Observations on the maintenance mechanisms of metapopulations, with special reference to the early reproductive process of the manila clam Ruditapes philippinarum (Adams & Reeve) in Tokyo Bay.

ABSTRACT To obtain basic ecological information on declining Tokyo Bay stocks of the Manila clam Ruditapes philippinarum, life history (including spawning, larval transport and settlement, and growth and survival of benthic clams) was investigated in local habitats, mainly in Sanbanse and Kisarazu. Spawning-stage adult clams were histologically detected from May to October at both sites, with two peak periods in summer and autumn. From May to November, between 13 and 23 larval cohorts appeared annually in each site, even during periods when spawning-stage adults were not detected. Spawning over many months plus larval transport from other habitats may contribute to the continuous appearance of larvae at both sites. This confirms the hypothesis that the clam population is maintained and supported by the reproduction of local populations. High density cohorts of newly settled clams, which later formed adult populations, may originate from larval cohorts spawned during peak spawning periods. A number of summer cohorts of benthic juveniles disappeared within 34 mo after settlement, whereas autumn cohorts maintained a relatively high density through the following spring, and subsequently they formed adult populations. Severe mortality resulting from extreme environmental factors may have affected the summer cohorts during summer and autumn. The recent decline of the adult stocks may be related to the unstable recruitment of summer cohorts, which was one of the two major seasonal components contributing to recruitment.

KEY WORDS: Manila clam, Ruditapes, metapopulation, larval supply, benthic abundance

INTRODUCTION

Fisheries production of the Manila clam Ruditapes philippinarum (Adams & Reeve) has greatly declined in Japan since the 1980s. Identifying the causes and proposing a remedy are important not only for the fishing industry but also for recreational harvest (Tomizuka 2004). Locally, the decline has not been the result of mass mortality caused by environmental disasters such as anoxic and hypoxic seawater (Tsuda & Kawai 1980, Kakino 1986), flood events (Yamashita 1988), and poor food availability (Kakino et al. 1992), but rather because of the long-term reduction in natural recruitment of juvenile clams (Toba 2004). Therefore, early life-history studies on reproduction of Manila clams have been conducted (Sekiguchi 2003).

However, the ecological information on early life-history of the manila clam is limited, but some recently initiated studies are utilizing the development of a rapid and mass identification method of clam larvae using species-specific fluorescent antibody staining (Hamaguchi 1999). Hence, studies on larval dynamics, such as seasonal patterns of appearance, transportation with tidal advection and larval transport to distant habitats, have been reported (Suzuki et al. 2002, Kasuya et al. 2003a, 2003b). Based on these results, the significance of larval exchange among local habitats in maintaining the clam population has now been studied (Furota 2002, Hinata 2003). Thus, in general, the adult population maintenance mechanism is recognized as a function of the interrelationship among several local populations, or metapopulations (Harrison & Taylor 1997), comprising the whole population of a bay area (Hamaguchi 2005).

To understand the maintenance mechanism in the reproduction of metapopulations, information is required on the entire life-history, beginning with spawning, larval dynamics, and including settlement and development of adult stocks. Whereas most studies have focused on the ecological characteristics of the planktonic larval stage (Hinoshita 1995, Yamada et al. 1996, Iwasaki et al. 1999, Matsumura et al. 2001, Nasu 2004) and settlement of benthic clams (Sekiguchi et al. 1995, Miyawaki & Sekiguchi 2000, Ishii et al. 2001), the relationship between spawning and larval supply remains poorly described. The present study attempts to describe the entire life history of the Manila clam, including adult spawning, transport of larvae and settlement of benthic clams, and growth and survival of adults in Tokyo Bay.

MATERIALS AND METHODS

Study Sites

The major habitats of the clams in the Kanto area, located on the central Pacific coast of Japan, are limited to the inner area of Tokyo Bay. Three sites, where intensive commercial harvesting of clams occurs, were chosen for this study. They are located at some distance from each other along the coast of the inner area of Tokyo Bay (Fig. 1).

[FIGURE 1 OMITTED]

The first site, Sanbanse, is a shallow coastal area located at the northernmost end of Tokyo Bay. The area is an inlet, and is surrounded by artificially reclaimed land with one dredged channel in the center and another on the eastern side. Only a part of the sandy flat, which extends 4 km onshore to offshore, is exposed at mean low water spring (MLWS), with a total tidal amplitude of 2.0 m. In Sanbanse, larval clams were sampled at 3 stations (st. A-C: -3 to--m-depth at MLWS) and benthic clams at 10 stations (st.1-10: [+ or -] 0 to -1.2-m depth at MLWS).

Banzu is a tidal flat, located at the Obitsu River estuary on the east coast of Tokyo Bay, consisting of the Obitsu River mouth, dredged channels and sandy flats, with a length of 12 km and offshore width of 1.0-1.5 km at MLWS. The second site, Kisarazu, is adjacent to the river mouth, which flows into Tokyo Bay in southern Banzu. Sampling of larval and benthic clams was carried out, as described earlier, at 10 stations (larval clams: st.D-F: -3 to -10 m and benthic clams: st. 11-17: [+ or -] 0.6 to [+ or -] 0 m). Stations were arranged from onshore to offshore (Fig. 1).

The third station, Futz, is the tidal flat closest to the mouth of Tokyo Bay, and is located north of Cape Futtsu. The length and offshore width of the tidal flat is approximately 3 km and 1 km, respectively, at MLWS. Adult clam samples were obtained from 2 stations (st.18 and 19) located at the tidal and subtidal areas, respectively.

From April 2001 to July 2004, larval and benthic clams were collected from Sanbanse and Kisarazu. The collection was carried out biweekly during April and December and monthly during the rest of the year. Adult clams (SL [anterior/posterior shell length] > 30 mm) were sampled at 5 stations (st.3, 13, 16, 18, and 19).

Sampling of Larval and Benthic Clams

Planktonic samples were collected by sampling with a submerged pump suspended from a vessel. At each station, 160-320 L of seawater was pumped from the surface (surface--0.3 m) and bottom (bottom + 0.5 m) within 2-4 min. Because additional sampling layers were added every 2-3 m when the water depth was more than 4 m, a total of 2-5 layers were sampled at each station. At all stations, water temperature and salinity were measured from the surface to the bottom at every 1-m depth by using a portable apparatus (Model 85D; YSI Inc., Ohio). The samples were filtered through a 50-[micro]m mesh-nylon net on the vessel, and residuals were transported chilled on ice to the laboratory on the sampling day. Larval samples, from which coarse matter was removed with a 200-[micro]m mesh filter, were stored at -80[degrees]C until staining for microscopic examination.

Three sediment core samples (46-mm diameter, >5 mm-depth) for newly settled clams and 10 samples (68-mm diameter, >10-cm depth) for large clams were collected at each station. Samples for the newly settled clams were preserved until sorting by adding 1% formalin with 0.1% rose bengal, and that of the large clams by adding 10% formalin to the samples after sieving with a 0.5-mm mesh.

Sample Treatments

Prior to microscopic examination, thawed subsamples of clam larvae were stained with FITC conjugated goat antimice monoclonal antibody that is specific to the velum protein of Manila clam larvae (Hamaguchi 1999). The bivalve larvae that showed fluorescence for the velum under the fluorescence microscope (BKF-II; Nikon, Tokyo) were identified as Manila clam larvae. SL of the identified Manila clam larvae was measured using an ocular micrometer.

The newly settled clams were manually separated from the sediment samples under a dissecting microscope after removing silt and clay components with a 0.072-mm mesh sieve. (Yanagibashi 1992). The following procedure was used for sorting. Whole sediment samples from one station were mixed with tap water and placed in a beaker. After stirring, the mixture was allowed to stand for approximate 1 min. The surface material of the settled contents was transferred to a Petri-dish using a capillary pipette to pick the clams. This procedure was repeated until no clams could be found in two successive operations. Large clams were separated under a loupe.

Condition Factor and Gonadal Observations

Adult clams sampled at st.3 in Sanbanse, st.13 and 16 in Kisarazu, and st.18 and 19 in Futtsu were preserved in 10% formalin for evaluating the condition factor (CF) and histological gonadal development. CF, a ratio of soft tissue weight to shell size, was calculated using the wet meat weight and shell dimensions (Toba et al. 1993).

Histological tissue sections of samples collected from st.3, 13, and 16 from April to November 2003 were observed for gonadal development. The fixed abdomens of at least 20 clams obtained on each sampling date were embedded in paraffin, sectioned transversely and observed under an optical microscope after staining with Mayer hematoxylin and eosin. Gonad developmental stages were classified into 6 categories according to Toba and Miyama (1991); undifferentiated, early active, late active, spawning, spent, and regressive. Certain clams were included as "parasitized," because they appeared to spawn abnormally because of a serious cercaria infection of a digenetic trematode.

Data Analysis

Size frequency distributions of larval and benthic clams from each sampling date, which were obtained as an average of size frequencies from all the sampled layers and stations, were analyzed by the least squares method to identify cohorts. Assuming that the components of cohorts were normally distributed, calculations to divide the observed polymodal distributions into several normal distributions were performed using the software Solver (MS-Excel 2003, Microsoft Corp.) as reported in Gorie (2002). The previously reported growth rate of natural clam populations (Shibata 2004) was referenced to determine the initial parameters to be applied to the calculations. Cohorts whose density was estimated at < 10ind./100 L in larva and 5ind./1,000 [cm.sup.2] in benthic clams were excluded from further analysis.

Mean water temperatures on each sampling date at the study sites were the averages of the water temperatures measured at multiple layers in the larval sampling stations. Seasonal daily changes in the water temperatures at the study sites were estimated, assuming that water temperatures changed linearly between the sampling dates. Growth rates and spawning dates of the identified larval cohorts were calculated according to the experimentally obtained larval growth rate (Toba 1992) using mean SL of the cohorts and the estimated daily water temperatures. Any cohorts estimated to be spawned within two successive days were treated as identical cohorts in further analysis.

RESULTS

Seasonal Changes in Water Temperature and Salinity

The annual water temperature recorded in Sanbanse (8[degrees]C to 9[degrees]C in February and 27[degrees]C to 28[degrees]C in August to September) showed greater variation than that of Kisarazu (9[degrees]C to 10[degrees]C and 25[degrees]C to 26[degrees]C) from 2001-2003 (Fig. 2).

[FIGURE 2 OMITTED]

Salinity in Sanbanse varied from 27-28 psu to 22-23 psu, with a sharp temporary decline to 15 18 psu in summer. In Kisarazu, the annual salinity was stable and ranged from 25-29 psu, except for a slight decline to 20-25 psu in summer.

Seasonal Changes in CF and Gonadal Development

The CF of clams from Sanbanse and Kisarazu was lowest in winter and highest during spring and summer (Fig. 3). The CF of clams from Sanbanse (st.3) showed similar bimodal seasonal changes from 2001-2003, and attained an annual maximum peak (CF = 22-23) in March to April and a lower secondary peak (CF = 16) from August to September. In Kisarazu (st. 13 and 16), the first peak (CF = 17-19) was delayed until June to July, and a secondary peak was not observed from 2001-2003. Peak periods of CF in Futtsu (st.18 and 19) fluctuated annually between March and June, with minor peak periods in June to October.

[FIGURE 3 OMITTED]

In Sanbanse (st.3), all the female gonads, excluding the parasitized clams, were in late active stage from April to early June in 2003 (Fig. 4A). Annual first spawning, revealed by the appearance of individuals in spawning stage, was observed from mid-June to early July. Thereafter, a second spawning period was initiated in August after the temporary cessation of spawning in late July. The second spawning period continued until late November with dual peak periods represented by the prevalence of spawning stage from late August to early September and October. The synchronization of the peak spawning periods was determined by gonadal observations and periods of statistically significant decline in CF (from mid-May to mid-August and from early September to early November, P < 0.05) (Fig. 3). These observations indicated that the major spawning of the clam population occurred during these periods.

[FIGURE 4 OMITTED]

Onset of annual first spawning in 2003 at the onshore side of the tidal flat in Kisarazu (st. 13) was at the end of June, peaking in late July (Fig. 4B). After a 2-too resting period, the second spawning period began in late September and continued until mid-October. On the other hand, the spawning period at the offshore side in Kisarazu (st.16) continued from late June to late October, with no resting period in 2003 (Fig. 4C). Peak spawning periods were observed in early August and September. Similarly, in Sanbanse the peak periods of spawning in Kisarazu generally overlapped with the periods of statistically significant decline in CF (from mid-September to mid-October in st. 13 and from mid-September to early November in st. 16).

Gonadal observations showed that vacant areas among matured oocytes resulting from partial spawning of oocytes were often observed in spawning-stage clams at both Sanbanse and Kisarazu (Fig. 5).

[FIGURE 5 OMITTED]

Larval Abundance and Estimated Spawning Dates

In Sanbanse, clam larvae were present in plankton samples throughout the warm season (May to November) during 2001 and 2003 (Fig. 6). Larval densities during these periods varied greatly between 2-463, 1-5504, and 2-1999 ind./100 L in 20012003 respectively. The peak larval densities were recorded two or three times annually; August and November in 2001, May and October to November in 2002, and July and August to September in 2003.

[FIGURE 6 OMITTED]

Clam larvae were documented from April to November in Kisarazu, with large fluctuations in densities; 2-1057, 1-5739, and 4-530 ind./100 L in 2001 2003, respectively (Fig. 7). Although two peaks of larval density were recorded annually, the periods of peak density varied between years; September and November in 2001, July and September in 2002, and June and August in 2003.

[FIGURE 7 OMITTED]

The numbers of larval cohorts identified from 2001 2003 were 13, 15, and 23 in Sanbanse and 19, 18, and 21 in Kisarazu, respectively (Fig. 8). Spawning intervals between successive larval cohorts were 3-30 days at both sites during 2001 and 2003, with the exception of one interval of 56 days during July to August 2002 in Sanbanse. Conversely, the calculated spawning periods in both sites coincided during 2001 and 2002; May to October and April to October, respectively. However, in 2003 spawning continued from mid-April to late October in Kisarazu, whereas in Sanbanse, it lasted from late May to early October. Comparisons with the histological data from adult clams indicated that several larval cohorts were estimated to have been spawned during the period when the spawning-stage adult clams were not observed at each site, that is from early June to late July and mid-April to late May in Sanbanse (Fig. 4).

[FIGURE 8 OMITTED]

Abundance of Benthic Clams

From 2001 2003, 17 cohorts of benthic clams were identified in Sanbanse (S1-17) (Fig. 9) and 18 in Kisarazu (Kl-18) (Fig. 10). In Sanbanse, the number of newly settled cohorts from 2001-2003 were 4(S3-6), 3(S8-10), and 5(S11-15), and in Kisarazu, 3(K3-5), 5(K6-10), and 7(K 11-17), respectively. The newly settled cohorts were observed mainly during May to October, and this was when the larvae were transported to these areas.

[FIGURES 9-10 OMITTED]

The initial density of almost all the newly settled cohorts in Sanbanse was [10.sup.1.0]-[10.sup.2.3] ind./1,000 [cm.sup.2] during 2001-2003, excluding S10, which had greater recruitment ([10.sup.3.7] ind./1,000 [cm.sup.2]) in October 2002 (Fig. 9A). Although no significant difference was detected (P < 0.05) between the initial density of the cohorts that settled in summer ([10.sup.0.9]-[10.sup.2.1] ind./1,000 [cm.sup.2]; S3-6, 8-9, 11-14) and autumn ([10.sup.1.5]-[10.sup.1.9] ind./1,000 [cm.sup.2], S7 and 15), density of the summer cohorts were greatly reduced to <[10.sup.1.0] ind./1,000 [cm.sup.2] within 2-3 mo after settlement.

In Kisarazu, a majority of the initial cohort densities were between [10.sup.2.0]-[10.sup.3.0] ind./1000 [cm.sup.2] (Fig. 10A). Similarly, in Sanbanse the initial density of summer cohorts ([10.sup.1.6] - [10.sup.3.1] ind./1,000 [cm.sup.2]; K3-4, 6-9, 12-16) did not differ significantly from those of the autumn cohorts (K5, 10, 17). Whereas the density of the summer cohorts was greatly reduced to approximately 10[degrees] ind./1,000 [cm.sup.2] within 2-4 mo after settlement, the autumn cohorts maintained their density at [10.sup.2]ind./ 1,000 [cm.sup.2] until the following spring.

Growth After Settlement

Six cohorts of benthic clams (S7, 8, 10, 11 in Sanbanse and K5, 10 in Kisarazu) were successively identified with a mean SL of 20 mm (Fig. 9B, 10B). Of these, 2 cohorts (S8, 10) had settled in summer, and the remaining 4 cohorts settled from October to November.

In Sanbanse, summer cohorts (S8, 11) settled in May to June, grew rapidly to >5 mm in mean SL during August and reached 20 mm in November (Fig. 9B). Autumn cohorts (S7, 10) that settled from October to November grew little until March of the following year. The cohorts began to grow after April, reached 20 mm in mean SL from August to September, and they entered a period of little growth in November, at a SL of 25-30 mm. Growth resumed in the following April, and in July the cohorts attained a mean SL of 35 mm.

Summer cohorts in Kisarazu (K3, 6) and in Sanbanse grew to 5 mm in August, but later growth ceased and the cohorts reached 15-20 mm in mean SL the following spring (Fig. 10B). Autumn cohorts (K5, 10) maintained a SL of <1 mm until the following spring. The active growing period in Kisarazu was shorter than that in Sanbanse. In Kisarazu, the cohorts initiated growth in May, stopped growing in October at approximately 20 mm in SL, and resumed growth the following May.

During 2001-2003, the total number of the identified cohorts of larvae, newly settled clams, and adult clams (>20 mm in mean SL) were 51, 12, and 4 in Sanbanse and 58, 15, and 2 in Kisarazu, respectively. Eight percent of the larval cohorts in Sanbanse successfully recruited and grew to adults, whereas 3 % did so in Kisarazu (Table 1).

DISCUSSION

Spawning of Adult Clams and Appearance of Larvae

The histology of female adult gonads revealed that at both Sanbanse and Kisarazu, spawning-stage clams appeared during the study period from June to November, with two peaks in summer and autumn. Whereas similar observations have been previously reported in Sanbanse (Toba et al. 1993), this study has further revealed that the clam population in Kisarazu had the same pattern of spawning activity characterized by successive spawning that continued for 6 mo during the warm season. The extended spawning activity from late spring to autumn has also been observed in the clams in Brittany coast, France (Beninger & Lucas 1984, Laruelle et al. 1994) and in Lagoon of Venice, Italy (Meneghetti et al. 2004).

The histology determined that partial spawning activity occurred in the spawning stage, indicating that a clam spawned intermittently and discharged some of the matured oocytes from the gonad. Certain physiological properties, such as intermittent spawning of individual clams, may contribute to the successive spawning pattern of clam populations in Tokyo Bay as in other areas (Laruelle et al. 1994, Sbrenna & Campioni 1994).

The peak periods of spawning at both sites showed discrepancy during 2003. The influence of external environmental conditions, such as water temperature and food availability, on growth rate of clam gonads has been experimentally confirmed (Mann 1979, Toba 1989, Toba & Miyama 1995). Differences in the seasonal water temperature at both sites (Fig. 2) may be an important factor responsible for the diversity of the seasonal spawning pattern.

Additionally, the gonadal assessment also revealed differences in the duration of the peak spawning period in the Kisarazu tidal area between the offshore side (st.16, +0.2 m in MLWS) and the onshore side (st.13, +0.6 m). On the offshore side, abundant food is indicated by the presence of photosynthetic pigment, present in the water (Shibata et al. 1999). Sea water exchange resulting from tidal currents and resuspension of organic detritus by wave action may both promote clam growth (Nishizawa et al. 1992). The seasonal reproductive patterns were dissimilar between the two adjacent lagoons located in the same estuary, possibly because of the different trophic conditions resulting from artificial eutrophication (Sbrenna & Campioni 1994).

Decline of CF in the clam population may result from both spawning events and temporary fluctuation of environmental trophic factors. However, the decrease of nutritional content in the body tissue during spawning has been experimentally recognized in the Manila clam (Mann 1979, Beninger & Lucas 1984), and decline in the ratio of meat to shell size during spawning has been noted in wild populations (Toba et al. 1993, Laruelle et al. 1994). Assuming that CF declines significantly during peak spawning, we suggest that during 2001-2003, the spawning periods varied among major local habitats in Tokyo Bay, because the periods of decline in CF differed among the four sites, including the tidal and subtidal areas, in Futtsu (Fig. 3).

Therefore, successive spawning over several months and the regional diversity of spawning periods caused by the variation in local environmental factors all result in abundant and extended periods of larval appearance in Tokyo Bay during the warm season. Moreover, seasonal variation in reproductive activity associated with growth of the adult clam (Tsuji et al. 1993) may result in further diversity in spawning patterns.

Because of a sampling error, which resulted in a 2-wk interval of no larval collection, the actual number of transported larval cohorts may have been larger than that observed during this study. Nevertheless, the estimated intervals between the spawning dates of larval cohorts observed from May to October ranged from a few days to less than a month. The frequent transport of newly spawned larval cohorts at both sites apparently resulted from the successive spawning patterns of the adults.

In Sanbanse and Kisarazu, the fact that larval cohorts appeared during periods when the adults were not in spawning condition indicated the input of larvae from other sites. In addition, the number of identified larval cohorts and their estimated spawning dates were not identical at both sites. This strongly implies that the larval cohorts transported to Sanbanse and Kisarazu were spawned from different spawning populations of adults.

The larvae spawned in the local habitats along the coast of Tokyo Bay dispersed over the entire bay within a few days (Kasuya et al. 2003a, 2003b), and the possible exchange of larvae between the distant habitats was estimated by calculations from a numerical model constructed based on hydrodynamic and meteorological surveys (Hinata 2005). Frequent and long periods of local larval appearance, including periods when resident adult clams were not in spawning condition and the difference in source of larval supply between sites, indicate that the spawning pattern of the adult population plays a substantial role that affects larval supply and exchange within Tokyo Bay. This study in Tokyo Bay confirms for the first time the hypothetical relationship between local seasonal spawning patterns and larval transport (Kasuya et al. 2003a, b, Kasuya et al. 2004, Hinata 2005). The fundamental mechanism for maintenance of the Manila clam population through the reproduction of metapopulations has been substantiated.

The continuous appearance of larvae during the warm season has been reported in most of the principal habitats in Japan; Mikawa Bay (Yamada et al. 1996, Matsumura et al. 2001), Ise Bay (Sekiguchi et al. 1995), Seto Inland Sea (Momoyama & Iwamoto 1979, Hinoshita 1995), and Ariake Sound (Ishii et al. 2001). In addition, histology has shown successive spawning cycles in adult clams with two peak periods from spring to autumn in most of the wild populations in the southern coast of Japan (Hiroshima Pref. 1954, Ko 1957, Momoyama & Iwamoto 1979, Tsuji et al. 1993). Thus, frequent re-supply of larvae from the successive spawning of adult clams may be a common phenomenon within the principal habitats in the southern coast of Japan.

Growth and Survival of the Benthic Clam Population

In Sanbanse and Kisarazu, cohorts of newly settled clams were identified from May to November, corresponding with the presence of larvae. Settled cohorts that grew to a SL of >20 mm (size at sexual maturity) (Holland & Chew 1974, Yap 1977) were estimated to have settled from May to June and October. Principal cohorts of newly settled clams, which resulted in adult populations originated from larval populations supplied during the peak period of spawning. These principal cohorts amounted to <10% of the total number of larval cohorts documented, so cohorts other than these may only partially contribute to the adult population.

Although the larval sampling regimen of this study (time intervals, number, and spatial extent of stations) may not have been sufficient to estimate the entire size of the larval population, the initial density of settlement appears to be related to the overall size of the larval population. This was confirmed by the identification of newly settled cohorts, which corresponded temporally with those of the dense larval cohorts. This may indicate a density dependant relationship between the larval supply and adult occurrence for clams in Tokyo Bay, as well as Ariake Sound (Ishii et al. 2001) and within a Portuguese coastal lagoon (Chicharo & Chicharo 2001).

However, survival of the summer and autumn cohorts differed greatly, particularly in Kisarazu, where summer cohorts disappeared within 3-4 months after settlement (<2-4 mm SL), whereas autumn cohorts maintained a relatively high density after the spring of the following year (SL < 1 mm). Severe mortality from unfavorable environmental factors may have affected clams <2-4 mm during summer to autumn.

In recent years, the adult stock at Sanbanse has decreased drastically, accompanied by poor juvenile recruitment (Toba 2004). In the 1980s, when larger adult stocks were maintained compared with more recently, juvenile clams appear to have originated from both summer and autumn spawnings and successfully recruited (Kakino & Toba 1990). The recent decline of adult stocks may be related to the unstable recruitment of summer cohorts, which was one of the major components of recruitment in the past. Unstable recruitment of summer cohorts was more obvious in Kisarazu, where juvenile clams have been decreasing since the early 1990s (Toba 2004). This fact correlates with a similar observation that recruitment of summer cohorts, which was strong around 1990 (Nishizawa et al. 1992), became unstable in 1994-1998 (Shibata 2004). Therefore, mortality of summer cohorts in both Sanbanse and Kisarazu has become evident from the late 1980s to early 1990s.

In the southern coast of Japan, either summer (Ishii et al. 2001) or autumn cohorts (Tsuji et al. 1996, Miyawaki & Sekiguchi 2000) have been reported to recruit to adult stock. Fisheries production has greatly declined in the Seto Inland Sea and Ariake Sound since the late 1980s. However, substantial recruitment of juvenile clams spawned during autumn was previously documented in the 1950s, prior to the stock decline (Yasuda & Takamori 1952, Ikematsu 1957). Unstable recruitment of summer cohorts may not necessarily be directly related to the stock decline of the Manila clam.

ACKNOWLEDGMENTS

The authors thank A. Kaneko, H. Tsuchiya, N. Shoji, Y. Shoji, T. Hayashi, and T. Kawashima for their helpful technical assistance in the field survey; Dr. J. Higano, Dr. M. Hamaguchi, Dr. T. Kasuya, Dr. H. Hinata, and Dr. J. Kakino for the valuable discussions. The field work could never have been accomplished without the help of a number of student volunteers. This study was completed under the program of Shellfish Field Investigation in Chiba Pref. Fish. Res. Ctr. collaborating with Tokyo Univ. Mar. Sci. Tech.

LITERATURE CITED

Beninger, P. G. & A. Lucas. 1984. Seasonal variations in condition, reproductive activity, and gross biochemical composition of two species of adult clam reared in a common habitat: Tapes decussatus L. (Jeffreys) and Tapes philippinarum (Adams & Reeve). J. Exp. Mar. Biol. Ecol. 79:19-37.

Chicharo, L. & M. A. Chicharo. 2001. Effects of environmental conditions on planktonic abundances, benthic recruitment and growth rates of the bivalve mollusk Ruditapes decussatus in a Portuguese coastal lagoon. Fish. Res. 53:235-250.

Furota, T. 2002. Importance of network among local population in benthos inhabits bay area. In: Abstracts of Symposium "Problems in environment for organic production which supports Manila clam production, and new view point of research." Fish. Res. Agen. pp. 8-9 (in Japanese).

Gorie, S. 2002. Estimation of parameters in a mixture of normal distributions from length frequency composition and growth formula by MS-Excel. Suisanzoshoku 50:243-249 (in Japanese).

Hamaguchi, M. 1999. Development of species-specific monoclonal antibodies technique to Ruditapes philippinarum larvae for field survey. In: Report on development of laboratory techniques in species identification for early ecological study of finfish and shellfish. Tokyo: Agri. Forest. Fish. Res. Council. pp. 66-77 (in Japanese).

Hamaguchi, M. 2005. New method to analyze metapopulation dynamics. Kaiyo Monthly 37:125-132 (in Japanese).

Harrison, S. & A. D. Taylor. 1997. Empirical evidence for metapopulation dynamics. In: I. Hanski, editor. Metapopulation biology. Oxford: Oxford University Press. pp. 27-42.

Hinata, H. 2003. Restorative strategy of tide land based on understanding of larval dynamics of Manila clam in whole bay area, for Tokyo Bay. In: Abstracts of Symposium "Emergency in tide land ecosystem-present status and scheme for restoration." Joint Council of the Societies Concerning Coastal Environment. pp. 81-82 (in Japanese).

Hinata, H. 2005. Numerical simulation on advective process of planktonic larvae of the clam Ruditapes philippinarum in Tokyo Bay. Bull. Fish. Res. Agen. (Suppl. 3):59-66 (in Japanese).

Hinoshita, Y. 1995. Investigation on appearance of juvenile Manila clam. Ann. Rep. Ooita Shallow Water Fish. Exp. Stn. in 1993. pp. 41-44 (in Japanese).

Hiroshima Pref. 1954. On the spawning period in Manila clam. Hiroshima Suishi Dayori 35:9-11 (in Japanese).

Holland, D. A. & K. K. Chew. 1974. Reproductive cycle of the Manila clam (Venerupis japonica), from Food Canal, Washington. Proc. Natl. Shellfish. Assoc. 64:53-58.

Ikematsu, W. 1957. Ecological study on the clam, Tapes japonica (REEVE) II; on the setting season and the growth in early young stage. Bull. Jap. Soc. Sci. Fish. 22:736-741 (in Japanese).

Ishii, R., H. Sekiguchi, Y. Nakahara & Y. Jinnai. 2001. Larval recruitment of the Manila clam Ruditapes philippinarum in Ariake Sound, Southern Japan. Fish. Sci. 67:579-951.

Iwasaki, K., N. Kuroda, S. Okamoto & T. Matsumura. 1999. Investigation on developing a promotive technique for a function of Manila clam fishery area. Report on the investigation on maintenance and development of coastal fishery area in 1998. Aichi Pref. Fish. Exp. Stn. 18 pp (in Japanese).

Kakino, J. 1986. The cases of accidental mortality of shellfish in Tokyo Bay, especially on the influence of anoxic sea water. Fish. Eng. 23:41-47 (in Japanese).

Kakino, J. & M. Toba. 1990. Characteristics of Manila clam stock in shellfish fishery ground in Northern Chiba District. Bull. Chiba Pref. Fish. Exp. Stn. 48:5%71 (in Japanese).

Kakino, J., M. Toba, A. Kaneko & Y. Miyama. 1992. Characteristics in winter mortality of Manila clam in Kisarazu, Tokyo Bay. Bull. Chiba Pref. Fish. Exp. Stn. 50:21 30 (in Japanese).

Kasuya, T., M. Hamaguchi, K. Furukawa & H. Hinata. 2003a. Short-term spatial and temporal variations in abundance and size-frequency distribution of planktonic larvae of clam Ruditapes philippinarum in Tokyo Bay. Res. Rep. Nat. Int. Land Infrastr. Mang. 8. 13 pp (in Japanese).

Kasuya, T., M. Hamaguchi, K. Furukawa & H. Hinata. 2003b. Larval abundance, distribution, and size composition of planktonic larvae of the clam Ruditapes philippinarum in the fall season in Tokyo Bay. Res. Rep. Nat. Int. Land Infrastr. Mang. 12. 12 pp (in Japanese).

Kasuya, T., M. Hamaguchi & K. Furukawa. 2004. Detailed observation of spatial abundance of clam larva Ruditapes philippinarum in Tokyo Bay, Central Japan. J. Oceanogr. 60:631-636.

Ko, Y. 1957. Some histological notes on the gonads of Tapes japonica DESHAYES. Bull. Jap. Soc. Sci. Fish. 23:394-399 (in Japanese).

Laruelle, F., J. Guillou & Y. M. Paulet. 1994. Reproductive pattern of the clams, Ruditapes decussatus and R. philippinarum on intertidal flats in Brittany. J. Mar. Biol. Assoc. UK. 74:351-366.

Mann, R. 1979. The effect of the temperature on the sexual maturation in Manila clam, Ruditapes philippinarum (Adams & Reeve). J. Exp. Mar. Biol. Ecol. 38:121-133.

Matsumura, T., S. Okamoto, N. Kuroda & M. Hamaguchi. 2001. Temporal and spatial distributions of planktonic larvae of the clam Ruditapes philippinarum in Mikawa Bay; application of an immunofluorescence identification method. Jap. J. Benthol. 56:1-8 (in Japanese).

Meneghetti, F., V. Moschino & L. Da Ros. 2004. Gametogenic cycle and variations in oocyte size of Tapes philippinarum from the Lagoon of Venice. Aquaculture 240:473-488.

Miyawaki, D. & H. Sekiguchi. 2000. Long-term observations on larval recruitment processes of bivalve assemblages on temperate tidal flat. Benthos Res. 55:1-16.

Momoyama, K. & T. Iwamoto. 1979. On the spawning period of Manila clam in Yamaguchi Bay and Oomi Bay. Bull. Yamaguchi Inland Fish. Exp. Stn. 7:19-34 (in Japanese).

Nasu, H. 2004. Present situation and future theme in Manila clam stock in Kumamoto coastal area in Ariake Sound. Seasonal Report of Marino-Forum 21.49:25-34 (in Japanese).

Nishizawa, T., J. Kakino, K. Nakata & K. Taguchi. 1992. Growth and mortality of Manila clam Ruditapes philippinarum in Banzu tidal flat, Tokyo Bay. Fish. Eng. 29:61-68 (in Japanese).

Sbrenna, G. & D. Campioni. 1994. Gametogenic and spawning patterns of Manila clams Tapes philippinarum (Bivalvia: Veneroida) in two lagoons of the river Po Delta, Italy. J. Shellfish Res. 13:37-46.

Sekiguchi, H., M. Uchida & A. Sakai. 1995. Post-settlement processes determining the features of bivalve assemblages in tidal flat. Benthos Res. 49:1-14.

Sekiguchi, H. 2003. Environmental disturbance in Ariake Sound, on the causes of drastic decline in fisheries production of Manila clam in Ariake Sound. Oceanogr. Jap. 12:21 36 (in Japanese).

Shibata, T., M. Toba, M. Sakai & A. Kaneko. 1999. Availability of photo-pigment as an index of productivity in the culture ground of Japanese little neck clam Ruditapes philippinarum. Bull. Chiba Pref. Fish. Exp. Stn. 55:67-72 (in Japanese).

Shibata, T. 2004. Settlement, growth and survival of the juvenile of Manila clam, Ruditapes philippinarum, on Banzu tidal flat in Tokyo Bay. Bull. Chiba Pref. Fish. Res. Ctr. 3:57-62 (in Japanese).

Suzuki, T., T. Ichikawa & 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. Jap. Soc. Fish. Oceanogr. 66:88-101 (in Japanese).

Toba, M. 1989. Significance of amount of ingested food on the sexual maturation of short-necked clam Ruditapes philippinarum Adams et Reeve (Pelecypoda) rearing in the tank. Suisanzoshoku. 37:63 69 (in Japanese).

Toba, M. & Y. Miyama. 1991. Gonadal development and spawning induction in artificially conditioned Manila clams Ruditapes philippinarum. Nippon Suisan Gakkai Shi. 57:126%1275 (in Japanese).

Toba, M. 1992. Relationship between temperature and larval growth rate in Manila clam Ruditapes philippinarum. Bull. Chiba Pref. Fish. Exp. Stn. 50:17-20 (in Japanese).

Toba, M., Y. Natsume & H. Yamakawa. 1993. Reproductive cycles of Manila clam collected from Funabashi waters, Tokyo Bay. Nippon Suisan Gakkai Shi. 59:15-22 (in Japanese).

Toba, M. & Y. Miyama. 1995. Influence of temperature on the sexual maturation in Manila clam Ruditapes philippinarum. Suisanzoshoku 43:305-314.

Toba, M. 2004. The decline of Manila clam stock in Tokyo Bay. Bull. Fish. Res. Agen. (Suppl.1):13-18.

Tomizuka, N. 2004. Establishment of "National Conference on Asari Clam Resource." Fish. Sci. 70:225-230 (in Japanese).

Tsuda, H. & H. Kawai. 1980 On the causes of weakening and mortality of shellfish in the estuary of Kiso Rivers. Ann. Rep. Mie Isewan Fish. Exp. Stn. in 1978. pp. 72-83 (in Japanese).

Tsuji, S., M. Munekiyo, M. Itani & A. Doke. 1993. Reproductive cycles of a Manila clam in Maizuru Bay. Bull. Kyoto Mar. Ctr. 17:1-9 (in Japanese).

Tsuji, S., M. Munekiyo, M. Itani & A. Doke. 1996. Settlement, growth and disappearance of spats of Manila clam Ruditapes philippinarum in Maizuru Bay. Suisanzoshoku. 44:25-30 (in Japanese).

Yamada, S., Y. Iwata & T. Yanagisawa. 1996. Spatial distribution of Manila clam larvae in Mikawa Bay--transporting, dispersing, recruiting processes. Kaiyo Monthly 28:150-156 (in Japanese).

Yamashita, T. 1988. Abnormal mass mortality caused by river flood during rainy season. Bull. Seikai Reg. Algae Shellfish Assoc. 5:61-64 (in Japanese).

Yanagibashi, S. 1992. Selective settlement Japanese littleneck larvae and their distribution in Mikawa Bay. Fish. Eng. 29:55-99 (in Japanese).

Yap, W. G. 1977. Population biology of the Japanese little-neck clam, Tapes philippinarum in Kaneohe Bay, Oahu, Hawaiian Islands. Pacif. Sci. 31:223-244.

Yasuda, J. & S. Takamori. 1952. Study on seeds of Asari (Venerupis philippinarum). Bull. Inland Reg. Fish. Res. Inst. 2:1-11 (in Japanese).

TABLE 1.

Number of the identified cohorts of larva, settled juvenile, and
adult in Manila clam Ruditapes philippinarum during
2001-2003 in Sanbanse and Kisarazu.

                        Sanbanse

              2001    2002    2003    Total (%)

Larva          13      15      23        51
Settled
  juvenile      4       3       5        12
Adult *         1       2       1         4

                        Kisarazu

              2001    2002    2003    Total (%)

Larva          19      18      21        58
Settled
  juvenile      3       5       7        15
Adult *         1       1       0         2 (3)

* Shell length >20 mm.

MITSUHARU TOBA, (1) * HIROSHI YAMAKAWA, (2) YUTAKA KOBAYASHI, (1) YOSHIO SUGIURA, (2) KEN HONMA (2) AND HIDETOSHI YAMADA (2)

(1) Tokyo Bay Fisheries Laboratory, Chiba Prefectural Fisheries Research Center, Kokubo 3091, Chiba 293-0042, Japan; (2) Tokyo University of Marine Science and Technology, Konan 4-5-7, Minato, Tokyo 108-8477, Japan

* Corresponding author. E-mail: [e-mail]