Tracing nitrate sources in one of the world’s largest eutrophicated bays (Hangzhou Bay): insights from nitrogen and oxygen isotopes

1. Introduction

Nutrients support marine primary productivity, but excess nutrient loads can lead to eutrophication and harmful algal blooms and hypoxia (Anderson et al., 2002; Rabalais et al., 2002; Diaz and Rosenberg, 2008; Wang et al., 2016; Glibert, 2017). In recent decades, due to human activity, riverine inputs of nutrients (mainly nitrate, ${{\rm {NO}}_3^-} $) to estuaries and coastal oceans have increased significantly (Galloway et al., 2004; Gruber and Galloway, 2008; Dai et al., 2011), and eutrophication in estuarine and coastal aquatic ecosystems has emerged as a global environmental problem. Thus, effective management practices are needed to reduce nutrient inputs to these valuable ecosystems. That goal requires that the sources of relevant nutrients are identified and quantified and that the biogeochemical processes controlling nutrient distributions are well understood.

Hangzhou Bay (Fig. 1) is the largest estuary of the East China Sea, covering an area of approximately 4800 km² (length ~270 km and bay-mouth width ~100 km). This funnel-shaped bay is characterized by a large tidal range, strong tidal currents, high turbidity, and high nutrient concentrations (Che et al., 2003). The mean tidal height is 2.5 m at the bay mouth, increasing up-estuary to 4 m to 6 m at the bay head (He et al., 2016; Shi, 2011). The Qiantang River, which carries large amounts of terrestrial material, discharges directly into the upper bay. In autumn, the famous Qiantang tidal bore has a mean height of ~5 m and a maximum height of ~10 m.

Figure  1.  Map of study area: Hangzhou Bay and adjoining waters. The basemap colors indicate water depth, and the labeled dots indicate August 2019 sampling sites.

DownLoad: Full-Size Img PowerPoint

In recent decades, due to increasing nitrogen (N) loads associated with agricultural, industrial, and domestic activities, the waters of Hangzhou Bay have become seriously eutrophic, with nitrate concentrations of up to ~150 μmol/L (Jia et al., 2014; Wu et al., 2019). One significant contributor to dissolved inorganic nitrogen (DIN) in Hangzhou Bay is ${{\rm {NO}}_3^-} $ from Qiantang River discharge (Dong et al., 1986; Gao et al., 1993; Wu et al., 2019). Several small rivers also connect to the bay, but their contribution to bay DIN is negligible because their runoff is much smaller than that of the Qiantang River (Gao et al., 1993). The Changjiang River (Fig. 1), where nitrate concentrations can be >100 µmol/L (Tang, 1990; Dai et al., 2011), contributes large amounts of ${{\rm {NO}}_3^-} $ to the area just south of the Changjiang River Estuary, and this ${{\rm {NO}}_3^-} $ has also been suggested to contribute significantly to Hangzhou Bay DIN (Wu et al., 2019).

Despite this general knowledge of the likely sources of Hangzhou Bay DIN, the relative contributions of ${{\rm {NO}}_3^-} $ from the different sources remains unknown. Variation of δ¹⁵N in nitrate (δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $) has been proven useful in identifying nitrogen sources and transformations in aquatic ecosystems (Kendall, 1998; Middelburg and Nieuwenhuize, 2001; Mayer et al., 2002; Ye et al., 2016; Yang et al., 2018; Chen et al., 2022), and many such studies have been conducted in the Changjiang River Estuary and adjacent area (Liu et al., 2009; Chen et al., 2013; Yan et al., 2017; Jiang et al., 2021), but to our knowledge such work has not been reported for Hangzhou Bay.

In this study, nitrogen and oxygen isotope ratios of ${{\rm {NO}}_3^-} $ were measured in water samples collected in August 2019 from Hangzhou Bay, the Qiantang River, the Changjiang River Estuary, and adjacent coastal waters. Concentrations of DIN [i.e., nitrate (${{\rm {NO}}_3^-} $), nitrite (${{\rm {NO}}_2^-} $), and ammonium (${{\rm {NH}}_4^+} $)] were also measured. These data were analyzed to investigate DIN transport from the Changjiang River to Hangzhou Bay and to quantify the relative contributions of DIN from different sources into Hangzhou Bay.

2. Material and methods

2.1 Sampling sites

In late August 2019, hydrographic data and seawater samples were collected at 39 stations during a research cruise (R/V ZHEHAIKE 1 HAO) in Hangzhou Bay, the Qiantang River, the Changjiang River Estuary, and adjacent coastal areas (Fig. 1). A rosette sampling system fitted with 6 Niskin bottles (10 L) and a CTD (conductivity–temperature–depth) sensor (SBE 917 plus, Sea-Bird Scientific) was used to measure in situ salinity and temperature and to bring seawater samples onboard. Generally, 1 to 3 depths were sampled at each station, depending on water depth. The analysis that follows focuses primarily on surface samples (collected approximately 2 m below the air/sea interface) and bottom samples (collected approximately 2 m above the sea bottom).

2.2 Sample collection

The seawater samples drawn from the Niskin bottles were pre-filtered through pre-combusted (at 450℃ for 4 h) GF/F membranes (Whatman, 0.7 µm) into 125 cm³ HDPE bottles that had been previously rinsed with 10% HCl, distilled water, and the filtrate solution (three times). The samples were then stored at –20℃ for later analysis of nutrient concentrations and the nitrogen and oxygen isotope ratios of nitrate.

2.3 Measurement of nutrient concentrations

Concentrations of nitrate [${{\rm {NO}}_3^-} $] (where the square brackets denote concentration) and phosphate [${{\rm {PO}}_4^{3-}} $] were measured with an automated discrete chemical analyzer (Smartchem 600, AMS Alliance, Italy). Analytical precision for the ${{\rm {NO}}_3^-} $ analyses was better than 2% at concentrations >5 μmol/L. Analytical precision for the ${{\rm {PO}}_4^{3-}} $ analyses was better than 2% for concentrations >0.8 μmol/L. [${{\rm {NO}}_2^-} $] and [${{\rm {NH}}_4^+} $] were determined manually, using standard colorimetric methods (Hansen and Koroleff, 2009). Analytical precision for the ${{\rm {NO}}_2^-} $ analyses was better than 3% at concentrations >0.5 μmol/L. For the ${{\rm {NH}}_4^+} $ analyses, analytical precision was approximately ±4% at 0.5 μmol/L and better than 3% at concentrations >1 μmol/L.

2.4 Analysis of nitrogen and oxygen isotope ratios of ${{{NO}}_{3}^-} $

The nitrogen and oxygen isotope ratios of ${{\rm {NO}}_3^-} $ were determined according to the method of Sigman et al. (2001) and Casciotti et al. (2002). First, ${{\rm {NO}}_3^-} $ was converted to gaseous nitrous oxide (N₂O) by cultured denitrifying bacteria (Pseudomonas aureofaciens, ATCC#13985). The product N₂O was then extracted from the sample vials, purified, and analyzed with an isotope ratio mass spectrometer (DELTA V Advantage, ThermoFisher Scientific). The international nitrate reference materials IAEA-NO-3 (δ¹⁵N = 4.7‰; δ¹⁸O = 25.6‰), USGS34 (δ¹⁵N = −1.8‰; δ¹⁸O = −27.9‰), and USGS35 (δ¹⁵N = 2.7‰; δ¹⁸O = 57.5‰) (Gonfiantini et al., 1995; Böhlke et al., 2003) were incorporated into every sample batch. This method is applicable to samples with [${{\rm {NO}}_3^-} $] ≥ 1 μmol/L, yielding an average standard deviation of 0.2‰ for δ¹⁵N and 0.5‰ for δ¹⁸O. For samples with [NO₂⁻] ≥ 0.5 μmol/L_, the nitrite was removed with sulfamic acid according to the procedure of Granger and Sigman (2009). Isotope ratios are reported in delta (δ) notation in units of per mil (‰):

where (¹⁵N/¹⁴N)_reference denotes N₂ in air and (¹⁸O/¹⁶O)_reference denotes Vienna Standard Mean Ocean Water (VSMOW).

3. Results

In late August 2019, salinity across our study area (Fig. 1) ranged from 0.1 to 29.7 (Fig. 2a), with lowest values in the Qiantang River (station QTJ-1) and Changjiang River (stations G1–G7) and highest values in nearshore waters. Concentrations of dissolved oxygen (DO) ranged from 131 μmol/L to 232 μmol/L (Fig. 2b), with highest concentrations in upper Hangzhou Bay and lowest concentrations in nearshore coastal waters (the Qiantang River was not sampled for DO). Phosphate concentrations ranged from 0.8 μmol/L to 3.6 μmol/L (Fig. 2c), with most values being between 1 μmol/L and 2 μmol/L. Highest concentrations were found at the head of Hangzhou Bay (station QTJ-3), and lowest concentrations were found in nearshore waters.

Figure  2.  Distributions of surface-water (left panels) and bottom-water (right panels) properties in the Qiantang River Estuary, Changjiang River Estuary, Hangzhou Bay, and adjacent nearshore waters in late August 2019: salinity (a), DO (b), and [${{\rm {PO}}_4^{3-}} $] (c).

DownLoad: Full-Size Img PowerPoint

Nitrate concentrations were highest (>120 μmol/L) in the Qiantang and Changjiang rivers (Fig. 3a) with lower concentrations (down to 22 μmol/L) in coastal waters. This pattern is opposite to that of salinity (Fig. 2a). At most stations, nitrite was <1 μmol/L (Fig. 3b) and ammonium was approximately 1 μmol/L (Fig. 3c). Consequently, DIN concentrations (Fig. 3d) were highest in the Qiantang River (158 μmol/L at station QTJ-1). Lowest DIN concentrations were found in coastal waters (down to 24 μmol/L in the near-bottom sample from station D7; Fig. 3d).

Figure  3.  Distributions of surface-water (left) and bottom-water (right) nutrients in the study area in late August 2019: nitrate (a), nitrite (b), ammonium (c), and DIN (d) concentrations.

DownLoad: Full-Size Img PowerPoint

Values of δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ (Fig. 4a) ranged from 5.3‰ to 7.3‰ (averaging 6.8‰ ± 0.3‰), with generally small variations across the study area. The lowest value occurred in the Qiantang River. Values of δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ (Fig. 4b) ranged from −0.4‰ to 3.5‰ (averaging 1.4‰ ± 0.9‰), with values increasing seaward from river to coastal waters. Low values of δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ occurred in the Qiantang and Changjiang rivers, and high values occurred in coastal waters.

Figure  4.  Distributions of surface-water (left) and bottom-water (right) δ¹⁵N and δ¹⁸O of nitrate in late August 2019: δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ (a) and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ (b).

DownLoad: Full-Size Img PowerPoint

4. Discussion

4.1 Conservative mixing of nitrate and DIN from different sources to Hangzhou Bay

Over most of our study area, ${{\rm {NO}}_3^-} $ accounted for more than 95% of DIN (Fig. 5a), while ${{\rm {NO}}_2^-} $ and ${{\rm {NH}}_4^+} $ accounted for less than 1% and 4%, respectively (Figs 5b, c). High values of [DIN]/[${{\rm {PO}}_4^{3-}} $] (>30, nearly twice the Redfield ratio of 16; Fig. 5d) suggest that excessive input of DIN (mainly in the form of ${{\rm {NO}}_3^-} $) to Hangzhou Bay has become a serious environmental problem.

Figure  5.  Distributions of nitrogen and phosphate concentration ratios in late August 2019: nitrate/DIN (a), nitrite/DIN (b), ammonium/DIN (c), and DIN/phosphate (d) concentration ratios. Ratios at each station were computed using vertically averaged nutrient concentrations.

DownLoad: Full-Size Img PowerPoint

The Qiantang River seems to have been one important source of ${{\rm {NO}}_3^-} $, as indicated by the contrasting spatial patterns of [${{\rm {NO}}_3^-} $] and [DIN] (with values being highest in the river, decreasing seaward; Figs 3a and d) and salinity (lowest in the river, increasing seaward; Fig. 2a). Changjiang River water seems to intrude along the bay’s north shore in both surface and bottom waters, evident as a plume of low-δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ water (Fig. 4b). To the south, high-δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ and high-salinity tongues of nearshore coastal water can be seen entering the mouth of the bay through the Zhoushan islands. The distributions of nitrate δ¹⁸O and salinity (Fig. 4b) suggest that the Qiantang River, the Changjiang River, and nearshore coastal waters may all be contributing nitrate to Hangzhou Bay.

Previous studies have suggested that biological removal of nutrients in Hangzhou Bay is relatively insignificant (Dong et al., 1986; Gao et al., 1993) due to high turbidity and consequent light limitation of phytoplankton production (Li et al., 1993; Che et al., 2003). To assess whether the spatial distributions of nitrate and DIN might therefore be controlled mainly by water mixing, we examined relationships between nitrate and salinity and between DIN and salinity. We first divided the study area into three subregions (Fig. 6a) according to which inflowing water body tended to dominate inputs, as indicated by station location (Fig. 1) and values of salinity (Fig. 2a) and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ (Fig. 4b): (1) subregion CJA, deemed to be most strongly affected by Changjiang River inputs; (2) subregion QTJA, deemed to be most strongly affected by Qiantang River inputs; and (3) subregion CWA, where coastal-water inputs dominate. Stations B1 through B4 were included in the CJA subregion because the δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ distribution (Fig. 4b) indicates a Changjiang River source for nitrate at these locations. Within subregion CWA, the average salinity at every station (surface water and bottom water) was >25.

Figure  6.  Study area subregions and mixing properties in each. a. Map of study area, showing the three subregions defined according to predominating water mass inputs: QTJA, influenced primarily by Qiantang River inputs (blue dots); CJA, influenced primarily by Changjiang River inputs (pink dots); and CWA, influenced primarily by coastal water inputs (dark cyan dots). b. Nitrate concentrations as a function of salinity. c. DIN concentrations as a function of salinity.

DownLoad: Full-Size Img PowerPoint

Significant linear relationships were found between [NO₃⁻] and salinity in the river-influenced QTJA and CJA areas (Fig. 6b) (R² = 0.93 for QTJA and 0.99 for CJA), consistent with observed overall patterns of δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ and salinity (i.e., with lowest values in the rivers, increasing seaward to highest values in nearshore coastal waters; Figs 4b and 2a). Significant linear relationships were similarly found for DIN and salinity (Fig. 6c) (R² = 0.97 for QTJA and 0.99 for CJA). These results suggest that as river waters and nearshore waters meet in the vicinity of Hangzhou Bay, their DIN (and ${{\mathrm{NO}}_3^-} $) are conservatively mixed, with little influence from nonconservative biological processes. These findings are consistent with earlier suggestions (mentioned above in this section) that the effects of phytoplankton assimilation on nitrate concentrations in Hangzhou Bay are insignificant due to high concentrations of suspended sediments.

The process of denitrification (i.e., the conversion of nitrate and nitrite to gaseous forms of nitrogen, thus removing DIN from seawater) in Hangzhou Bay is likely similarly insignificant. Denitrifying microbes use nitrite as an electron receptor only when oxygen is limited, and in the course of our cruise DO was always >130 μmol/L (Fig. 2b). The process of nitrification (i.e., conversion of ammonium to nitrate, which does not change DIN concentrations) does seem to have been occurring in the head of Hangzhou Bay, as evidenced by the unchanging concentration of nitrate from Qiantang River to the head of Hangzhou Bay concurrent with decreasing concentrations of ammonium and nitrite (Fig. 3). Still, nitrification seems not to have significantly affected the signal of conservative mixing of DIN from the various sources to Hangzhou Bay.

4.2 Relative DIN contributions to Hangzhou Bay

Conservative mixing models have been widely used to study nutrient dynamics in estuarine and coastal regions (Dähnke et al., 2008; Han et al., 2012; Wang et al., 2014). To estimate relative DIN contributions from different sources to Hangzhou Bay, we likewise used a three-endmember salinity-based conservative mixing model:

where the ƒ terms represent the fractional contributions of the three endmember water masses denoted by the subscripts CJ (Changjiang River), QTJ (Qiantang River), and CW (coastal water). The terms S and DIN represent the parameters being mixed: salinity and dissolved inorganic nitrogen. The terms S_MIX and DIN_MIX represent the quantities in a mixture of the three endmembers.

The first step in the analysis was to assign S and DIN values to each contributing water mass (Table 1). For the Changjiang River, there was little variation in salinity and DIN concentration between stations G1 and G7 (Figs 1, 2a, and 3d), so we used the average of all values along that stretch. For the Qiantang River (water mass QTJ), we used the average values of salinity and DIN concentration measured at stations QTJ-1 and QTJ-3 (Figs 1, 2a, and 3d). For the nearshore coastal water (water mass CW), highest salinity was observed in the bottom water of station D7, along with the lowest concentration of DIN and the highest δ¹⁸O of nitrate (Figs 1, 2a, 3d, and 4b). Thus, these values of salinity and DIN concentration were chosen as the CW endmember parameters.

Table  1.  Endmember parameters used in the mixing model of Eqs 1–3. For an explanation of each endmember salinity and DIN, see Section 4.2

Endmember	Salinity	DIN/(μmol·L−1)
CJ (Changjiang River)	0.16 ± 0.01	120.2 ± 4.1
QTJ (Qiantang River)	0.86 ± 1.06	153.9 ± 6.7
CW (coastal water)	29.74	24.1

 | Show Table

DownLoad: CSV

By inserting endmember characteristics (Table 1) into the mixing model (Eqs 1, 2, and 3) and solving for ƒ_CJ, ƒ_QTJ, and ƒ_CW at each station, we can estimate the fractional contributions of riverine and coastal water DIN across Hangzhou Bay in August 2019. The results indicate that the Qiantang River (Fig. 7a) was a significant source of DIN between the bay head and the central bay, possibly contributing more than 50% of DIN in the bay head area. The Changjiang River (Fig. 7b) notably contributed DIN to the northern bay mouth area, with its influence (i.e., percent contribution) decreasing southward. At the time of our cruise, discharge from the Changjiang River was relatively low and incursion of the nearshore water mass was relatively strong, as indicated by the distributions of salinity (Fig. 2a) and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ (Fig. 4b). Water exiting the Changjiang River was basically squeezed in by intruding coastal water to hug the northern shore of Hangzhou Bay. Coastal waters (Fig. 7c) contributed most of the DIN in the lower bay, from the bay mouth to the central bay.

Figure  7.  Percent contributions of DIN to Hangzhou Bay from the Qiantang River (a), the Changjiang River (b), and nearshore coastal waters (c).

DownLoad: Full-Size Img PowerPoint

4.3 Primary sources of nitrate to Hangzhou Bay

As discussed in Section 4.1, the Qiantang River, the Changjiang River, and nearshore coastal waters all contributed nitrate to Hangzhou Bay, and nitrate generally behaved conservatively in our study area. Thus, the classical dual isotope approach (Kendall et al., 2007; Xue et al., 2009; Yang et al., 2018) could be used to identify nitrate sources.

In the Changjiang River, concentrations of ${{\rm {NO}}_3^-} $ ranged from 114.0 μmol/L to 124.7 μmol/L [averaging (119.0 ± 4.1) μmol/L, stations G1 to G7] in August 2019, close to previously published values (Dai et al., 2011; Jiang et al., 2021). Multiple sources of ${{\rm {NO}}_3^-} $ have been suggested for this river, such as chemical fertilizer, urban sewage effluent, degradation of terrestrial organic nitrogen, and atmospheric deposition (Li et al., 2010; Dai et al., 2011). Among these possibilities, soil N and synthetic N fertilizer (mostly in the form of ammonium salt or urea) have been suggested to be the most important contributors to Changjiang River nitrate, based on isotopic signatures (Li et al., 2010; Jiang et al., 2021). Soil nitrogen is thought to be a primary source of nitrogen in various aqueous environments (Deutsch et al., 2006; Kendall et al., 2007; Chen et al., 2022). Most of the N in soils is bound in organic forms, with dissolved inorganic nitrogen (mainly ${{\rm {NO}}_3^-} $) constituting ~1% of the total N and δ¹⁵N signatures, possibly reflecting the effects of various anthropogenic activities (Kendall and Aravena, 2000). Soil ${{\rm {NO}}_3^-} $ produced from fertilizer generally has a lower δ¹⁵N value (4.7‰ ± 5.4‰) than does animal waste (δ¹⁵N = 14.0‰ ± 8.8‰) (Kendall et al., 2007). N fertilizer applications have greatly increased in China since the 1980s and are well correlated with the increase of [${{\rm {NO}}_3^-} $] in the Changjiang River (Jiang et al., 2021). In our Changjiang River samples, δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ ranged from 6.1‰ to 6.7‰ and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ ranged from −0.4‰ to 0.4‰, likely reflecting a mixed source of nitrate, mainly from soil N and synthetic N fertilizer to the river (Fig. 8).

Figure  8.  Values of δ¹⁵N and δ¹⁸O in various nitrate source reservoirs and in seawater samples in the study area. The boxes indicate ranges of typical isotopic compositions for five potential sources of nitrate: NP : ${{\rm {NO}}_3^-} $ in precipitation; NF : ${{\rm {NO}}_3^-} $ in fertilizer; NFP : ${\mathrm{NH}}_4^+$ (or urea) in fertilizer and precipitation; SN : ${{\rm {NO}}_3^-} $ derived from soil N; and M&S : ${{\rm {NO}}_3^-} $ derived from manure and sewage (Kendall et al., 2007; Xue et al., 2009; Yang et al., 2018). The colored dots show values of δ¹⁵N and δ¹⁸O in seawater samples in each of the three subregions defined in Fig. 6.

DownLoad: Full-Size Img PowerPoint

In the Qiantang River and at its mouth (Stations QTJ-1 and QTJ-3), concentrations of ${{\rm {NO}}_3^-} $ were slightly higher than in the Changjiang River, ranging from 125.0 μmol/L to 127.4 μmol/L [averaging (126.2 ± 1.7) μmol/L]. Values of δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ (5.3‰ to 7.3‰) and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ (0.4‰ to 1.3‰) suggest that soil N was an important source of nitrate in the Qiantang River (Fig. 8). Synthetic N fertilizer could also be an important contributor, as [${{\rm {NO}}_3^-} $] in the Qiantang River Estuary has greatly increased since the 1980s (from <70 μmol/L in the 1980s to >120 μmol/L in 2019; Dong et al., 1986, and Fig. 3a), concurrent with greatly increasing fertilizer applications in China (Jiang et al., 2021). Significantly high concentrations of ${{\rm {NH}}_4^+} $ were also observed in the Qiantang River (18.0 μmol/L to 20.0 μmol/L) in combination with high concentrations of phosphate and nitrite (Figs 2c and 3a), which suggests a local input of ${{\rm {NO}}_3^-} $ derived from manure and sewage. A local manure and sewage input is also suggested by the slight increase in δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ between upriver station QTJ-1 and river-mouth station QTJ-3, as nitrate sourced from manure and sewage is generally characterized by relatively high δ¹⁵N (Fig. 8).

Intrusions of coastal water into Hangzhou Bay and the Changjiang River Estuary are indicated by the high-salinity tongues in those areas during our study period (Fig. 2). This coastal water was possibly formed in the summer (the flood season), when a low-salinity, high-nitrate river plume could have dispersed offshore to mix with high-salinity, oligotrophic surface waters of the East China Sea shelf and also the nearshore Kuroshio Branch Currents (Chang and Isobe, 2003; Zhou et al., 2009). According to previous studies, this shelf water is almost depleted of ${{\rm {NO}}_3^-} $ and the branch-current waters generally contain [${{\rm {NO}}_3^-} $] of ~13.0 μmol/L (Wang et al., 2016; Yan et al., 2017). Thus, the Changjiang River ([${{\rm {NO}}_3^-} $], ~119 μmol/L) still seems to be a key source of nitrate in the coastal water.

In conclusion, based on the relative DIN (mainly ${{\rm {NO}}_3^-} $) contributions from different water sources to Hangzhou Bay (Fig. 7) and the overall ranges of δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ (5.3‰ to 7.3‰) and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $ (−0.4‰ to 3.5‰) in our study area, we suggest that the ${{\rm {NO}}_3^-} $ in Hangzhou Bay in late August 2019 was likely derived mainly from soils, synthetic N fertilizer, and manure and sewage.

5. Conclusions

Over most of Hangzhou Bay in late August 2019, DIN ranged from 54 μmol/L to 149 μmol/L, with nitrate accounting for more than 95% of DIN in most areas. The [DIN]:[${{\rm {PO}}_4^{3-}} $] ratio ranged from 40 to 56, suggesting that excessive input of DIN to Hangzhou Bay has become a serious environmental problem. The spatial distributions of nutrients (nitrate, nitrite, ammonium, and phosphate) and the signatures of the stable isotopes of nitrate (δ¹⁵${\mathrm{N}}_{{\mathrm{NO}}_3^-} $ and δ¹⁸${\mathrm{O}}_{{\mathrm{NO}}_3^-} $) together suggest that the DIN in Hangzhou Bay originated mainly from soils, synthetic N fertilizer, and manure and sewage. In the upper bay, the Qiantang River was a significant DIN source, possibly accounting for more than 50% of DIN in the bay head area. In the lower bay, coastal water inputs contributed most of the DIN. DIN coming directly from the Changjiang River made a relatively small contribution to the Hangzhou Bay DIN at that time.