Introduction

Continental shelves have generally been characterized as areas where biological productivity is much higher than that in the open ocean, due to the enormous input of terrigenous nutrients (see Walsh 1987). The East China Sea possesses a wide continental shelf shallower than 200 m, occupying approximately 70% of the total sea area. Two major rivers, viz. Changjiang and Huanghe, drain into the shelf region. This is, thus, one of the most productive areas in the world’s oceans.

Studies conducted as a part of the MASFLEX program (Tsunogai et al. 2003), which particularly focused on biochemical cycles in the East China Sea, clearly revealed that biomass and productivity of phytoplankton are highest during the spring (e.g. Furuya et al. 1998). Satellite images obtained by the coastal zone color scanner (CZCS) and sea-viewing wide field of view sensor (SeaWiFS) have also shown a prolonged period of high concentration of phytoplankton pigments between March and June (Takagi et al. 1993; J. Ishizaka, personal communication).

By subjecting the pigment data obtained by high-performance liquid chromatography (HPLC) analyses to interpretation by CHEMTAX, Furuya et al. (2003) estimated the contribution of various phytoplankton classes to the total concentration of chlorophyll a from the Okinawa Islands to the inner continental shelf in the East China Sea during the spring of 1996. Eight major phytoplankton groups were classified, viz. prochlorophytes, cyanobacteria, chrysophytes, chlorophytes, prymnesiophytes, cryptophytes, dinoflagellates and diatoms. Among them, diatoms were observed to be the most abundant group on the shelf, contributing 40–66% of the total chlorophyll a, followed by chlorophytes, cryptophytes, chrysophytes and primnesiophytes.

Many diatoms have been recorded as producing “resting spores” and “resting cells” (hereafter referred as “resting stages” for both types) during their life histories (Hargraves and French 1983; Garrison 1984). Numerous studies have reported that resting stages play crucial roles in survival under adverse conditions for vegetative cells and in seeding blooms (see McQuoid and Hobson 1996; Itakura et al. 1997). Since light is an essential factor in inducing germination of diatom resting stages (Hollibaugh et al. 1981; French and Hargraves 1985; Imai et al. 1996), resuspension of bottom sediments into the water column must be an important process for the cells to recruit as vegetative forms (see McQuoid and Hobson 1996; Itakura et al. 1997).

On the shelf of the East China Sea, winter monsoon, storm and tidal currents often resuspend bottom sediments throughout the water column (Milliman et al. 1985; Tanaka et al. 1987; Tanaka 1992; Iseki et al. 1995; Furuya et al. 1996). This resuspension in the East China Sea makes it a good location to study the role of the resting stages. On the basis of sediment incubations in laboratory, Ishikawa et al. (2001) postulated that the resting stages on the shelf are important for seeding the bloom. In this study we investigated the abundance and species composition of the diatoms in the spring bloom of 1996 in the East China Sea, along with enumeration of the viable resting stages in the sediments on the shelf. As a whole, this study was designed to evaluate the importance of diatom resting stages in the onset of the bloom and eventually to draw the relationship between diatom bloom formation and hydrographic conditions in the East China Sea.

Materials and methods

Field observations and sample collection

Physical oceanographic observations and water sampling were carried out at 10 stations along the PN line in the East China Sea during the SE96–01 cruise of T/S Seisui Maru (Mie University) from 15 April 1996 to 19 April 1996 (Fig. 1). Although the PN line originally comprised 12 stations (e.g. Furuya et al. 1996), stations PN-2 and PN-12 were not surveyed in this cruise. As described in Furuya et al. (2003), stations PN-1, PN-3 and PN-4 are located off the continental shelf, station PN-5 at the shelf edge, and stations PN-6 through PN-11 on the shelf. A Kuroshio water mass was observed around PN-4. Water temperature, salinity and density were obtained using a Niel Brown Instrument Systems CTD (mark III) at stations PN-1 and PN-4, and a Sea Bird electronic CTD at other stations (i.e. PN-3, and PN-5 through PN-11). Seawater samples were collected at all the stations, using rosette Niskin bottles fitted to the CTD.

Fig. 1
figure 1

Location of sampling stations on the PN line in the East China Sea

Full size image

Sediment samples were also obtained from seven stations (i.e. PN-5 through PN-11) on the shelf and at the shelf edge with a Smith-McIntyre grab sampler. Four sub-samples were taken from the bucket of the sampler, using polycarbonate coring tubes of 3.5 cm diameter. The top 1 cm of each core was then sliced off and pooled together into a sample bottle to avoid small-scale heterogeneity in the distribution of resting stages within a sampled area (26×23 cm). These samples were brought to the laboratory and stored at 4°C in the dark.

Phytoplankton biomass and composition

For estimation of chlorophyll a, a 200-ml aliquot of seawater was filtered through a Whatman GF/F glass fiber filter (25 mm) at <100 mm Hg pressure. The filters were soaked in N, N-dimethylformamide, kept at −20°C (Suzuki and Ishimaru 1990) for pigment extraction and the chlorophyll a concentration was determined fluorometrically (Parsons et al. 1984) using a Turner Design fluorometer (Model 10-005).

For species identification and cell counts of diatoms, seawater samples were immediately fixed by adding borax-buffered formaldehyde at a final concentration of 2% (v/v). To investigate the horizontal distribution of diatoms, samples collected from three depths (0, 10 and 40 m) along the PN line were used for microscopic observation. In addition, at stations PN-8, PN-10 and PN-11, cell counts were particularly made on all samples collected vertically for analysis of the vertical distribution. An aliquot (50–100 ml) of the water sample was settled in an Utermöhl chamber for at least 24 h and the cells counted under an inverted microscope at 200× magnification.

During this cruise, analyses of phytoplankton pigments using HPLC, and nutrients (nitrate and phosphate) were carried out simultaneously (Furuya et al. 2003).

Enumeration of viable diatom resting stages in the sediments

Sediment samples stored at 4°C in the dark for more than 3 months in the laboratory were used for enumeration of viable resting stages of planktonic diatoms in the sediments. The samples were incubated following the extinction dilution method [most probable number method (MPN)] (Imai et al. 1984). This method allows us to enumerate the number of viable (i.e. able to germinate) resting stages in the sediment. Wet sediment (1 g) was suspended in 10 ml of f/2 medium (Guillard and Ryther 1962) to make a suspension at a concentration of 0.1 g wet sediment ml-1 (100 dilution). This suspension was subsequently diluted with f/2 medium and dilutions from 10-2 to 10-5 were obtained. Each 1-ml aliquot of the diluted samples (10-2–10-5) was incubated in five wells on a 24-well tissue culture plate (Corning) at 15°C under illumination at 170 μmol m-2 s-1 with a 12 h light/12 h dark photoperiod. The temperature maintained during the incubation simulated the average surface temperatures recorded in situ between stations PN-5 and PN-11, and light intensity during the incubation experiment was similar to levels within the euphotic zone at these stations (Ishikawa et al. 2001). The number of positive wells where vegetative cells germinated was counted on the seventh day using an inverted microscope. The most probable number of the resting stages in the sediments (MPN g-1 wet sediment) was then calculated from a statistical table (Throndsen 1978) and the number was finally converted into density cm-3 of wet sediment by measuring the specific gravity of the original sediment.

Results

Hydrographic conditions

Detailed hydrographic conditions along the PN line during the cruise are described in Furuya et al. (2003). In brief, the water mass along the line was distinguished as being of two types, viz. a less saline (32.7–34.3) and cold water mass (<15°C), and a saline (34.0–34.5) and warmer (16–22°C) mass that was either the Kuroshio water or modified Kuroshio water. The former water mass was present on the shelf west of station PN-7, and the latter water mass occurred off the shelf, at the shelf edge, station PN-5, and station PN-6. Thus, the boundary between the two water masses was located between stations PN-6 and PN-7. These water masses are referred to here as “shelf water” and “Kuroshio water” respectively. The shelf water was weakly stratified at stations PN-8, PN-9 and PN-11(Fig. 2). A marked vertical gradient of σ t was observed at depths of 10–16 m at station PN-9 and at depths of 30–38 m at station PN-11 with Δσ t /Δdepth values of 0.023 and 0.065, respectively. At station PN-8, σ t increased steadily along with depth from the surface. These observations indicate vertical stability in the upper water column. The pycnocline was mainly associated with the halocline (Furuya et al. 2003). In contrast, no stratification in the water column was observed at station PN-10. Concentration of nitrate was high in the shelf water, varying from 1.9 to 8.0 μM, whereas it was mostly <1 μM in the Kuroshio water (Furuya et al. 2003).

Fig. 2
figure 2

Distribution of σ t along the PN line during the period 15 April 1996 to 19 April 1996

Full size image

Chlorophyll a

In the shelf water, chlorophyll a was generally >1.0 μg l-1 in surface waters at stations PN-8, PN-9 and PN-11 (Fig. 3). The highest concentration of chlorophyll a (2.75 μg l-1) was observed at 20 m at station PN-8. However, at station PN-10, where the water column was not stratified, the concentration was consistently low (<0.5 μg l-1) throughout the water column. In the Kuroshio water, chlorophyll a was always low (usually <0.5 μg l-1), although a chlorophyll a patch (>1.0 μg l-1) was found at around 30 m at the shelf edge (station PN-5) (Fig. 3). The results of chlorophyll a detected by the fluorometric method in this study corresponded to those detected by HPLC analysis (Furuya et al. 2003).

Fig. 3
figure 3

Distribution of chlorophyll a along the PN line during the period 15 April 1996 to 19 April 1996

Full size image

Phytoplankton community

The abundance of diatoms varied horizontally along the PN line (Fig. 4). In the shelf water, at 0 and 10 m at stations PN-7, PN-8, PN-9 and PN-11, where less saline and cold water dominated and the pycnocline was formed, diatoms were abundant (>104 cells l-1). In contrast, at station PN-10, where the water column was mixed, the cell concentration was low and uniform throughout the water column (3.0–3.3×103 cells l-1). In the Kuroshio water, the cell concentration was generally low, except at station PN-5, where saline and warmer water predominated. A maximum concentration of 1.0×105 cells l-1 was observed at 10-m depth at station PN-8. In this study, 111 species belonging to 37 genera of diatoms were identified. In the shelf water, the diatom community was dominated by Chaetoceros debilis and three other Chaetoceros species, Thalassiosira sp. 1, Pseudo-nitzschia pungens and Paralia sulcata (Fig. 4). Each of these species contributed >10% to the total number of diatoms. In the waters where diatom concentration exceeded 104 cells l-1, C. debilis was dominant (mean=54±25%), followed by Thalassiosira sp. 1 (mean=12±17%) and Pseudo-nitzschia pungens (mean=7±8%). The contribution of C. debilis to the total diatoms was particularly high (72–83%) at 0- and 10-m depths at stations PN-8 and PN-11. Paralia sulcata, another characteristic species in the shelf water, comprised 34–60% of the diatom community throughout the water column at station PN-10, and 29% and 89% at 40-m depth at stations PN-9 and PN-11, respectively. In contrast, east of station PN-6, the dominant species varied among the stations with no consistent trend (Fig. 4). At station PN-5 where the diatoms were abundant (>104 cells l-1), 78–100% of the diatoms had a relative abundance less than 10%, indicating a high species diversity.

Fig. 4
figure 4

Variations in diatom abundance (upper panel) and relative abundance of dominant species to the total diatom numbers (lower panel) at 0-, 10- and 40-m depths along the PN line during the study period. Species showing relative abundance >10% of total diatoms are considered dominant species

Full size image

At stations PN-8 and PN-11 (Fig. 5), diatoms showed higher abundance (104–105 cells l-1) above 30-m depth and decreased below this depth. C. debilis was the dominant species, contributing 57–80% of the diatom community above 50 m at station PN-8, and 76–86% above 30 m at station PN-11. Concentration of diatoms was relatively uniform ranging between 1.5–4.9×103 cells l-1 throughout the water column at station PN-10, where P. sulcata dominated (34–80%) the diatom community (Fig. 5). This species was also a dominant component below 60-m (18–44%) and 30-m (48–89%) depths at stations PN-8 and PN-11, respectively.

Fig. 5
figure 5

Vertical distributions of diatom abundance (left panel) and relative abundance of dominant species within total diatom numbers (right panel) at stations PN-8, PN-10 and PN-11 during the study period. Species showing relative abundance >10% of total diatoms are considered dominant species

Full size image

In general, along the PN line, diatom abundance was significantly correlated with chlorophyll a (r 2=0.66, P<0.001).

Diatom resting stages in the sediments

Seven diatom taxa, C. debilis, C. curvisetus, unidentified Chaetoceros spp., Skeletonema costatum, Thalassiosira rotula, Thalassiosira sp. 1 and Thalassiosira sp. 2, were detected in the sediment incubation based on the extinction dilution method. Since C. debilis, T. rotula and Thalassiosira sp. 1 were frequently observed in the experiment, the MPNs of these three species were selectively enumerated (Fig. 6). The resting stages of C. debilis were distributed in the sediments on the shelf west of station PN-6. C. debilis resting stages were relatively abundant (around 4×103 cells cm-3) at stations PN-7, PN-10 and PN-11, with a maximum number of 4.4×103 cells cm-3 observed at station PN-11. Cells of this species were not detected at station PN-5. The resting stages of both the Thalassiosira species, with an exception at station PN-10 where the cells of T. rotula were not detected, were widely distributed in the sediments over the shelf and its edge, and the cell number ranged between 2.9×102 and 2.1×103, and between 3.0×102 and 1.9×103 cells cm-3 for T. rotula and Thalassiosira sp. 1, respectively.

Fig. 6
figure 6

Chaetoceros debilis, Thalassiosira rotula, Thalassiosira sp. 1. Horizontal distribution of the numbers of viable resting stages in the sediments on the continental shelf along the PN line during the study period

Full size image

Discussion

In the shelf water between stations PN-7 and PN-11, the chlorophyll a concentration observed in spring (Fig. 3) was much higher than is usually observed in other seasons (see Furuya et al. 1996; Hama et al. 1997), indicating that spring bloom occurred on the shelf during this study period. Based on the HPLC analyses and subsequent CHEMTAX interpretation of the pigment data, Furuya et al. (2003) reported that the spring bloom in 1996 in the shelf water of the East China Sea was mostly made up of diatoms.

The present study reveals that diatom resting stages were distributed in the sediment on the shelf (Fig. 6). Since light is essential for their germination (Hollibaugh et al. 1981) and the bottom depths from stations PN-5 to PN-11 are deep (53–128 m), germination of resting stages must follow resuspension. Furuya et al. (1996) reported that, during winter, the water column in the East China Sea was well mixed and the benthic diatom, P. sulcata, dominated throughout the water column, proposing that the occurrence of this species near the surface is a good indicator of enhanced vertical mixing. This species was reported to be widely distributed on the shelf sediments of the PN line (Asaoka 1975; Furuya et al. 1996). In the present study, direct microscopic observation of the sediments also revealed that the abundance of P. sulcata cells in the sediments ranged from 4.4×103 cells cm-3 to 3.8×104 cells cm-3 along the PN line on the shelf (data not shown). Furthermore, uniform distribution of P. sulcata was found from the bottom to the surface at station PN-10 (Fig. 5), suggesting that vertical mixing processes were sufficient to resuspend benthic diatoms and any resting stages of planktonic diatoms. In contrast, the water column at other stations on the shelf, west of stations PN-7, started to be weakly stratified (Fig. 2). These hydrographical conditions on the shelf coincide with those generally hypothesized for diatom resting stages to promote blooms, namely resting stages in the sediments are resuspended throughout the water column by vertical mixing and germination of theses cells occur when the mixed layer becomes shallower, potentially resulting in the formation of diatom blooms (see McQuoid and Hobson 1996).

The present study further revealed microscopically that the main component of the diatom community in the shelf water was C. debilis (Figs. 4, 5), whose resting stages were also the most abundant in the sediments on the shelf (Fig. 6). C. debilis is a species distributed mainly in cooler water (Hasle and Syvertsen 1997). Asaoka (1975) reported the occurrence of the vegetative cells of this species in winter of 1963. Although no data on the diatom community in winter of 1996 is available, this species might occur even at low concentration in the water column. If so, the vegetative cells possibly originated from the resting stages, which could have been resuspended to the surface light field from the bottom sediments after late autumn. However, the hydrographic conditions in winter did not allow the vegetative cells to build a large population because cells would have been mixed below the critical depth. In contrast, stratification in early spring reduced the depth of the mixed layer and kept the germinated cells above the euphotic layer, a more favorable condition for the growth of this species. This is most probably the reason that the diatom bloom led by C. debilis occurred in the upper layer of the stratified stations on the shelf during early spring. Thus, this study strongly supports the view proposed by Ishikawa et al. (2001) that diatom resting stages resuspended in winter seed the onset of the spring bloom on the shelf of the East China Sea.

During this study period, concentration of chlorophyll a at the surface shallower than 10 m at station PN-8 increased drastically from 2 μg l-1 in the day time (0900 hours) to 9 μg l-1 in the night time (2100 hours) (Furuya et al. 1998). The night community was dominated by C. debilis, contributing 95% of the total diatom cells (8.2×105 cells l-1) (data not shown). The rapid increase of the biomass, being equivalent to a growth rate of 3.0 day-1, cannot be explained only by the net growth rate (0.54 day-1) which was estimated by the “dilution method” (Landry and Hassett 1982) for this spring phytoplankton community at 0 m at station PN-8 (Furuya et al. 1998). Therefore, increased diatom cell concentration within a short period was probably the result of combined biological growth and advection of cells. Concurrent with the rapid increase in biomass was a decrease of nitrate from 3 μM during the day to 0.1 μM during the night (Furuya et al. 2003), suggesting the possibility that the diatom bloom would decline immediately after this study period. However, the observations of phytoplankton pigments by CZCS and SeaWiFS have shown a prolonged period of high phytoplankton abundance from March through June (Takagi et al. 1993; J. Ishizaka, personal communication). This may suggest that phytoplankton blooms are repeated in the East China Sea from spring to the early summer.

On the shelf of the East China Sea, typhoon, storm and tidal currents cause intermittent mixing in the water column (Milliman et al. 1985; Tanaka et al. 1987; Tanaka 1992; Iseki et al. 1995). This mixing supplies the sea surface with nutrients from deep waters and resting stages from the sediments. The intermittent mixing enables the resting stages of various species, such as C. curvisetus, C. debilis, S. costatum, T. rotula, Thalassiosira sp. 1 and others, to exploit the favorable conditions at the sea surface, light, temperature and photoperiod, for their germination. Since vegetative growth after germination is largely controlled by complex factors, e.g. water temperature, light intensity, nutrient concentrations, competition with other phytoplankton and grazing by zooplankton, vegetative growth can then develop into a bloom not only in spring but also in other seasons if such environmental factors are favorable when the water column is stratified again. The shelf of the East China Sea may also receive allochthonous phytoplankton cells via water from the Kuroshio current (Kawarada et al. 1966, 1968; Guo 1991). Once they are brought in, the stratified waters of the shelf may promote the growth of these allochthonous cells. Phytoplankton growth on the shelf may also be enhanced by introduction of nutrients from Changiang River (Furuya et al. 1998). The effects of the riverine nutrient enrichment, however, are limited to local areas around the estuary. Over the growing season, meteorological and topographical characteristics in the East China Sea lead to complex hydrographic conditions that might prolong the bloom period from spring to early summer. Variations in hydrography may also lead to change in the dominance of autochthonous and allochthonous species in the phytoplankton.