Spatial and Temporal Variations of Chlorophyll a and Primary Productivity in the Hangzhou Bay

: The Hangzhou Bay (HZB) is an important part of the Zhoushan ﬁshing ground, the most productive region in the Eastern China Seas. Although HZB remains eutrophication all year round, its chlorophyll a (Chl) and primary productivity (PP) are usually signiﬁcantly lower than those in the adjacent waters. In the present study, we presented the Chl and PP distributions in the HZB and analyzed their correlations with environmental factors in four seasons. The ﬁeld observation showed that Chl and PP had signiﬁcant seasonal variations, and was highest in the summer (1.66 ± 0.61 mg · m − 3 and 12.11 ± 12.25 mg C · m − 3 · h − 1 , respectively). Total suspended matters (TSM) concentration was the key environmental factor that constrains PP in the study area. High concentration of TSM reduced light exposure (LE, the annual mean value was 0.92 ± 0.81 Einstein · m − 2 · day − 1 ) in the mixed layer of the HZB, which was much lower than the saturated light intensity of phytoplankton growth, and thus caused a strong light limitation in the HZB. However, the seasonal variations in the photosynthesis rates (P B ) and Chl did not coincide. This fact suggested that the growth rate was not the only factor controlling seasonal variations of phytoplankton biomass. In winter, the very high TSM and strong mixing might reduce the zooplankton grazing rate, and lead to a relatively high concentration of Chl during the very low LE and P B period. These results implied that, in the HZB, the extremely turbid water could affect both phytoplankton growth and loss, which was probably the major mechanism responsible for the complex phytoplankton spatial and temporal variations.


Introduction
Primary productivity (PP) of phytoplankton is critical to marine fishery resources. In the coastal area of the East China Sea (ECS), terrigenous input, tidal mixing and coastal upwelling jointly control the material and energy required for phytoplankton growth and thus have important impacts on the distribution of coastal fisheries [1,2]. Affected by the Changjiang River and Zhejiang coastal upwelling, the largest fishing ground of the ECS appears around the Zhoushan Islands (29 • 30 S-31 • 00 S, 120 • 30 E-125 • 00 E), including the Zhoushan Islands, Hangzhou Bay (HZB) and the Changjiang Estuary (CJE) diluted water region, etc. [3][4][5]. The Zhoushan fishing ground is also the highest annual PP area in the ECS [3,4].
As an important part of the Zhoushan fishing ground, the HZB is the spawning and nursery ground of many economic species [6,7]. The HZB is one of the strongest tidal regions in the world [8] and is a highly eutrophic bay [9]. Influenced by its trumpet-like

Study Area
The HZB, adjacent to the CJE, is located in the east coast of China (29 • 50 -31 • 00 N, 121 • 00 -122 • 00 E). The bay mouth of the HZB is adjacent to the CJE on the north, connected to the Zhoushan islands sea area on the east. The HZB is subjected to strong tides due to its trumpet-shaped characteristic and flat topography with an average water depth of less than 10 m [22]. Moreover, the HZB is significantly influenced by the Changjiang Diluted Water and a secondary plume of the Changjiang River entering the HZB on the north side [23].

Environmental Parameters
A Rosette sampler was used to collect seawater samples at the surface and bottom layers of the HZB. 100 mL of seawater filtered by 0.45 μm acetate membrane was collected and fixed with 0.3 mL of 35 g·dm −3 HgCl2 for storage. Dissolved inorganic nitrogen (DIN, NOX + NH4 + ), dissolved inorganic phosphorus (DIP) and dissolved silicate (DSi) were measured by Skalar San++ according to the National Standard of the People's Republic of China (GB12763.4-91).
TSM was measured using the weighing method. The surface and bottom layer seawater samples were filtered by Whatman GF/F filter membrane (prepared at 450 °C for 2 h and weighed before). The membrane was cryopreserved and brought back to the laboratory. In the laboratory, the thawed membrane was placed in the muffle furnace at 500 °C for 2 h, equilibrated with a silica gel dryer and weighed until the weight no longer changed. Finally, the TSM was calculated using the result of weighing and the volume of the sample.
LE can be calculated from the sea surface light intensity (i.e., photosynthetically available radiation (PAR), obtained from NASA's sea surface light intensity data product (https://oceancolor.gsfc.nasa.gov, accessed on 2 April 2021) [24]), and the formula is as follow [3]: where I0 is the sea surface light intensity (Einstein·m −2 ·day −1 ), MLD is the mixed layer depth (m), K is the light attenuation. MLD is related to water stability and K is controlled by TSM. In this study, MLD is the first depth where ΔTemperature > 0.5 °C·m −1 , K is calculated from Secchi disk depth according to the method proposed by Castillo-Ramirez et al. [25].

Chlorophyll a
Chl concentrations (mg·m −3 ) were determined by fluorimetry after 50 mL surface and bottom layer seawater samples were collected, respectively. The seawater samples were filtered by 200 μm bolting-silk to remove zooplankton, and then filtered by Whatman

Environmental Parameters
A Rosette sampler was used to collect seawater samples at the surface and bottom layers of the HZB. 100 mL of seawater filtered by 0.45 µm acetate membrane was collected and fixed with 0.3 mL of 35 g·dm −3 HgCl 2 for storage. Dissolved inorganic nitrogen (DIN, NO X + NH 4 + ), dissolved inorganic phosphorus (DIP) and dissolved silicate (DSi) were measured by Skalar San++ according to the National Standard of the People's Republic of China (GB12763.4-91).
TSM was measured using the weighing method. The surface and bottom layer seawater samples were filtered by Whatman GF/F filter membrane (prepared at 450 • C for 2 h and weighed before). The membrane was cryopreserved and brought back to the laboratory. In the laboratory, the thawed membrane was placed in the muffle furnace at 500 • C for 2 h, equilibrated with a silica gel dryer and weighed until the weight no longer changed. Finally, the TSM was calculated using the result of weighing and the volume of the sample.
LE can be calculated from the sea surface light intensity (i.e., photosynthetically available radiation (PAR), obtained from NASA's sea surface light intensity data product (https://oceancolor.gsfc.nasa.gov, accessed on 2 April 2021) [24]), and the formula is as follow [3]: where I 0 is the sea surface light intensity (Einstein·m −2 ·day −1 ), MLD is the mixed layer depth (m), K is the light attenuation. MLD is related to water stability and K is controlled by TSM. In this study, MLD is the first depth where ∆Temperature > 0.5 • C·m −1 , K is calculated from Secchi disk depth according to the method proposed by Castillo-Ramirez et al. [25].

Chlorophyll a
Chl concentrations (mg·m −3 ) were determined by fluorimetry after 50 mL surface and bottom layer seawater samples were collected, respectively. The seawater samples were filtered by 200 µm bolting-silk to remove zooplankton, and then filtered by Whatman GF/F filter membrane. The Whatman GF/F filter membrane containing phytoplankton was extracted with 90% acetone for 18 h in darkness and at −20 • C. Then, the Turner Designs 10 AU Fluorometer was used to determine the concentration in the extraction fluid. The determination and calculation were performed by following the method in [26].

Primary Productivity and Photosynthetic Rate
PP and photosynthetic rate (P B ) were determined using the 14 C isotope tracer method developed by Nielsen and modified by Ning et al. and Evans et al. [27][28][29]. The surface seawater samples were filtered with 200 µm bolting-silk to remove zooplankton. Then, they were injected into two white culture bottles and one black culture bottle. A certain amount of NaH 14 CO 3 tracer was added into each culture bottle successively, and placed on the deck to simulate field culture at surface temperature. Different neutral light fading materials were selected to control the light intensity of the simulated incubator. Meanwhile, surface seawater was cyclically pumped to control the temperature. The incubator was cultured for 6 h. After culture, the phytoplankton samples were collected using Whatman GF/F filter membrane. The membrane was fumigated with concentrated hydrochloric acid, and then placed in scintillation bottles for drying and preservation at low temperature. The samples were taken back to the laboratory for determination on a PE 2900 liquid scintillation counter. PP was calculated according to the formula recommended by Parsons [30], and P B was calculated as the ratio of surface PP to Chl.

Data Analysis
Considering the strong mixing in the HZB, in this study, the mean Chl concentrations and chemical variables in the mixed layer were used in the analysis. The distribution of each parameter was drawn using Ocean Data View 5.5.1 [31]. R language was used for data analysis, and Pearson was used to test the correlation between Chl, PP and environmental factors.

Environmental Parameters
In the HZB, the surface temperature and salinity varied greatly in different seasons, and their distributions are shown in Appendix A ( Figure A1). The surface temperature was higher in summer (27.87 ± 1.08 • C) and autumn (19.83 ± 0.62 • C), and lower in spring (14.93 ± 0.63 • C) and winter (7.83 ± 1.05 • C); while the mean surface salinity was higher in winter (19.50 ± 3.09) and spring (19.06 ± 6.31), and lower in summer (13.90 ± 8.19) and autumn (13.58 ± 3.84) ( Table 1). In addition, there was a significant intrusion of high salinity seawater in the bay mouth in summer ( Figure A1). Table 1. Chl, PP, P B and environmental parameters recorded in the HZB in each cruise (mean ± SD).

Parameters
July The nutrient concentrations in the HZB were relatively high and had significant spatial distributions ( Figure 2). DIN, DIP and DSi showed decreasing trends from west to east in each season (   As shown in Figure 3, the TSM in the HZB had significant seasonal variation. TSM in summer (0.56 g·dm −3 ) was much lower than in other seasons (mean TSM in winter, spring and autumn were 2.34 g·dm −3 , 3.01 g·dm −3 , and 2.41 g·dm −3 , respectively), about a quarter to a fifth of them (Table 1). Although PAR in the HZB was not low, influenced by high TSM concentrations, LE was usually only 1-5% of PAR. The seasonal variation of LE was ranked in the order summer > spring > autumn > winter ( Figure A2), not wholly consistent with TSM. The mean PAR in summer (34.31 ± 3.50 Einstein·m −2 ·day −1 ) and spring (33.50 ± 4.19 Einstein·m −2 ·day −1 ) was similar, but the mean LE in summer (1.89 ± 0.98 Einstein·m −2 ·day −1 ) was about twice as high as that in spring (0.87 ± 0.35 Einstein·m −2 ·day −1 ) ( Table 1). In addition, there was no significant difference between mean LE in autumn (0.46 ± 0.13 Einstein·m −2 ·day −1 ) and winter (0.43 ± 0.19 Einstein·m −2 ·day −1 ) ( Table 1). The highest LE (4.29 Einstein·m −2 ·day −1 ) occurred in the middle of the bay mouth in summer ( Figure 3I), consistent with the lowest TSM ( Figure 3A).

Distribution of Chl
As shown in Figure 4, Chl in the HZB had significant seasonal variations. Chl concentrations were higher in summer (1.66 ± 0.61 mg·m −3 ) and winter (1.50 ± 0.51 mg·m −3 ), and lower in spring (1.00 ± 0.31 mg·m −3 ) and autumn (0.90 ± 0.41 mg·m −3 ). In addition, high Chl concentrations (>2 mg·m −3 ) only occurred in summer and winter, but there are no found in spring and autumn ( Figure 4).
In summer, the range of Chl concentrations was 0.84-2.97 mg·m −3 . High Chl (>2 mg·m −3 ) in summer occurred in the east of the HZB (Figure 4), and the lowest Chl concentration (0.84 mg·m −3 ) occurred in the south side of the HZB ( Figure 4A). The distribution of Chl in summer was contrary to that of TSM, and both exhibited significant variation trends from west to east. In winter, the range of Chl concentrations was 0.91-2.34 mg·m −3 . And high Chl (>2 mg·m −3 ) in winter occurred in the middle of the HZB, the northwest side and north of the bay mouth ( Figure 4B). In spring, the range of Chl concentrations was 0.45-1.62 mg·m −3 . The highest and lowest Chl in spring occurred in the south and north of the study area ( Figure 4C), respectively, opposite to TSM. In autumn, the range of Chl concentrations was 0.28-1.69 mg·m −3 . Chl in autumn was higher in the east and west and lower in the north and south ( Figure 4D). In summer, the range of Chl concentrations was 0.84-2.97 mg·m −3 . High Chl (>2 mg·m −3 ) in summer occurred in the east of the HZB (Figure 4), and the lowest Chl concentration (0.84 mg·m −3 ) occurred in the south side of the HZB ( Figure 4A). The distribution of Chl in summer was contrary to that of TSM, and both exhibited significant variation trends from west to east. In winter, the range of Chl concentrations was 0.91-2.34 mg·m −3 . And high Chl (>2 mg·m −3 ) in winter occurred in the middle of the HZB, the northwest side and north of the bay mouth ( Figure 4B). In spring, the range of Chl concentrations was 0.45-1.62 mg·m −3 . The highest and lowest Chl in spring occurred in the south and north of the study area ( Figure 4C), respectively, opposite to TSM. In autumn, the range of Chl concentrations was 0.28-1.69 mg·m −3 . Chl in autumn was higher in the east and west and lower in the north and south ( Figure 4D).

Surface PP and P B
PP in the HZB ranged from 0.11-30.38 mg C·m −3 ·h −1 and had a significant variation of summer > spring > winter > autumn ( Figure A2). The highest PP in summer (30.38 mg C·m −3 ·h −1 ) occurred in the north side of the bay mouth ( Figure 5A). PP in summer experienced a decreasing trend from east to west ( Figure 5A). The mean PP in winter was 1.32 ± 1.02 mg C·m −3 ·h −1 , and its distribution trend was similar to that of Chl, where both highvalue areas occurred in the middle of the bay mouth ( Figure 5B). Although Chl in spring was lower than in winter, the mean PP in spring (2.10 ± 1.18 mg C·m −3 ·h −1 ) was higher than in winter. High PP in spring occurred in the north of the HZB with a downward trend from north to south ( Figure 5C). The mean PP in autumn was only 0.20 ± 0.12 mg C·m −3 ·h −1 , almost one order of magnitude lower than those in other seasons, indicating that autumn was a typical period of low productivity.

Surface PP and P B
PP in the HZB ranged from 0.11-30.38 mg C·m −3 ·h −1 and had a significant variation of summer > spring > winter > autumn ( Figure A2). The highest PP in summer (30.38 mg C·m −3 ·h −1 ) occurred in the north side of the bay mouth ( Figure 5A). PP in summer experienced a decreasing trend from east to west ( Figure 5A). The mean PP in winter was 1.32 ± 1.02 mg C·m −3 ·h −1 , and its distribution trend was similar to that of Chl, where both high-value areas occurred in the middle of the bay mouth ( Figure 5B). Although Chl in spring was lower than in winter, the mean PP in spring (2.10 ± 1.18 mg C·m −3 ·h −1 ) was higher than in winter. High PP in spring occurred in the north of the HZB with a downward trend from north to south ( Figure 5C). The mean PP in autumn was only 0.20 ± 0.12 mg C·m −3 ·h −1 , almost one order of magnitude lower than those in other seasons, indicating that autumn was a typical period of low productivity.

Correlations among Biological and Environment Parameters
We compared Pearson correlation among biological and environmental parameters in the different seasons. As shown in Figure 7, nutrients usually had significant negative correlations with salinity ( Figure 7) in the four seasons, inferring nutrients' variations dominated by eutrophic terrestrial input in the HZB. However, biological parameters, such as Chl, PP and PB, had no relationship with nutrients. It suggested that nutrient variations neither affected phytoplankton biomass nor its growth. Although high TSM is the major reason for low LE in the HZB, the LE also had significant correlations between PAR or MLD in different seasons. This meant that LE variability in each season was complex, and LE distributions also could be partly affected by PAR and MLD. Note that correlations among biological and environmental parameters were not consistent in different seasons. It meant that, in a year, phytoplankton variations could not be explained by using a single environment parameter. Our results showed that PP had the similar spatial distribution as PP in each season ( Figure 6). P B also had significant seasonal variations ( Figure A2). In this study, P B in the HZB ranged from 0.19-10.32 mg C·(mg Chl·h) −1 , with the annual mean value of 2.36 ± 2.82 mg C·(mg Chl·h) −1 . P B was higher in summer (5.46 ± 3.64 mg C·(mg Chl·h) −1 ) and spring (2.33 ± 1.45 mg C·(mg Chl·h) −1 ), while lower in winter (0.88 ± 0.30 mg C·(mg Chl·h) −1 ) and autumn (0.20 ± 0.01 mg C·(mg Chl·h) −1 ) ( Table 1).

Correlations between phytoplankton and environment factors
As shown in Figure 7, the correlations between phytoplankton and environment factors were not consistent in different seasons. In summer, PP and P B had high correlations with TSM and LE (Figure 7). However, in other seasons, the correlations between them phytoplankton biomass nor its growth. Although high TSM is the major reason for low LE in the HZB, the LE also had significant correlations between PAR or MLD in different seasons. This meant that LE variability in each season was complex, and LE distributions also could be partly affected by PAR and MLD. Note that correlations among biological and environmental parameters were not consistent in different seasons. It meant that, in a year, phytoplankton variations could not be explained by using a single environment parameter.

Dynamics of Chl
Historical studies showed that the variation of Chl in the HZB was usually ranged within one order of magnitude, and the high Chl usually occurs in summer [14,15,17]. Long-term observations of Jia et al. showed that the Chl concentrations varied from 0.54 to 3.42 mg·m −3 during 1992-2012 [14], and Chl concentrations were higher in summer but lower in spring [17]. Our results showed that the Chl level and seasonal variation were consistent with those in previous studies. Our study found that Chl had significant spatial and temporal variations in the HZB, ranging from 0.45 to 2.97 mg·m −3 . The annual highest Chl (2.97 mg·m −3 ) occurred in the northern side of the bay mouth in summer ( Figure 4A), and the annual lowest Chl (0.28 mg·m −3 ) occurred in the northern side of the bay mouth in autumn ( Figure 4D). The Chl in HZB increased with the water depth in summer, and the high Chl occurred in the area near the CJE, consistent with the historical data [17]. However, in other seasons, the spatial distribution of Chl in HZB seemed to have no clear rules.
Our results showed that the annual mean concentrations of DIN, DIP and DSi in the HZB were 89.14 µmol·dm −3 , 1.42 µmol·dm −3 and 65.85 µmol·dm −3 (Table 1), respectively, much higher than the limitation thresholds (DIN = 1 µmol·dm −3 , DIP = 0.1 µmol·dm −3 and DSi = 2 µmol·dm −3 , suggested by Justić et al [32]). The mean values of N/P and Si/P were higher four and three times than Redfield ratio [33], respectively. Notably, although N/P and Si/P indicated that DIP would deplete first, the DIP concentration is still higher than the limitation threshold [34]. This suggested that the HZB was a potential phosphorus limited region [35]. However, there was no significant correlation between Chl and nutrients in the HZB (Figure 7). This suggested that nutrients were not the control factor for phytoplankton, which was consistent with the previous study [14].
Some literature suggested that water temperature was an important factor affecting the ECS nearshore phytoplankton [36,37]. However, our results showed (Figure 7) that Chl and temperature changes in HZB lacked consistency, and the correlation coefficient between the two was not high (r 2 < 0.3). This suggests that temperature is not an important factor affecting phytoplankton changes and that its effect on phytoplankton may be indirect [3].
We compared the Chl levels in different highly turbid estuaries ( Table 2) and found that the Chl in the HZB varied within a smaller range than other estuaries. In addition, the Chl level in the HZB was lower than in other estuaries. For example, the CJE is adjacent to the HZB with similar nutrient levels, but the Chl in the CJE is much higher than the HZB. The reason could be that the very low LE inhibited the phytoplankton blooming, and there was no bloom season because of the all-year-round strong light limitation in the HZB. The annual mean LE in the HZB was merely 0.92 ± 0.81 Einstein·m −2 ·day −1 , which was about only 10% of the mean LE in the coastal waters of China [3]. Hao et al. suggested that, due to the seasonal increase of TSM and MLD, light limitation usually occurred in the winter half-year in the offshore waters of the ECS [3]. However, in the HZB, the light limitation existed in any season because of the extremely high TSM induced by tide mixing. In other sea areas, the light limitation tends to be intermittent, resulting in localized or seasonal blooms, and thus the Chl in these areas usually change greatly.

Dynamics of PP and P B
In temperate oceans, the saturated light intensities of phytoplankton usually vary from 33 to 400 µmol Einstein·m −2 ·s −1 [43]. As the dominant species of the HZB [18,19,[44][45][46][47], diatoms' saturated light intensities are usually above 50 µmol Einstein·m −2 ·s −1 [48]. This suggested that diatom light limitation would occur once LE lower than 2.2 Einstein·m −2 ·day −1 . Our result showed the annual mean LE in the HZB was only 0.92 Einstein·m −2 ·day −1 , much lower than saturated light level, indicating strong light limitation in the HZB. LE was higher than the saturated light intensity of phytoplankton only in the bay mouth of the HZB in summer. This was due to the significant increase of LE in the HZB caused by the northeastward movement of the Changjiang River secondary plume and the intrusion of low turbidity seawater. In this study, PP and P B in the HZB showed significant negative correlations with TSM only in summer (r = −0.82, r = −0.93, respectively) ( Figure 7). This is mainly because the P B of phytoplankton will increase with light when the phytoplankton was in the period of light limitation [48,49]. When the TSM decreases, PP and P B will increase rapidly. Interestingly, there was no significant correlation between TSM and P B in other seasons. This might be because TSM in these seasons is about five times higher than in summer, leading to LE being much lower than the saturated light intensity of phytoplankton. Thus, the variations of P B and LE in the HZB in the same season lacked gradient, leading to weak correlations between them. In fact, LE and P B had relatively consistent seasonal variations (i.e., summer > spring > winter ( Figure A2)), indicating that TSM controlling LE was the main factor affecting phytoplankton photosynthesis.
Strong light limitation also led to excess nutrient in the HZB. Since light limitation cause that PP in the HZB was one order of magnitude lower than that of the CJE [4]. The phytoplankton in the HZB cannot uptake nutrients efficiently due to the light limitation. This causes terrigenous inputs of nutrients to accumulate in the HZB, leading to eutrophication. If the TSM concentration of the HZB decreases and the light limitation disappears, the HZB would turn into a potential high-PP region due to its abundant nutrient stocks.

Possible Effects of Zooplankton on Phytoplankton
In fact, the standing stock of phytoplankton is not only affected by light and nutrients but also related to the losses caused by respiration, zooplankton grazing, sinking and viral lysis [50,51]. Considering the re-suspension process caused by strong tidal mixing in the HZB, the loss of phytoplankton sinking might be negligible, so the losses of phytoplankton in the HZB are mainly controlled by zooplankton grazing.
In winter, although LE and P B were relatively low, Chl were higher than those in spring and autumn in the HZB ( Figure A2). This was probably because extremely high TSM in winter might affect the zooplankton grazing rate. Arruda et al. suggested that suspended particles can get into the zooplankton gut, potentially reducing food assimilation efficiency [52]. Herzig et al. suggested that sediment may also cause the abrasion of zooplankton exoskeletons during mixing [53]. On the other hand, the low temperature would prolong zooplankton development, reduce hatching abundance and decrease adult zooplankton survival [54]. This will lead to lower zooplankton biomass in winter, which reduces the grazing on phytoplankton. These factors might cause the low zooplankton grazing rate in highly turbid waters in winter, which made the accumulation of phytoplankton biomass in the water column. As a result, phytoplankton still had a relatively high Chl level in winter when P B was low.

Conclusions
This study confirmed that LE levels of HZB were much lower than the requirement for phytoplankton growth, and light is the main limiting factor controlling Chl and PP levels in the area. Due to the low LE restricted by the extremely high TSM concentration, phytoplankton in most sea areas of the HZB is under the strong light limitation throughout the year. Nutrient concentrations did not affect the temporal and spatial distribution of phytoplankton. The nutrient-uptake rate of phytoplankton will decrease under the light limitation, resulting in the excess of nutrients, which is likely the important reason for the perennial eutrophication in the HZB. The inconsistent seasonal variation of Chl and P B meant that the phytoplankton dynamics in the HZB were not only controlled by the bottom-up effect but also affected by the phytoplankton loss relevant to zooplankton. However, due to lacking of synchronous zooplankton grazing observations, it is difficult to assess seasonal variations in phytoplankton loss rates and their environmental control mechanisms. Further studies are required to understand the effects of phytoplankton dynamics in the HZB on the food web and eutrophication in the Zhoushan fishing ground. This requires not only longer-term interdisciplinary observations but also the studies of primary and secondary production coupling, especially the influence of highly turbid waters and strong tidal mixing on the phytoplankton loss rate.