tokyo Bay. Recently, it has been realized that shallows and tidal flats have a water purification function, in addition to their role as a nursery for juveniles, and the construction of artificial shallows or tidal flats around ports has been planned. Artificial shallows and tidal flats may increase the habitat of R. philippinarum, and could potentially contribute to restoring the R. philippinarum resource. Nevertheless, although port regions have not been recognized as a habitat and spawning area for fishery organisms, the abundance and distribution of organisms in the port area should be closely investigated before constructing shallows or tidal flats in Tokyo Bay. In this experiment, other zooplankters found in the formalin-fixed 100-µm net sub-samples, including the heterotrophic dinoflagellate Noctiluca scintillans, were also identified and counted. In addition, we made similar observations, including measuring the dissolved oxygen content, on August 6 and 10, 2001, in Tokyo Bay. We intend to combine these data to clarify the factors affecting the distribution and abundance of R. philippinarum planktonic larvae, i.e. hydrography, hypoxic/anoxic water mass, predation by carnivorous plankters, and larval transport processes in Tokyo Bay. Other spawning areas, e.g. the Sanbanse and Kanazawa areas, where benthic R. philippinarum occurs abundantly (Fig. 1), will also be evaluated. Fig. 5. Horizontal density distributions of Ruditapes philippinarum D-shaped larvae with a 100-µm SL from three depth layers on August 2, 2001. The density of 100- µm-sl larvae was obtained by multiplying the density (Table 1) by the size-frequency of respective samples. Acknowledgements We thank the staff of the Marine Environment Division at the National Institute for Land and Infrastructure Management for their collaboration in collecting samples, and we greatly appreciate the advice given by Dr. Mitsuharu Toba. We also thank two anonymous reviewers for their helpful comments on the manuscript. This study was partly supported by the Program for Promoting Fundamental Transport Technology Research from the Corporation for Advanced Transport & Technology (CATT). References Furota, T. (1988): Effects of low-oxygen water on benthic and sessile animal communities in Tokyo Bay. Bull. Coastal Oceanogr., 25, 104 113 (in Japanese). Hamaguchi, M., H. Usuki and H. Ishioka (1997): Interrelationship of Japanese littleneck clam, Ruditapes philippinarum, and other animals on tidal flat. Fisheries Eng., 33, 201 211 (in Japanese). Distribution of Ruditapes philippinarum Planktonic Larvae 635

Kakino, J. (1992): Recent situation on the Japanese littleneck fisheries. Fisheries Eng., 29, 31 39 (in Japanese). Kuwabara, R. (1990): Sandy and muddy fauna of macrobenthos in the inner area of Tokyo Bay. J. Agricultural Sci., The Tokyo University of Agriculture, 35, 152 166 (in Japanese with English abstract). Matsumura, T., S. Okamoto, N. Kuroda and M. Hamaguchi (2001): Temporal and spatial distributions of planktonic larvae of the clam Ruditapes philippinarum in Mikawa Bay; application of an immunofluorescence identification method. Japanese J. Benthol., 56, 1 8 (in Japanese with English abstract). Suzuki, T., T. Ichikawa and M. Momoi (2002): The approach to predict sources of pelagic bivalve larvae supplied to tidal flat areas by receptor mode model: a modeling study conducted in Mikawa Bay. Bull. Japan Soc. Fish. Oceanogr., 66, 88 101 (in Japanese with English abstract). Tanaka, Y. (1982): Identification of bivalve larvae (16). Aqua biology, 18, 23 26 (in Japanese). Toba, M. (1987): On the seedling production of short-necked clam Ruditapes philippinarum, Adams et Reeve (Bivalvia) I. Bull. Chiba Pref. Fish. Exp. Sta., 45, 41 48 (in Japanese). Toba, M. (1992): Relationship between temperature and larval growth rate in manila clam Ruditapes philippinarum. Bull. Chiba Pref. Fish. Exp. Sta., 50, 17 20. Toba, M., Y. Natsume and H. Yamakawa (1993): Reproductive cycles of manila clam collected from Funabashi waters, Tokyo Bay. Nippon Suisan Gakkaishi, 59, 15 22 (in Japanese with English abstract). 636 T. Kasuya et al.