The spatial and seasonal dynamics of total and size-fractionated phytoplankton biomass (chlorophyll- a) as well as physical and chemical factors in the Yangtze River Estuary and adjacent East China Sea coastal waters were investigated from April 2002 to February 2003. Average surface total and water column integrated chlorophyll a biomass showed a clear seasonal variation in response to the Yangtze River discharge, with the highest in summer (∼4 mg m−3 and >60 mg m−2), intermediate in spring and autumn (∼1 mg m−3 and 26–28 mg m−2), and the lowest in winter (0.5 mg m−3 and <20 mg m−2). Summer maximum chlorophyll a concentrations (>10 mg m−3) occurred at intermediate salinities (∼20–30) region beyond the front zone between 112.5°E and 123°E with sufficient nutrients replenishment for phytoplankton growth. Generally, spatial distribution of size-fractionated phytoplankton showed that phytoplankton biomass was dominated by the large size fraction (>20 μm) in the turbid eutrophic estuarine and near-shore waters, while the small-sized phytoplankton (<5 μm) were dominant in the offshore stations. Phosphate was the main limiting nutrient of phytoplankton biomass in river diluted water and most near-shore stations, while dissolved inorganic nitrogen became the potential limiting nutrient in some offshore stations, except for summer when phosphate limited almost all in the whole investigation region. Controlling the inputs of phosphate loading from the Yangtze River is one of the most effective strategies for reducing the increasing eutrophication and occurrences of harmful algal blooms in Yangtze River Estuary and adjacent East China Sea coastal waters.
Being the main primary producers, phytoplankton are the base of the marine food chain and provide a vital link between inorganic compounds and organic matter available to higher trophic levels (Zhou et al., 2004; Yin et al., 2004). Chlorophyll a (Chl a), the main photosynthetic pigment, is an index reflecting the amount of phytoplankton biomass (Song et al., 2008). The size-fractionated phytoplankton biomass (chl a), plays a significant role in the efficiency of export of biogenic carbon from the upper ocean/costal waters to deep layers (Malone, 1980; Takahashi and Bienfang, 1983). In general, the potential export of biogenic carbon is high in ecosystems dominated by large-sized phytoplankton and low in those dominated by small-sized phytoplankton (Cermeño et al., 2005).
The Yangtze River (YR), about 6300 km long, and flowing into the East China Sea (ECS), ranks third in the world in terms of discharge volume (Ning et al., 1988). It provides 85% of the freshwater runoff to the ECS, with a value of 9.32 × 1011 m3 a−1 (Chen et al., 2001). During the dry season (November to April), 29% of the runoff occurs, while 71% appears during the wet season (May to October) (Zhang, 1996). There is evidence that indicates phosphate limitation occurs in the Yangtze River Estuary (YRE), while nitrogen is more limited in offshore waters (Harrison et al., 1990; Chai et al., 2006; Wang et al., 2008). Due to the rapid development in industry and agriculture, the riverine nutrients loads increased in the past few decades (Zhou et al., 2008). This induced increasing occurrences of eutrophication and red tides/harmful algal blooms (HABs) in the YRE and adjacent waters (Chai et al., 2006; Zhou et al., 2008). HABs can produce toxins, provide mechanical harm (spines), produce excess organic matter, deplete oxygen and be responsible for light shading. The spatial extents of HABs have increased from <2000 km2 in late 1980s to >7000 km2 in early 2000s. At the same time, the dominant species has recently shifted from the diatom (Skeletonema costatum) to toxic forms, such as dinoflagellates (Prorocentrum dentatum, Noctiluca scintillans and Alexandrium spp.) (Gao and Song, 2005). Furthermore, the construction of Three Gorges Dam (TGD) has resulted in an increasing concern about the link between nutrients availability and the succession of phytoplankton communities in the YRE and adjacent waters (Song et al., 2006, 2008; Jiao et al., 2007; Zhou et al., 2008).
Past research has documented the phytoplankton biomass in this region (Ning et al., 1988; Li and Luan, 1998; Gao and Song, 2005; Luan et al., 2006), but reports on size-fractionated phytoplankton after the construction of TGD were scarce (Song et al., 2006, 2008), and almost no references on seasonal changes of phytoplankton size distribution were documented. Our goal was to determine the seasonal variations of total and size fraction phytoplankton biomass (<5 μm, 5–20 μm, and >20 μm) in the YRE and adjacent waters. Their regulations of physical and chemical processes may be important.
Materials and methods
Study sites and sampling
The ECS, the largest continental marginal sea in the western North Pacific Ocean, is strongly influenced by freshwater nutrient enrichment from YR (Gong et al., 1996; Wong et al., 1998), the oligotrophic Kuroshio Current and Taiwan Warm Current (Guan, 1994; Jiao et al., 2007). Four seasonal multidisciplinary investigations were carried out from April 25–May 20, August 26–September 10, November 5–11, 2002, and February 25–March 19, 2003, representing 4 seasonal conditions of spring, summer, autumn and winter, respectively. The study area, including 28 stations, 7 transects, covered YRE and adjacent waters near Zhoushan Islands, ECS (Figure 1).
Water samples were collected at the surface (1 m below), middle, and bottom (2–4 m above the seabed) using Niskin bottles. A multi-parameter YSI 6600 water quality sonde (Yellow Spring, OH, USA) was used to take vertical profiles of temperature and salinity.
Total and size-fractionated chlorophyll a biomass
Water samples (100–250 ml according to phytoplankton density) for total chl a were passed through 0.7 μm Whatman GF/F filters. Samples of 300–1000 ml for size fraction chl a were sequentially filtered through 20 μm, 5 μm Poretics polycarbonate membranes and Whatman GF/F filters under low-vacuum pressure (<100 mm Hg) to collected <5 μm, 5–20 μm and >20 μm size fraction chl a, respectively.
Filters containing chl a were extracted with 10 ml 90% acetone and sonicated for 10 min in an ice-cold water bath, and then, extracted at 4°C in the dark for 24 h. The fluorescence of the extract was measured using a Turner Designs Fluorometer, which was pre-calibrated using pure chl a (Sigma). Water column depth-integrated chl a (Ichl a) concentrations were calculated by trapezoidal integration (Parsons et al., 1984).
Nutrients and criteria for stoichiometric nutrient limitation
Water samples for nutrients analysis were filtered through pre-combusted (450°C, 4 h) Whatman GF/F filters on board and immediately stored in a deep freezer (−20°C) until analyzed. Nutrients including nitrate (NO3−), nitrite (NO2−), ammonium (NH4+), orthophosphate (PO43−) and silicate (SiO43−) were analyzed using a Skalar San Plus Autoanalyzer (Skalar Analytical B V, Breda, the Netherlands) after the colorimetric methods described in JGOFS protocols (Knap et al., 1996).
Dissolved inorganic nitrogen (DIN), DIN = NO3 + NO2 + NH4, PO4 and SiO4 concentrations were used to calculate atomic ratios of N:P, N:Si and Si:P. In a nutrient-replete ecosystem, the atomic N:P:Si ratio of marine diatom is about 16:1:16 (Redfield, 1958). Deviation from the Redfield ratio suggests the potential limitation of N, P or Si for phytoplankton growth (Yin et al., 2001). To assess nutrient stoichiometric limitations, we calculated two ambient nutrient ratios and applied the Redfield ratio to predict: (1) when N:P < 16 and N:Si < 1, N limits; (2) when N:P > 16 and Si:P > 16, P limits; (3) when N:Si > 1, Si:P < 16, Si limits. A similar approach was employed by Justić et al. (1995) to assess nutrient limitation in the Mississippi River Estuary.
Quality control and data statistical analysis
Sampling and analysis bottles or glassware were pre-cleaned by soaking in 10% HCl overnight. The chemicals and solutions used in this study were analytical reagent grade. A one-way analysis of variance (ANOVA) and standard regressions were performed using SPSS 13.0 software. Differences were considered statistically significant when p < 0.05. Contours were obtained with the SURFER 8.0 program (Golden Software).
According to Huang et al. (2001), the statistics of discharge (1950—1993) from the Datong hydrographic station at the lower reach of YR, the closest Gauging Station to the YRE, was <3% of the annual runoff which occurred in February (<1 × 104 m3 s−1), while ∼18% occurred in July (∼5 × 104 m3 s−1) (Figure 2). Forty three percent of annual discharge flowed in summer, only 9% in winter, 23% and 25% in spring and autumn, respectively (Figure 2). The annual rainfall recorded by the Hangzhou Observatory (1951—2008) near YRE was 1412 mm, with the lowest in December (∼50 mm), and highest in June (∼210 mm). On average, ∼37% of annual rainfall occurred during dry season, and 63% during wet season (Figure 2).
Surface temperature showed a regular seasonal fluctuation between the highest (26.6–29.1°C) in summer and the lowest (7.8–14.3°C) in winter (Table 1 and Figure 3). In summer, vertical profile of temperature at transect C (31°N) showed that it decreased gradually with depth, indicating the presence of seasonal stratification. As to spring, autumn and winter, the lowest temperature appeared at the estuary (Figure 4). The temperature of YR diluted water (YRDW) was lower than that of offshore waters. Surface isotherms of autumn and winter were almost parallel to the shoreline (Figure 3).
In spring and autumn, the spatial distribution patterns of YR estuarine plume were similar (Figure 3). The eastward extension range of the river plume was much smaller than that of summer due to the lower runoff and rainfall (Figures 2– 4).
In summer, the tongue of YRDW (salinity <20) expanded eastward farther than 123°E. Surface salinity was <30 in most study area, except in the northeast and southeast stations (Figure 3), indicating the dilution of high YR freshwater runoff (Figure 2). A highly stratified water column was observed, accompanied by a decrease in temperature and an increase in salinity with depth (Figure 4).
In winter, salinity was >30 in the two offshore latitudinal transects (Figure 3). Temperature and salinity kept uniform in the whole water column due to strong winter northeast monsoon induced vertical mixing (Figure 4). YR runoff had little effect on the waters in the east of 123°E, with salinity >32.
In general, DIN and SiO4 decreased gradually from estuary to the offshore waters. All year, both the DIN and SiO4 were significantly correlated with salinity (p < 0.01), indicating the source of DIN and SiO4 was mainly from YR outflow. Dissolved inorganic nitrogen was highest in summer due to the high input of YR outflow, sewage and rainfall. The highest PO4 concentrations were generally recorded at the river mouth station (Stn B1) both in spring (0.9 μM) and summer (1.1 μM). Average PO4 was highest in spring and lowest in summer (Table 1). Results showed that SiO4 concentration was generally low during the transition period from dry to wet season (April–May in spring), while it was high during the transition period from wet to dry season (November in Autumn) (Table 1).
In the spring, nutrient limitation was in the offshore stations of latitudinal transects at 123°E and 123.5°E, were N limited, except for stations A3, A4, C3 and B3, while the near-shore stations were potential P limited. Si only limited at stations A4, D4, G2, G3 and G4. In summer, N:P ratios were >16 in the whole study area (Table 1), and P was limited at two near-shore latitudinal transects and stations C3, D3, E3 and F3 at 123°E, while other offshore stations were N limited. Si was limited at the latitudinal transect of 123.5°E. In autumn, N:Si ratios were <1 and Si:P > 16 in the whole study region (Table 1), and P was limited at stations of two near-shore latitudinal transects and stations B3, C3 and C4, other offshore waters were N limited. Si was not limited at all the stations. In winter, almost all the stations were P limited, except Si limited in stations B4, E3 and E4, and N limited in stations D4, F3 and G4.
Total phytoplankton biomass
Spatial distribution of surface chl a along the salinity gradient showed that the highest phytoplankton biomass did not appear in the turbid estuarine waters with low salinity and the clear oligotrophic offshore waters with high salinity, but among the intermediate salinities (∼20–33) zone from ∼30–33 in spring, ∼20–30 in summer, ∼25–33 in autumn and ∼28–30 in winter (Figure 5). The summer bloom (chl a > 10 mg m−3) was obviously appeared beyond the front zone of YRDW between 122.5°E and 123°E, a stable circumstance suit for phytoplankton growth, with stable stratification occurred in the water column and sufficient nutrients replenishment supplied from the YR outflow (Figures 3–5).
There was a clear seasonal variation in surface chl a, with the highest in summer (∼4 mg m−3), the intermediate in spring and autumn (∼1 mg m−3), and the lowest in winter (0.5 mg m−3) (Table 1). Summer surface chl a was ∼8-fold higher than that of winter (Table 1). Water column depth-integrated chl a (Ichl a) exhibited similar seasonal variations as the surface chl a, with the highest in summer (62 mg m−2), moderate in spring and autumn (28 and 26 mg m−2, respectively), and the lowest in winter (16 mg m−2) (Table 1). Summer Ichl awas 2-fold higher than that of spring/autumn, and ∼4-fold of the winter concentration.
Size fractionated phytoplankton biomass
In spring, the >5 μm (5–20 μm + >20 μm) fraction made up more than 70% of the total chl a biomass in the estuarine and near-shore waters, whereas small-sized (<5 μm) phytoplankton dominated at the surface and middle layer of the offshore stations in the northeast and southeast region. Average proportion of large-sized (>20 μm) phytoplankton was <15% in the bottom layer (Figure 6).
In summer, large-sized (>20 μm) phytoplankton exceeded 50% at surface and the >5 μm size fraction chl a biomass dominated the total chl a in the whole water column, except stations in the southeast of the investigation region, where the waters were mainly influenced by high saline oligotrophic waters of Taiwan Warm Current and a branch of Kuroshio Current (Figure 3).
Compared with other seasons, the proportion of 5–20 μm size fraction phytoplankton was dominated both in the surface and middle layer in autumn, averaging ∼40% and 50%, respectively (Figure 6). In winter, large-sized (>20 μm) chl a biomass accounted for a low proportion in the surface and middle layer (averaged ∼25%), except Stn C1 in the river month (Figure 6).
Average proportions of 3 size fractions were almost the same in the whole water column in summer, except the large-size fraction (>20 μm) at the bottom layer were slightly lower than that of surface and middle layer (Figure 6). In addition to summer, the highest average proportion of >20 μm phytoplankton appeared at the bottom layer all year, especially in spring and autumn (Figure 6).
Compared with Stn E2 (shallow, 21 m), Stn D3 (deep, 63 m) was more influenced by YR outflow. The vertical salinity profile of Stn D3 in spring suggested that the water body was homogeneous at the upper 40 m, and a sudden increase of salinity from ∼33 to >34 between 40–50 m depth (Figure 7). A small proportion of the phytoplankton community was >20 μm in spring. But >20 μm size phytoplankton dominated in summer, when the water column was deeply stratified, >50% at the surface and ∼50% below 10 m depth. More than 90% of chl a biomass was >20 μm size at the bottom in autumn (Figure 7). In comparison, no halocline occurred at the shallow near-shore Stn E2, the freshwater from YRE and Hangzhou Bay influenced the whole water column. The small size fraction (<5 μm) occupied only a small part of total chl a all year, except in autumn (∼40% at surface and middle layer). The large size fraction (>20 μm) contributed ∼4 mg m−3 chl a biomass at surface in summer (Figure 7). Total chl a was lower than 0.5 mg m−3 in winter, with similar size structure at both sites, indicating strong vertical mixing induced by winter monsoon.
Seasonal dominated species is one of the main factors effecting the size fraction of phytoplankton biomass (chl a). According to Luan et al. (2006) and He and Sun (2009), Skeletonema costatum, a small-cell-chained species (in long chains, generally >20 μm long), dominated all year in YRE and adjacent coastal waters, while some seasonal dominated species appeared at different seasons, such as Noctiluca Scientillans (ellipsodial, >150 μm in diameter) and Prorocentrum dentatum (Asymmetric pear-shaped, 15–22 μm long, 9–14 μm wide) in spring, Proboscia alata f. gracillima (slender and straight tubular, ∼1000 μm long, 3–7μm in diameter) in summer, Coscinodiscus jonesianus (discal, >100 μm in diameter) in autumn and Coscinodiscus asteromphalus (discal or short cylindrical, >100 μm in diameter) in winter.
Typically, maximum chl a biomass and productivity occur at intermediate salinities and coincide with the non-conservative decrease in nutrients along the salinity gradient in the estuarine and coastal waters (Lohrenz et al., 1999). Phytoplankton biomass is governed by light penetration, nutrients and physical forces (such as tide, wind, mixing, etc.) in the estuarine and adjacent coastal waters (Cloern, 1996; Yin et al., 2001, 2004; Ho et al., 2010). The distribution pattern of chl a biomass in YRE was comparable with other large estuarine waters in the world, such as the northern Gulf of Mexico, which receives the discharge from nutrient enriched Mississippi and Atchafalaya rivers (Lohrenz et al., 1999), the Chesapeake Bay, the largest estuary in US, where the Susquehanna River freshwater flows (Schubel and Pritchard, 1986), the Adriatic Sea receiving the Po River discharge (Revelante and Gilmartin, 1976), and the Pearl River plume in the northern South China Sea (Yin et al., 2001, 2004).
Physical processes regulation
Spatial and temporal variations in biological processes (e.g. phytoplankton biomass) are largely linked to physical forces such as tidal cycle, and river outflow dilution, as well as winds (e.g. monsoon), induced vertical mixing in estuaries and coastal wasters (Cloern, 1996; Yin et al., 2004).
Since an estuary is an interface between freshwater and seawater, the outflow of freshwater from rivers becomes an important factor in driving the spatial variations of chl a biomass. There is often a progression downstream or upstream associated with seasonal changes in river flow (Yin et al., 2004). There was strong estuarine circulation in the YRE during wet season, which was indicated by the sharp halocline in summer and autumn (Figures 4 and 7).
In the turbid river-mouth waters, the dilution rate was too fast to allow phytoplankton blooms, and phytoplankton growth rates were often limited by light. Yin et al. (2001) indicated that phytoplankton grew slowly relative to the dilution and vertical mixing, and could not accumulate in the euphotic zone in the estuary in spite of high nutrients. As to the open coastal receiving waters, the dilution rate decreased and the light penetration improved. As a result, phytoplankton blooms occurred in the plume region in YRE. The similar results were found by Gong et al. (1996) in the middle and outer shelves of ECS, with high chl a biomass appearing in the plume of the YRDW during summer.
In the southeast stations of study area, the <5 μm size fraction was dominated all year both in surface and middle waters. According to Guan (1994) and Su (1998), a branch of oligotrophic Kuroshio Current from northeast Taiwan strongly controlled the waters in this area (Figure 3). Studies showed that small-sized algae, such as picophytoplankton (<2 μm), are more competitive in the nutrient depleted open oceanic waters (Li and Luan, 1998; Yin et al., 2004). The chl a biomass and size fraction composition were almost consistent with depth at Stn D3 in winter. It was mainly caused by winter northeast monsoon induced strong vertical mixing, which was indicated by the vertical profile of salinity.
Phytoplankton biomass varied seasonally (i.e. high in summer and low in winter) in response to the fluctuation of YR discharge. Highly significant correlations were between surface nutrients (DIN and SiO4) and salinity (p < 0.01). It suggested that the nutrients of the study area were mainly from YR outflow, similar to the results from the Pearl River Estuary (Ho et al., 2010). Wang et al. (2008) concluded that YR inputs and phytoplankton uptake were the leading factors regulating the distributions and concentrations of nutrients in the YRE and adjacent ECS coastal waters.
Nitrogen and phosphorus remarkably increased in YRE from 1960s to early 2000s (Chai et al., 2006). Dissolved inorganic nitrogen in the estuarine waters have increased ∼10-fold in the last 40 years, from <20 μM in 1960s (Chai et al., 2006) to >100 μM in 2002–2003 (this study). PO4 have increased about 3- to 4-fold since 1960s, reached ∼1 μM in the estuarine stations. As a response to nutrients enrichment, chl a biomass significantly increased both in YRDW and offshore waters (Zhou et al., 2008). The maximum chl a biomass in spring/summer in 2002 (24.2 mg m−3, this study) was about 4 times higher than that in the mid 1980s (6.0 mg m−3, Ning et al., 1988).
Relative to PO4, nitrate (including nitrite) was present in excess of the amount required to support algae growth in surface water of the northwestern ECS, including our investigation area (Wong et al., 1998; Wang and Wang, 2006). The presence of excess nitrogen suggest that effective removal of PO4 from the sewage runoff flowing into ECS by YR may be one of the valid strategies to reduce the increase in eutrophication, hypoxia and the high frequency of HABs in YRE and adjacent ECS waters.
Silicon is an essential macronutrient for diatom growth. Although concentrations still remain high, they may not became a limiting factor for phytoplankton growth currently, SiO4 has decreased since the TGD began to fill as its sluice gates started closing in 2003 (Song et al., 2006, 2008; Zhou et al., 2008). The N:Si ratios in this study (1.3 in spring and ∼7 in summer) were comparable with the results of Chai et al. (2006) in 2004 (Table 1). The N:Si ratio was 0.7 in May 2001 (before water storage), and 1.3 in May 2004 (after water storage) in YRE (Chai et al., 2006). One response of phytoplankton communities to such changing nutrients input is the decrease in the relative abundance of diatoms, accompanied by an increasing trend of dinoflagellates. As a result, large size (>20 μm) phytoplankton contributions may decrease. The percentage of diatom species decreased from 85% in the mid 1980s to 64% in 2002 (Zhou et al., 2008). The frequent occurrence of potentially harmful species shifted from diatom (e.g. Skeletonema costatum) to dinoflagellates (e.g. Prorocentrum donghaiense, synonym P. dentatum in this area) (Gao and Song, 2005; Zhou et al., 2008).
Phytoplankton biomass showed seasonal variations in relation to YR discharge, with the highest value in summer, lowest in winter and intermediate in spring and autumn. Generally, phytoplankton biomass was dominated by the small size fraction cells (<5 μm) in offshore waters, while large size phytoplankton (>20 μm) were dominant in the turbid eutrophic YRE, and near-shore region. Summer algal blooms mostly occurred beyond the front zone of YRDW with stable stratification and sufficient nutrients replenishment from YR outflow. In order to decrease eutrophication and occurrences of large scale harmful algal blooms in YRE and adjacent coastal waters, sewage discharges should be strictly monitored and under control, especially the phosphorus discharges.
This study was financially supported by the 973 Project (2001CB409703, 2010CB951201), the Natural Science Foundation of China (40806050 and 41106107), the CAS (SQ200803), the Ministry of Science and Technology of China (2008FY110100), the Open Fund of Jinan University (JN2010-4) and Laboratory of Marine Ecosystem and Biogeochemistry, Second Institute of Oceanography (200806). We thank Ecology and Oceanography of HABs in China Project team for providing temperature, salinity and part of the nutrient data. We are grateful to the two anonymous reviewers for valuable comments.