On influencing factors of hypoxia in waters adjacent to the Changjiang estuary

Author’s Accepted Manuscript On influencing factors of hypoxia in waters adjacent to the Changjiang estuary Xiaofan Luo, Hao Wei, Renfu Fan, Zhe Liu, Liang Zhao, Youyu Lu www.elsevier.com/locate/csr

PII: DOI: Reference:

S0278-4343(16)30669-0 https://doi.org/10.1016/j.csr.2017.10.004 CSR3684

To appear in: Continental Shelf Research Received date: 19 December 2016 Revised date: 30 August 2017 Accepted date: 10 October 2017 Cite this article as: Xiaofan Luo, Hao Wei, Renfu Fan, Zhe Liu, Liang Zhao and Youyu Lu, On influencing factors of hypoxia in waters adjacent to the Changjiang estuary, Continental Shelf Research, https://doi.org/10.1016/j.csr.2017.10.004 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

On influencing factors of hypoxia in waters adjacent to the Changjiang estuary Xiaofan Luoa, b, Hao Weia, *, Renfu Fana, Zhe Liub, Liang Zhaob, Youyu Luc a

School of Marine Science and Technology, Tianjin University, Nankai District, Tianjin, 300072, China

b

College of Marine Environmental Sciences, Tianjin University of Science & Technology, Tianjin, 300457, China

c

Ocean and Ecosystem Sciences Division, Fisheries and Oceans Canada, Bedford Institute of Oceanography,

Dartmouth, Nova Scotia, B2Y 4A2, Canada

Abstract Based on observational data from ten cruises carried out in 2012 and 2013, the distribution of dissolved oxygen (DO) and the evolution of hypoxia (DO concentrations <2.0 mg L-1) in waters adjacent to the Changjiang estuary are studied. The linkage between summer hypoxia and hydrodynamic conditions is explored. The results suggest that hypoxia frequently occurred from June to October to the south of the Changjiang estuary near the 30-50 m isobaths and was prone to happening under strong stratification without the presence of the Kuroshio Subsurface Water (KSW). Over the Changjiang Bank, hypoxia mainly occurred in July, August and September. Low-oxygen areas initially appeared under strong stratification induced by the spreading of the Changjiang Diluted Water (CDW), and developed into hypoxic zones due to lack of DO replenishment from the relatively DO-rich Yellow Sea Water and the KSW. The yearly evolution of hypoxia was influenced by shelf circulation especially the path of the KSW in the bottom layer of the water to the south of the Changjiang estuary, and the extension of the CDW in ______________________________ *Corresponding author. Tel.: +86 022 27409515 (This author contributed equally to this work with the first author and should be considered co-first author.) E-mail address: [email protected](H. Wei). 1

the surface layer over the Changjiang Bank.

Keywords: Hypoxia, Changjiang estuary, Kuroshio Subsurface Water, Changjiang Diluted Water, Space-time variability, Data analysis.

1 Introduction The life of marine creatures depends on dissolved oxygen (DO) in the water and can be threatened by DO concentrations ([DO]s) less than 3.0 mg L-1 (e.g., Howell and Simpson, 1994; Breitburg et al., 1997). In coastal waters, the increasing occurrence of extremely low [DO]s, i.e. hypoxia ([DO]s <2.0 mg L-1), is becoming a global environmental issue (e.g., Diaz and Rosenberg, 2008; Conley et al., 2009). One of the main goals of regional environment management is to monitor and control the area and volume of hypoxia, and this requires the understanding and prediction of low-oxygen evolution (Feng et al., 2012). It is generally agreed that stratification and organic matter degradation are main causes for the formation of hypoxia. For estuarine regions, there have been continuous debates on the roles played by the riverine nutrient loads and freshwater discharge (Bianchi et al., 2010).

Since the middle of last century, seasonal surveys have revealed the occurrence of hypoxia in the lower water column adjacent to the Changjiang estuary, an area with a high level of primary production on the shelf of East China Sea (ECS) (Fig. 1) (Gu, 1980; Chen et al., 1988; Tian et al., 1993; Zhao et al., 2001). However, hypoxia was not identified as an important factor affecting the ecosystem in this region until the studies carried out by the Chinese GLOBEC (Global Ocean Ecosystem Dynamics) project in the summer of 1999 (Li et al., 2002). Subsequently, summer 2

hypoxia in this region has been continuously reported based on sparse ship-based observations (Shi et al., 2006a; Wei et al., 2007; Wang, 2009). Currently, the Changjiang estuary has been regarded as one of the largest coastal hypoxia areas in the world (Chen et al., 2007).

The spatial and temporal variations of [DO]s adjacent to the Changjiang estuary have been demonstrated by analyses of observational data collected so far. From May to October in 2006, intense observations were made in this region by various research teams (Zhang et al., 2007; Zhou et al., 2010; Li et al., 2011; Wei et al., 2011; Wang et al., 2012). Data from these surveys revealed that low [DO]s (< 3 mg L-1) were absent in May but a zone of depleted oxygen appeared in June in waters along the coast to the south of the Changjiang estuary. In July, a hypoxic zone appeared in the western part of the Changjiang Bank. Two hypoxic zones developed in August, one to the north and the other to the south of the Changjiang estuary, with the northern one being more severe. The hypoxic zone to the south of the Changjiang estuary persisted into September and October. A re-analysis of the monthly data from September 1958 to September 1959 (Liu et al., 2012) demonstrated that the low-oxygen area occurred to the south of the estuary in May and June, near the estuary in July and August, and back to the south of the estuary in September. This evolution of the low-oxygen area showed correspondence with the extension of the Taiwan Warm Current (TWC) which carries Taiwan Strait Water in the upper layer and the Kuroshio Subsurface Water (KSW) in the lower layer. The lowest DO value of 0.34 mg L-1 was observed in August of 1959, a highly severe hypoxic event that had drawn little attention for about 40 years.

Generally, the formation of hypoxia is related to high rates of primary production, long resident

3

times of bottom water, persistent presence of stratification and specific bathymetry of continental shelf (Rabouille et al., 2008). For hypoxia in waters adjacent to the Changjiang estuary, previous studies have identified the important roles played by the pycnocline between the freshwater from river discharge and the upwelled water from the TWC, and the abundant bottom detritus from surface phytoplankton blooms (Li et al., 2002; Zhang et al., 2007). However, the presence of pycnocline and rich detritus does not consistently correspond to the existence and spatial evolution of hypoxia. Zhou et al. (2010) analyzed the bottom [DO]s, water temperature and salinity in waters adjacent to the Changjiang estuary in June, August, October of 2006, and made comparison with data in August of 1998 and 1999. They concluded that the temporal evolution of hypoxia is mainly associated with variations in the distribution of water masses. Based on observations in 2006, Wei et al. (2011) suggested that the northward spreading of hypoxia is caused by eutrophication off the Changjiang estuary, intensification of the TWC, and the northward shift of the upwelling region. Based on analysis of bottom DO distributions in January and in each month from April to November in 2006, Wang et al. (2012) concluded that the location of hypoxic zone is controlled by stratification and the northward expansion of the TWC. Zhang et al. (2012) suggested that stratification caused by the Changjiang Diluted Water (CDW) is the main cause of hypoxia formation while waters advected from the Yellow Sea (YS) tends to weaken hypoxia; and the competition of the two processes induces the spatial and temporal variations of bottom [DO]s. Based on observations from June to August, and October in 2006, and from August to September in 2009, Zhu et al. (2015) related the occurrence of hypoxia to the presence of pycnocline. Through analyzing observational data from 2006 to 2007 and simulation results of an ocean circulation model, Wei et al. (2015b) suggested the plausible influence of seasonal variations of

4

circulation on the changes of hypoxia position in regions adjacent to the Changjiang estuary.

Previous studies have reached a consensus that the movement of water masses and the change of stratification induced by ocean circulation variations both influence the development of hypoxia in waters adjacent to the Changjiang estuary. However, there lacks an agreement on which factor, among the TWC, the CDW, fronts and the Yellow Sea Water (YSW), plays the dominant role. It is possible that each previous study only captured one aspect of the multifaceted mechanism of hypoxia evolution, due to the limited spatiotemporal coverage of the particular dataset being analyzed. For instance, the hypoxia in summer of 2006 was located farther north relative to 1959, but did this northward shift occur frequently in recent years? From the beginning to the end of summer, does the location of hypoxia regularly migrate from the south to the north and then back to the south? How does hypoxia evolve from formation to decay? Besides the blocking of oxygen exchange by stratification and fronts, how are DO distribution and hypoxia progression influenced by the TWC accompanied with the KSW? What role does the YSW play?

A synthesis of hypoxia data reported during the last decade (Fig. 1 and Table 1) identifies significant year-to-year variations in the location of hypoxia in waters adjacent to the Changjiang estuary. This study attempts to further explore the influence of changes in hydrodynamic conditions on the evolution of hypoxia through analyzing new observations carried out in 2012 and 2013. The bottom DO concentrations in these two years are mapped, and distributions of stratification and hydrography are described. By linking the DO distribution to the variations of hydrodynamic conditions, we analyze the relative contributions of the TWC, the CDW and the YS

5

tidal front to the formation of hypoxia in this region, with a special attention paid to the influence of the lateral advection of the KSW.

2 Field observations and data processing DO and hydrographic data are obtained from 10 cruises carried out in June, August and October in 2012, and monthly from May to September in 2013 (Table 2). The data for May and July in 2013 merge observations made from two synchronized ship-based surveys in each month. A RBR 620CTD with a Seapoint sensor of DO was used in all 10 cruises to measure profiles of water temperature, salinity, and DO concentration. An SBE-CTD was deployed in 8 cruises excluding that in May of 2013 to the north of the Changjiang estuary and that in July of 2013 in the estuary. When both the RBR620-CTD and the SBE-CTD were employed, the profiles of water temperature and salinity from the SBE-CTD are chosen because of the higher sampling frequency and precision of this instrument.

Data recorded during descending profiling are used. For each profile, the inverse pressure calibration is applied and spikes are removed as the basic data quality control. For each cruise, about 100 water samples were taken from the surface, middle and bottom layers using the carousel water sampler of the SBE-CTD. Salinity of these samples was measured with a SYA2-2 salinometer in laboratory to calibrate the CTD measured salinity in each cruise. The DO sensor on the RBR620-CTD was calibrated prior to each cruise as follows (Ruskin User Guide: https://rbrglobal.com/products/software). The sensor was first immersed into fully oxygen-aerated pure water to get the voltage of 100% DO saturation, and then into anoxia water (i.e., saturated sodium

6

sulfite solution) to get the voltage of zero DO reading. The [DO] was determined according to the linear scaling between these two voltages. During some cruises, the chemistry group also used the Winkler Titration method to measure [DO]s of water samples from different layers at selected stations. This enabled an alternative calibration of the [DO]s measured by the DO sensor. Data was sampled at high frequencies, for instance, 24 Hz for the SBE-CTD and 8 Hz for the RBR-620 CTD. The raw data was averaged over 0.2 m vertical intervals. The buoyant frequency (N2) was calculated from water temperature and salinity (Jackett and McDougall, 1995). The maximum N2 in a profile was used to represent the stratification intensity of the water column at a particular station (N2 denotes the maximum N2 in the following).

Stations to the south of 33° N are chosen for analyzing hypoxia in waters adjacent to the Changjiang estuary (Fig. 1). Sampling grids and sections for each cruise are shown in Fig. 2. To facilitate discussion, we divide waters adjacent to the Changjiang estuary into two regions (Fig.1) based on hydrodynamic conditions and topography features. Region I covers the fairly flat Changjiang Bank with depths less than 50 m. Region II covers the region to the south of the Changjiang estuary including a steep slope that is located approximately between the 30 and 50 m isobaths, the submarine river canyon, and part of the middle shelf of the ECS with depths between 50-100 m.

According to Su et al. (1996), salinity can be used to identify water masses in the ECS, with S <30.0 representing the diluted water, and S >34.0 for the TWC water that includes the core part of the KSW with S >34.5. In the following KSW specifically refers to the core part that keeps the

7

water properties most similar to source water. In general, the 17.0℃ isothermal can serve as the boundary between the YS cold water and its surrounding warmer water (Mao et al., 1964). Thus, we take the area with bottom temperature in the range of 13.0-17.0℃ to represent the summer tidal front of the YS (Zhao, 1985). According to China’s 1992 National Marine Investigation Standard (State Bureau of Technical Supervision, 1992), a pycnocline is defined to exist if the vertical gradient of density is larger than 0.1 kg m-4 for the region with total water depths less than 200 m. This is equivalent to the maximum N2 of a vertical profile being larger than 1.0×10-3.0 s-2.

3. Results 3.1. Bottom [DO]s in June, August and October 2012 There was no hypoxia in June 2012 (Fig. 3a). The bottom [DO]s were larger than 4.0 mg L-1, with values higher in the eastern part than in the western part of the survey area. A band of relatively low [DO]s of 4.0-5.0 mg L-1 was found along the 30 m isobath to the north of 29° N.

In August (Fig. 3b), the [DO]s were 2.0-3.0 mg L-1 lower than that in June. There were only a few stations located in Region II. Over the western side of Region I, [DO]s <3.0 mg L-1 were observed. Hypoxia appeared in the northwestern part of the survey area (station I1, at 33° N, 122.5° E).

In October, the [DO]s of bottom water were generally higher than in June and August (Fig. 3c). In Region II, two relatively low-oxygen spots were present, one located near the northern end of the submarine river canyon with [DO] = 4.8 mg L-1 (station PN1), and the other on the steep slope with [DO] <4.0 mg L-1, about 2.0 mg L-1 lower than that in the surrounding waters. 8

Overall, in summer of 2012 hypoxia off the Changjiang estuary was not severe. There was no hypoxia in June and October, though low-oxygen area existed. Hypoxia occurred in Region I in August.

3.2. Bottom [DO]s in May - September 2013 In May, the vertical (data not shown) and horizontal distributions of DO were fairly homogenous (Fig. 3d). Bottom DO values were 8.0-9.0 mg L-1 in the near-shore region and 7.0-8.0 mg L-1 in the offshore region. The minimum DO value, located at station kb, was 6.7 mg L-1.

In June, DO values at most stations in Region I were less than 5.0 mg L-1 (Fig. 3e). In Region II, [DO]s were lower than 4.0 mg L-1 and a hypoxic zone appeared near the slope between the 30-50 m isobaths. Along the 122.5° E section (black lines in Figs. 2a, e), [DO]s in the lower water column in Region I (31° N - 33.5° N) were substantially higher than that in Region II (28.5° N 31° N) where the hypoxic water had a thickness of about 10 m (Fig. 4b). In addition, bottom [DO]s in June 2013 were overall 2.0-3.0 mg L-1 lower than that in June 2012 (Figs. 3a, e and Figs. 4a, b), and hypoxia occurred earlier in 2013 than in 2012. Previously, the occurrence of hypoxia in June was only reported in 2003 in the submarine river canyon near 30.45° N (Xu, 2005).

In July, bottom [DO]s were larger than 7.0 mg L-1 in the Changjiang estuary and Hangzhou Bay, except that [DO]s <4.0 mg L-1 were found at stations X5 and X6 near the northern corner of the Changjiang estuary (Fig. 3f). In Region I, at all stations with depths greater than 30 m, bottom

9

[DO]s were less than 3.0 mg L-1; severe hypoxia occurred within 32-32.4° N, 122.4-123.5° E over the western part of Region I. The huge area with [DO]s <3.0 mg L-1 extended eastward. Over the middle shelf of the ECS, DO values were 3.0-5.0 mg L-1. Overall, [DO]s in the southern and eastern parts of the survey area were higher than that in the middle part.

In August, the survey area was smaller than that in July (Fig. 3g). In Region I, [DO]s <2.0 mg L-1 were observed at five stations and the hypoxic area was separated into several patches extending northward to 33° N and eastward to 125° E. Along section K (at 32° N, green lines in Figs. 2b, g), DO distribution showed that the thickness of the hypoxic water reached 20-30 m from stations K2 to K6, extending about 100 kilometers in length (Fig. 4d), with DO values much lower than in August 2012 (Fig. 4c). In Region II, a low-oxygen zone of [DO]s <3.0 mg L-1 was located in the northern end of the submarine river canyon.

In September, bottom [DO]s over the western part of Region I increased quickly, while [DO]s <3.0 mg L-1 were still found near the northeastern and eastern edges of this region with hypoxia occurring near 33° N (Fig. 3h). Another area with [DO] <3.0 mg L-1 was in the river canyon in Region II.

In 2013, hypoxia first appeared in Region II in June and was sustained in Region I in July, August and September. Multiple low-oxygen areas were observed from June to September. In Region II, the hypoxic zone was stably located in the coastal water near the 30-50 m isobaths. In Region I, the positions of hypoxia changed around the bank. The near-bottom increase of [DO]s at Q3 (Fig.

10

4b) may be related to the KSW northward intrusion. However, the causes for the similar nearbottom increase at K4 (Fig. 4d) are not clear. Overall, compared with historical reports, in the summer of 2013 the hypoxia occurred earlier, occupied a larger area, had a larger thickness, and was maintained longer over the Changjiang Bank.

4. Discussion 4.1 Hydrodynamic conditions Figures 5-8 present the spatial distribution and temporal evolution of N2, bottom salinity, surface salinity and bottom temperature observed from the 2012 and 2013 surveys. The hydrographic conditions in the two years show a general consistency with the climatology based on historical observations (Editorial Board for Marine Atlas-Hydrology, 1992), but also show notable year-toyear differences. Spatial and temporal variations of hydrography in this region are influenced by bathymetry, surface forcing associated with monsoons, runoff of the Changjiang River, and circulation patterns including Kuroshio, TWC, YSC and coastal currents (Su, 2001).

The northward TWC carries Taiwan Strait Water in the upper layer characterized with high temperature and moderately high salinity (S >34.0), and the KSW in the lower layer with low temperature and high salinity (S >34.5) (Weng and Wang, 1985). These two distinctly different water masses contribute to a sustained stratification within TWC in summer, and this leads to relatively low [DO]s in the TWC bottom water due to oxygen consumption. The KSW can be clearly identified by its characteristic [DO]s of 5.0-6.0 mg L-1 (Pan et al., 1993; Liu et al., 2012). Hence, the KSW provides the relatively low- and high-oxygen water before and after the

11

formation of hypoxia, respectively. In 2012, the KSW intruded northward beyond 31° N in June and then reached at the northern corner of the Changjiang estuary in August, covering the whole Region II and the southern part of Region I (Figs. 6a-b), and retreated in October (Fig. 6c). In 2013, the intrusion of KSW was much weaker than in 2012 (Fig. 6). The further-most northward intrusion occurred in July 2013 (Fig. 6f).

In Region I, a strong halocline is created by the fresher CDW (S <30.0) in the upper layer and the saltier lower layer water. The CDW extension is mainly influenced by the advancement and recession of the TWC, especially its inner branch that flows in parallel to the 50 m isobath into the river canyon and reaches the northern corner of the Changjiang estuary (Weng and Wang, 1985; Wei et al., 2015a). When the TWC intrudes further north over the continental shelf, the CDW tends to veer northeastward substantially (Wei et al., 2015a). This is the situation that occurred in 2012 (Figs. 6a-c and Figs. 7a-c). In 2013, with a weaker northward intrusion of the TWC combined with strong southerly wind (Wei et al., 2015a), the CDW expanded eastward significantly in July and resulted in strong stratification in the whole Region I (Figs. 5f and 7f).

The tidal front in the YS forms in May corresponding to surface heating and gets intensified gradually in the following three months. In Region I, this front is usually located near the northeastern edge and extends southeastward (Fig. 8). The YSW with relatively high [DO]s is transported by the YSC into Region I along the eastern side of the front. Here we use bottom T <15.0℃ to identify this water mass. In addition, the current along the western coast of the YS within the 20 m isobath (i.e., SBCC in Fig. 1) flows northward as a response to the wind in

12

summer (Liu and Hu, 2009). Both the tidal front and the northward coastal current are main obstacles for DO replenishment from the YSW to the low-oxygen water in most parts of Region I.

In summary, the hydrodynamic conditions, especially the TWC and the CDW, play a leading role in the evolution, duration and intensity of stratification in waters adjacent to the Changjiang estuary. According to laboratory experiments, Liu et al. (2012) proposed that the main cause for hypoxia formation was not oxygen consumption but the lack of replenishment after consumption. Hence hypoxia may occur in any area where stratification is sustained over a considerable duration without DO replenishment.

4.2. CDW spreading and stratification in Region I In Region I, the low-oxygen area is located where stratification is strong, and the hypoxic zone is influenced by the extension of the CDW.

In June 2012, both strong stratification and relatively low oxygen occurred in the western part of this region (Figs. 3a, 5a); strong stratification was maintained until August and [DO]s decreased to less than 3.0 mg L-1 (Figs. 3b, 5b). The CDW veered northeastward substantially (Fig. 7b), and below it a salinity front was produced as the TWC encountered the fresher coastal water (Fig. 6b). To the east of the salinity front, there existed a tidal front (Fig. 8b). The DO-rich YSW was transported to the eastern area along this tidal front, preventing DO replenishment to the north of the salinity front in the western part. This resulted in the development of the earlier low-oxygen into hypoxia (Fig. 3b). In October, the CDW switched back to flowing southward along the coast.

13

As stratification disappeared (N2 <10-3.5 s-2), bottom [DO]s increased to 7.0 mg L-1 (Figs. 3c, 5c).

In the summer of 2013, both oxygen depletion and hypoxia formation over the Changjiang Bank showed correspondence with strong stratification (N2 >10-2.0 s-2) (Fig. 3 and Fig. 5). In July, the CDW expanded eastward, covering a zone spanning from the Changjiang estuary to Jeju Island. This facilitated the formation of a large low-oxygen area with [DO]s <3.0 mg L-1 over the whole bank under strong stratification (Figs. 3f, 5f, 7f). In the meanwhile, DO supplement from the KSW and the YSW inhibited the formation of hypoxia in the southern and eastern bank, respectively (Figs. 6f, 8f). Hence, hypoxia events only happened in the western bank without DO supplement (Fig. 3f). In August, the KSW did not approach the Changjiang estuary, but retreated southward (Fig. 6g). Meanwhile, the northward migration of the tidal front (Fig. 8g) limited the DO replenishment from the YSW. Consequently, a severe hypoxia over the eastern bank evolved from low-oxygen conditions with sustained stratification (Figs. 3g, 5g). In September, the CDW shrank, and stratification was broken almost everywhere (Figs. 5h, 7h), leading to increases in [DO]s (Fig. 3h). Near the northeastern edge of Region I, however, stratification (N2 >10-3.0 s-2) still persisted due to the vertical temperature gradients, and therefore low [DO]s (<3.0 mg L-1) were maintained (stations I4, I5 and K6) with hypoxia occurring at the northeastern location (station I5).

In the bottom water of Region I, the YSW (T <15.0 °C), CDW (S <30.0), TWC water (34.0< S <34.5) and KSW (S >34.5) can be clearly identified, while the water of other salinity ranges is regarded as the Shelf Mixed Water (SMW) (Fig. 9a). Here, the TWC water excludes the KSW. Generally, the YSW in the north and KSW in the south provide relatively DO-rich water in this

14

region. Thus, hypoxia was unlikely to develop in areas occupied by the YSW and KSW, but usually formed in the SMW (Fig. 9a). This indicates that there is a limitation in using bottom temperature and salinity to track the development of hypoxia. The relationship of [DO]s with surface salinity and stratification (Fig. 9c) shows that the overall lower [DO]s and hypoxia appeared under the CDW in summer. Hypoxia developed only when stratification was strong, while DO values were always larger than 4.0 mg L-1 when stratification was very weak (N2 <10-3.0 s-2). However, high [DO]s did occur under strong stratification, suggesting that stratification is a necessary but insufficient condition for the formation of hypoxia over the bank. Despite this, the bottom DO concentration in Region I with depths greater than 30 m was overall negatively correlated with the strength of stratification in summer and autumn (Fig. 10a). In 2013, the lowoxygen area was initially formed under strong stratification associated with the CDW. Then [DO]s decreased continuously with the persistence of stratification even though the stratification weakened slightly (Fig. 10c). In areas without the replenishment of DO-rich water, low-oxygen evolved into hypoxia (Fig. 10c). In 2012, the pattern of DO evolution was essentially the same as that in 2013 (Fig. 10b). When stratification totally disappeared, DO recovered to higher values (Fig. 10b).

In summer, the CDW spreads over the bank but the orientation of its extension may vary. For instance, the CDW spread northward in June of 2012, and eastward in July of 2013. The largest values of N2 were reached when the CDW lay over salty waters. Overall, hypoxia usually formed over the northwestern Changjiang Bank when the CDW extended northwestward, and in the eastern part when the CDW spread eastward substantially.

15

4.3. KSW intrusion and stratification in Region II In Region II, the occurrence of low-oxygen zones is consistent with the distribution of strong stratification (Fig. 5), and the evolution of hypoxia can be mainly influenced by the lateral transport of the KSW.

A low-oxygen zone often first occurs in the confluent area of the coastal current (i.e., MZCC in Fig. 1) and the TWC, which is usually located between the 30 m and 50 m isobaths where stratification is stronger and more persistent than over the middle shelf of the ECS. In June and August of 2012, because of the intensive KSW intrusion toward the northern corner of the Changjiang estuary (Figs. 6a, b), hypoxia did not form despite the presence of strong stratification (N2 >10-2.5 s-2) (Figs. 5a, b). In October of 2012, owing to the TWC recession and surface cooling, stratification was overall weak and DO increased (Figs. 3c, 5c, 6c).

From May to September of 2013, the coastal area was strongly stratified (Figs. 5d-h). In May, the cumulative oxygen consumption was insufficient to substantially reduce [DO]s, despite the development of stratification. In June 2013, the southern branch of the CDW expanded farther relative to June 2012 (Figs. 7a, e), and the TWC reached 30.5° N (Fig. 6e). This led to large values of N2 (N2 >10-2.0 s-2) in the coastal area of Region II (Fig. 5e). With the presence of sustained stratification (May to June), [DO]s were rapidly depleted and a large low-oxygen area was formed. Meanwhile, the KSW was located to the south of 28° N (Figs. 6e, 8e) and could not provide DOrich water to that area. Consequently, low-oxygen developed into hypoxia (Fig. 3e). In July, with

16

the KSW intruding northward and occupying the majority of the bottom water in Region II (Fig. 6f), hypoxia faded away and did not re-appear afterward. Thus, the timely replenishment of the KSW could prevent the evolution of low-oxygen into hypoxia.

The bottom water of Region II is mainly composed of TWC water (34.0< S <34.5) and KSW, and with additions of Coastal Water (CW: S <31.0) and SMW (Fig. 9b). The distribution of DO against bottom T-S (Fig. 9b) indicates that hypoxia did not occur when S >34.5, i.e., in the KSW. TWC usually provides relatively low DO water to Region II in summer. With a sustained stratification, this low DO water tends to develop into hypoxia in the absence of the KSW. For example, hypoxia in the coastal area of Region II defined in the present study was maintained from August to October 2006 when the KSW did not intrude to the coastal area to the north of 29°N (Zou et al., 2008; Zhou et al., 2010; Wang et al., 2012). Another example is August 1959 when hypoxia occurred in the northern end of the canyon where bottom salinity was 33.0 (Liu et al., 2012).

4.4. An inter-comparison with hypoxia in the Gulf of Mexico Worldwide, the frequent occurrence of hypoxia has been related to eutrophication (Conley et al., 2009). For the Gulf of Mexico, a statistical model for hypoxia prediction has been established based on nutrient loads from Mississippi River (Turner et al., 2006). The inter-annual variation of the extent of hypoxia waters has a strong correlation with the volume of river discharge in the northern Gulf of Mexico in spring and summer (Rabouille et al., 2008). However, for waters adjacent to the Changjiang estuary, hypoxia is not directly related to the river discharge. In summer of 2006 the Changjiang runoff was extremely low (Fig. 11) while severe hypoxia

17

occurred (Zou et al., 2008; Zhou et al., 2010; Wang et al., 2012). In summer of 2012 the runoff was among the highest in the recent decade, while only a small area of hypoxia existed in August (Figs. 3a-c and Fig. 11). Compared with 2012, the runoff in 2013 was considerably lower (Fig. 11) while hypoxia occupied extremely large areas and sustained much longer (Figs. 3d-h). Clearly, the area or volume of hypoxia adjacent to the Changjiang estuary cannot be predicted based on river runoff. This is because, based on our analysis, hypoxia in the ECS tends to occur under a sustained stratification created by the CDW extension in the absence of DO replenishment. The CDW extension is not determined by the runoff (Wei et al., 2015a).

Over the Texas-Louisiana shelf in the northern Gulf of Mexico, the formation and destruction of hypoxia are primarily a local vertical process, and biological processes are responsible for the variation of hypoxia from east to west (Hetland and DiMarco, 2008; Bianchi et al., 2010). Hypoxia is predominantly caused by water column respiration in the eastern region with steep shelf break, and by benthic respiration over the wide shelf to the west. Similarly, the hypoxia area adjacent to the Changjiang estuary can be divided, according to bathymetry, into an area on the wide bank (Region I) and an area with steep slope (in Region II). Over the Changjiang Bank, lowoxygen is related to the presence of the CDW, while the CDW’s position is influenced by the TWC (Wei et al., 2015a). Strong tidal mixing and wind may both induce the detachment of freshwater from the CDW (Xuan et al., 2012). This results in stronger changes of the river plume on the Changjiang Bank, compared with that on the west part of the Texas-Louisiana shelf where tides are relatively weak. In Region II of the ECS, the occurrence of low-oxygen centers near the steep slope is related to bathymetric features, and the bottom DO in this region is significantly

18

influenced by the shelf circulation including the TWC (stratification) and shoreward intrusion of the Kuroshio (DO replenishment).

Circulation conditions in the northern Gulf of Mexico and the ECS are distinctly different. The absence of strong offshore currents in the Mississippi margin results in a long residence time of the bottom water, in favor of biological respiration and hence frequent occurrence of hypoxia over a large spatial scale (Rabouille et al., 2008). Circulation in the ECS is strongly influenced by the presence of the strong western boundary current (Kuroshio). This results in a shorter residence time of the bottom water and hence less regular occurrence of hypoxia. Thus, in water adjacent to the Changjiang estuary, physical processes dominate over biological processes in driving the evolution of hypoxia. The prediction of low-oxygen requires information about salinity distribution at the bottom in Region II and at the surface in Region I, that defines the extensions of the KSW and the CDW, respectively.

5. Conclusions This study demonstrates that the mechanisms dominating the evolution of stratification and DO distribution are different in Region I (the Changjiang Bank) and Region II (the region to the south of the estuary):

a) In Region I, there is a high possibility that hypoxia occurs from July to September. Bottom DO concentration is negatively correlated to the intensity of stratification. The appearance of lowoxygen areas is mostly related to the spreading of the Changjiang Diluted Water. The combination

19

of the coastal current (SBCC) and tidal front limits DO replenishment from surrounding waters. Therefore, under the condition of sustained stratification, low-oxygen tends to evolve into hypoxia. Generally, coastal currents transport the Yellow Sea Water to the Changjiang Bank after the transition of monsoon in autumn, when stratification has already disappeared. Thus, the inflow of the YSW probably plays a minor role in the relief of summer hypoxia over the bank.

b) In Region II, a low-oxygen zone usually exists in the confluent area of the Minzhe Coastal Current and the Taiwan Warm Current near the 30-50 m isobaths, and hypoxia may occur from June to October accompanied by strong stratification. Bathymetry-induced salinity fronts exist throughout the year in the northern end and block the DO exchange with waters to the north. Thus, DO-rich water mainly comes from the south. The Kuroshio Subsurface Water (S >34.5) carried by the inner branch of the TWC can substantially affect the bottom DO concentration in this region. Hypoxia is difficult to form if the KSW appears early in the middle shelf of the ECS and intrudes further northward in summer. This study suggests that hypoxia does not occur at locations with bottom S >34.5.

In summary, based on new observations in 2012 and 2013, especially the 2013 monthly data from May to September, it is evident that hypoxia in waters adjacent to the Changjiang estuary is substantially influenced by shelf circulation, including the CDW plume, the TWC and the KSW intrusion. Due to the lack of monthly observations covering the same region in different years, credible model simulations of inter-annual variations of hydrodynamic conditions are needed for further investigating the linkage of hypoxia with the KSW intrusion and the CDW spreading.

20

Combining observation and modelling shall eventually lead to better prediction of the spatial and temporal variations of hypoxia in the study area.

Acknowledgements This study was supported by Aoshan Science and Technology Innovation Project from Qingdao National Laboratory for Marine Science and Technology (Grant No. 2016ASKJ02), the National Basic Research Program of China (Grant No.2011CB403606), the National Natural Science Foundation of China (NSFC, Grant No.41376112) and the Natural Science Foundation of Tianjin (Grant No.16JCYBJC23000). Liang Zhao was funded by the National Key Research and Development Program of China (Grant No. 2016YFA0601301 and 2016YFC1401602) and the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA11020305). We thank two anonymous reviewers for very constructive comments on the originally submitted manuscript.

References Bianchi, T.S., DiMarco, S.F., Cowan Jr, J.H., Hetland, R.D., Chapman, P.J., Day, W., Allison, M.A., 2010. The science of hypoxia in the Northern Gulf of Mexico: a review, Sci. of the Total Environ., 408, 1471-1484, doi:10.1016/j.scitotenv.2009.11.047. Breitburg, D.L., Loher, T., Pacey, C.A., Gerstein, A., 1997. Varying effects of low dissolved oxygen on trophic interactions in an estuarine food web, Ecol. Monogr., 67(4), 489-507, doi: 10.1890/0012-9615(1997)067[0489:VEOLDO]2.0.CO;2.

Chen, C.C., Gong, G.C., Shiah, F.K., 2007. Hypoxia in the East China Sea: One of the largest

21

coastal

low-oxygen

areas

in

the

world,

Mar.

Environ.

Res.,

64,

399-408,

doi:10.1016/j.marenvres.2007.01.007. Chen, J.Y., Chen, X.L., Yang, Q.L., 1988. Report of Comprehensive Survey on Resources in Coastal Zone of Shanghai, Shanghai Science and Technology Press, Shanghai, China, 390 pp. (in Chinese) Conley, D.J., Carstensen, J., Vaquer-Sunyer, R., Duarte, C.M., 2009. Ecosystem thresholds with hypoxia, Hydrobiologia, 629, 21-29, doi:10.1007/s10750-009-9764-2. Diaz, R.J., Rosenberg, R., 2008. Spreading dead zones and consequences for marine ecosystems, Science, 321, 926-929, doi:10.1126/science.1156401. Editorial Board for Marine Atlas, 1992. Marine Atlas of Bohai Sea, Yellow Sea, East China Sea: Hydrology, China Ocean Press, Beijing. (in Chinese). Feng, Y., DiMarco, S.F., Jackson, G.A., 2012. Relative role of wind forcing and riverine nutrient input on the extent of hypoxia in the northern Gulf of Mexico, Geophys. Res. Lett., 39, L09601, doi:10.1029/2012GL051192. Gu, H.K., 1980. The maximum value of dissolved oxygen in its vertical distribution in Yellow Sea, Acta Oceanol. Sin., 2, 70-80. (in Chinese, with English Abstract.) Hetland, R., DiMarco, S., 2008. How does the character of oxygen demand control the structure of hypoxia on the Texas–Louisiana continental shelf? J. Mar. Syst., 70, 49-62, doi:10.1016/j.jmarsys.2007.03.002. Howell, P., Simpson, D., 1994. Abundance of marine resources in relation to dissolved oxygen in long island sound, Estuaries, 17(2), 394-402, doi: 10.2307/1352672. Jackett, D.R., McDougall, T.J., 1995. Minimal adjustment of hydrographic profiles to achieve

22

static

stability,

J.

Atmos.

and

oceanic

technol.,

12,

381-389,

doi:

https://doi.org/10.1175/1520-0426(1995)012<0381:MAOHPT>2.0.CO;2. Li, D.J., Zhang, J., Huang, D.J., Wu, Y., Liang, J., 2002. Oxygen depletion off the Changjiang (Yangtze River) estuary. Science in China (Series D), 32, 686-694. Li, H.L., Chen, J.F., Lu, Y., Jin, H.Y., Wang, K., Zhang, H.S., 2011. Seasonal variation of DO and formation mechanism of bottom water hypoxia of Changjiang River Estuary, J. Mar. Sci., 29, 78-87, doi:10.3969/j.issn.1001-909X.2011.03.010. Limeburner, R., Beardsley, R.C., Zhao J.S., 1983. Water masses and circulation in the East China Sea, International symposium on sedimentation on the continental shelf, with special reference to the East China Sea, China Ocean Press, Beijing, China, 285-294. Liu, H.X., Li, D.J., Gao, L., Wang, W.W., Chen W.Q., 2012. Study on main influence factors of formation and deterioration of summer hypoxia off the Yangtze River Estuary, Adv. in Mar. Sci., 30, 186-197. (in Chinese, with English Abstract.) Liu, Z.L., Hu, D.X., 2009. Preliminary study on the Huanghai Sea coastal current and its relationship with local wind in summer, Acta Oceanol. Sin., 31, 1-7. (in Chinese, with English Abstract.) Mao, H.L., Ren, Y.W., Sun, G.D., 1964. A preliminary study on the hydrographic characteristics and water types in summer in the Southern Yellow Sea and the East China Sea (28°-37°N, Studia Marina Sinica, 1, 23-77. (in Chinese) Pan, Y.Q., Su, J.L., Su, Y.F., 1993. Seasonal hydrographic characteristics in the Southern East China Sea, Selected papers of investigation and study on the Kuroshio current (Five), Beijing: Ocean Press,186-200. (in Chinese, with English Abstract.)

23

Rabouille, C., Conley, D.J., Dai, M.H., Cai, W.J., Chen, C.T.A., Lansard, B., McKee, B., 2008. Comparison of hypoxia among four river-dominated ocean margins: The Changjiang (Yangtze), Mississippi, Pearl, and Rhone rivers, Cont. Shelf Res., 28, 1527-1537, doi:10.1016/j.csr.2008.01.020. Shi, X.Y., Lu, R., Zhang, C.S., Wang, X.L., 2006a. Distribution and main influence factors process of dissolved oxygen in the adjacent area of the Changjiang estuary in autumn, Periodical of Ocean Univ. of China, 36, 287-290. (in Chinese, with English Abstract.) Shi, JZ., Zhang, S. Y., Hamilton, L.J., 2006b. Bottom fine sediment boundary layer and transport processes at the mouth of the Changjiang estuary, China. J. Hydrology, 327, 276-288, doi:10.1016/j.jhydrol.2005.11.039. Su, J.L., 2001. A review of circulation dynamics of the coastal oceans near China, Acta Oceanol. Sin., 23, 1-16. (in Chinese, with English Abstract.) Su, Y.S., Li, F.Q., Wang, F.Q., 1996. Water type distribution and water regime division in the Bohai, Yellow and East China Seas, Acta Oceanol. Sin., 18, 1-7. (in Chinese, with English Abstract.) The State Bureau of Technical Supervision, 1992. The specification for oceanographic survey & oceanographic survey data processing. (in Chinese) Tian, R.C., Hu, F.X., Martin, J.M., 1993. Summer nutrient fronts in the Changjiang (Yangtze River) estuary, Estuarine, Coastal and Shelf Sci., 37, 27-41, doi:10.1006/ecss.1993.1039. Turner, R.E., Rabalais, N.N., Justic, D., 2006. Predicting summer hypoxia in the northern Gulf of Mexico: Riverine N, P, and Si loading, Mar. Pollution Bulletin, 52, 139-148, doi:10.1016/j.marpolbul.2005.08.012.

24

Wang, B.D., Wang, X.L., 2007. Chemical hydrography of coastal upwelling in the East China Sea, Chinese J. Oceanol. and Limnol., 25, 16-26, doi:10.1007/s00343-007-0016-x. Wang, B. D., 2009. Hydromorphological mechanisms leading to hypoxia off the Changjiang estuary, Mar. Environ. Res., 67, 53-58, doi:10.1016/j.marenvres.2008.11.001. Wang, B.D., Wei, Q.S., Chen, J.F., Xie, L.P., 2012. Annual cycle of hypoxia off the Changjiang (Yangtze River) Estuary, Mar. Environ. Res., 77, 1-5, doi:10.1016/j.marenvres.2011.12.007. Wei, H., He, Y.C., Li, Q.J., Liu, Z.Y., Wang, H.T., 2007. Summer hypoxia adjacent to the Changjiang estuary, J. Mar. Syst., 67, 292-303, doi:10.1016/j.jmarsys.2006.04.014. Wei, H., Luo, X.F., Zhao, Y.D., Zhao, L., 2015a. Intraseasonal variation in the salinity of the Yellow and East China Seas in the summers of 2011, 2012, and 2013, Hydrobiologia, 754, 13-28, doi:10.1007/s10750-014-2133-9. Wei, Q.S., Wang, B.D., Chen, J.F., Xia, C.S., Qu, D.P., Xie, L.P., 2015b. Recognition on the forming-vanishing process and underlying mechanisms of the hypoxia off the Yangtze River estuary, Sci. China: Earth Sci., 45(2), 187-206, doi: 10.1007/s11430-014-5007-0. Wei, Q.S., Zhan, R., Wei, X.H., Zang, J.Y., Wang, S.Q., Wang, Z.X., 2010. Distribution of dissolved oxygen and hypoxia characteristics in the northeast sea of the Yangtze estuary, Adv. in Mar. Sci., 28, 32-40. (in Chinese, with English Abstract.) Wei, Q.S., Yu, Z.G., Xia, C.S., Zang, J.Y., Ran, X.B., Zhang, X.L., 2011. A preliminary analysis on the dynamic characteristics of the hypoxic zone adjacent to the Changjiang estuary in summer, Acta Oceanol. Sin., 33, 100-109. (in Chinese, with English Abstract.) Weng, X.C., Wang, C.M., 1985. A study on Taiwan Warm Current Water, Mar. Sci., 9, 7-10. (in Chinese, with English Abstract.)

25

Xu, S.M., 2005. The distribution and environmental significance of redox sensitive elements off the Changjiang estuary hypoxia zone and its contiguous sea area, Ph.D. Thesis, Ocean University of China. Qingdao, China, unpublished. (in Chinese, with English Abstract.) Xuan, J.L., Huang, D.J., Zhou, F., Zhu, X.H., Fan X.P., 2012. The role of wind on the detachment of low salinity water in the Changjiang estuary in summer, J. Geophys. Res., 117, C10007, doi:10.1029/2012JC008121. Zhang, Y.Y., Zhang, J., Wu, Y., Zhu, Z.Y., 2007. Characteristics of dissolved oxygen and its affecting factors in the Yangtze Estuary, Environ. Sci., 28, 1649-1654. (in Chinese, with English Abstract.) Zhang, Z.Q., 1990. On maximum and minimum dissolved oxygen with bottom layer in Yellow Sea and northern east China Sea in summer. Bulletin of Mar. Sci., 9, 22-26. (in Chinese, with English Abstract.) Zhang, Z., Zhang, Z.F., Han, G.C., Wang, Y., Wang, L.J., He, X., 2012. Spatio-temporal variation, formation and transformation of the hypoxia area off the Yangtze River Estuary, Mar. Environ. Sci., 31, 469-473. (in Chinese, with English Abstract.) Zhao, B.R., 1985. The frontal of the Huanghai Sea Cold Water Mass induced by tidal mixing, Oceanol. et Limnol. Sin., 16, 451-460. (in Chinese, with English Abstract.) Zhao, B.R., Ren, G.F., Cao, D.M., Yang, Y.L., 2001. Characteristics of the ecological environment in upwelling area adjacent to the Changjiang River Estuary, Oceanol. et Limnol. Sin., 32, 327-333. (in Chinese, with English Abstract.) Zhou, F., Huang, D.J., Ni, X.B., Xuan, J.L., Zhang, J., Zhu, K.X., 2010. Hydrographic analysis on the multi-time scale variability of hypoxia adjacent to the Changjiang River Estuary, Acta

26

Oceanol. Sin., 17, 4728-4740. (in Chinese, with English Abstract.) Zhu, J.R., Zhu, Z.Y., Lin, J., Wu, H., Zhang, J., 2015. Distribution of hypoxia and pycnocline off the Changjiang Estuary, China, J. Mar. Sci., http:// dx.doi.org/10.1016/j.jmarsys.2015.05.002. Zhu, Z.Y., 2007. Hypoxia in the Changjiang estuary and its adjacent area-started with phytoplankton pigments, Ph.D. Thesis, East China Normal University. Shanghai, China, unpublished. (in Chinese, with English Abstract.) Zou, J.J., Yang, G., Liu, J.H., Shi, X.F., Fang, X.S., 2008. Distribution characteristics of dissolved oxygen in the sea area adjacent to the Changjiang river in September, Adv. in Mar. Sci., 26, 65-73. (in Chinese, with English Abstract.)

27

Tables Table 1 Records of locations and values of minimum dissolved oxygen off the Changjiang estuary.

Table 2 Information on cruises analyzed in this study

Table 3. Average bottom DO ± SD (mg L-1) and the logarithm of the maximum squared buoyancy frequency (s-2) in Region I from different cruises in 2012 and 2013. K denotes the number of samples.

28

Figures Figure 1. Topography and schematic map of summer circulation in the Yellow Sea and East China Sea. Black arrows denote circulation including the Yellow Sea Current (YSC), Subei Coastal Current (SBCC), Minzhe Coastal Current (MZCC), Taiwan Warm Current (TWC), Kuroshio Subsurface Branch Current (KSBC) and Kuroshio. CE, HB and RC are short for the Changjiang estuary, Hangzhou Bay and submarine river canyon, respectively. Various symbols represent sampling stations of different cruises labelled in the right pane that observed the occurrence of hypoxia. Polygons outlined by green lines are (I) Changjiang Bank and (II) the region to the south of the Changjiang estuary. Contours in gray denote isobaths of 30, 50, 100, and 200 m. Dotted blue line marks the boundary between the Yellow Sea and the East China Sea.

Figure 2. Maps of sampling stations off the Changjiang estuary (a-c: June, August and October in 2012; d-h: May to September in 2013). The layout is designed to ease comparison between 2012 and 2013 for June and August. Black, green and blue bold lines denote section at 122.5°E and section K in Fig. 4, and section PN in Fig. 5, respectively.

Figure 3. Dissolved oxygen concentration (mg L-1) of bottom water in the same layout as Fig. 2.

Figure 4. Dissolved oxygen concentration (mg L-1) along section at 122.5°E (a: June 2012; b: June 2013) and section K (c: August 2012; d: August 2013). 29

Figure 5. Same as Fig. 3 except for maximum squared buoyancy frequency (N2: s-2) at logarithmic scale.

Figure 6. Same as Fig. 3 except for bottom salinity (psu). The contour of S = 30.0 in blue represents the extension of the Changjiang Diluted Water; S = 34.0 in magenta represents the Taiwan Warm Current; and S = 34.5 in red indicates the intrusion of Kuroshio Subsurface Water.

Figure 7. Same as Fig. 6 except for surface salinity (psu).

Figure 8. Same as Fig. 3 except for bottom temperature (°C). Shading area denotes frontal zone.

Figure 9. Variations of DO (mg L-1) corresponding to bottom salinity (psu) and bottom temperature (°C) for (a) Region I and (b) Region II, and (c) corresponding to surface salinity (psu) and the logarithm of maximum squared buoyancy frequency (N2: s-2) for Region I. Dots stand for samplings collected in summer (June to August); diamonds in spring (May) and autumn (September and October). TWCW, KSW, SMW, YSW, CDW and CW are abbreviations of Taiwan Warm Current Water, Kuroshio Subsurface Water, Shelf Mixed Water, Yellow Sea Water, Changjiang Diluted Water and Coastal Water.

Figure 10. The distribution of DO (mg L-1) versus log10 (N2) in Region I with water depth greater than 30 m for (a) all samplings in 2012 and 2013 and separately in (b) 2012 and (c) 2013. Color

30

symbols denote samplings collected during different monthly cruises. Black solid line in (a) denotes the linear regression relationship. In (b) and (c), black arrows illustrate the seasonal evolution of DO in 2012 and 2013, respectively; colored dots denote the averaged values of DO and N2 in different months (also listed in Table 3). The vertical black dashed lines represent N2 =10-3.0s-2, and the horizontal black dashed line in (a) represents [DO] = 2 mg L-1.

Figure 11. Monthly discharge of the Changjiang River during 2000–2013, averaged from daily monitoring data collected at the Datong Hydrologic Station (http://yu-zhu.vicp.net/).

31

Figure 1. Topography and schematic map of summer circulation in the Yellow Sea and East China Sea. Black arrows denote circulation including the Yellow Sea Current (YSC), Subei Coastal Current (SBCC), Minzhe Coastal Current (MZCC), Taiwan Warm Current (TWC), Kuroshio Subsurface Branch Current (KSBC) and Kuroshio. CE, HB and RC are short for the Changjiang estuary, Hangzhou Bay and submarine river canyon, respectively. Various symbols represent sampling stations of different cruises labelled in the right pane that observed the occurrence of hypoxia. Polygons outlined by green lines are (I) Changjiang Bank and (II) the region to the south of the Changjiang estuary. Contours in gray denote isobaths of 30, 50, 100, and 200 m. Dotted blue line marks the boundary between the Yellow Sea and the East China Sea.

32

A1 A2 A3 C1 C2 C3 C4 C5 C6 E1 E2 E3 E4E5 E6 E7 E8

34 (a) N 201206 33 32 31 30 29 28 27

34 (e) N 201306 33

30 I1 I2 I3 I4 I5 K1 K2 K3 K4 K5 M1 M2 M3 M4 PN1 N1 N2 PN2 PN3 PN4 P1 P2 P3 P4 Q1 P5 Q2 Q3 Q4 S1 S2 S3 100

32 50

31 30 29 28

34 (d) 30 N 201305 i4 i5 33

30 I1 I2 I3 I4 I5

d2

K2 K3 K4 K5 M2 M3 PN1 PN2 PN3 PN4 P2 P3 P1 A1 A2 A3 P4 P5 Q3 Q4 C1 C2 C3 C4 C5 C6II S2 S3 E2 E3E4 E5E6 100

32

K1

50

31

d5 ka k2 k3 k4 kb M14~18

M11 M12 M13 M7M8 M10

30

M4 M6

29

M1 M3 HV2 H02

50

H01 N02 A02 N04

28 C01 C02 C03 C04C05 100 1A04

27 27 120 121 122 123 124 125126E 120 121 122 123 124 125126E 120 121 122 123 124 125126E 34 (g) 34 34 (b) 30 30 30 (f) N 201308 N 201307 Z1 N 201208 I1 I4 I5 I1 I2 I3 I4 I5 50 MC1 R1 33 33 33

32

K2 K3 K4 K5 M1 M2 M3 M4

31

PN2 PN3 PN4

32 50

31 30

30 29

29

28

28

K1 K2 K3 K4 K5 K6 M1 M2 M3 M4 50 PN1 PN2 m+ PN3 P3 P1 P2 PN4 Q3 S1 S2 S3

32 31 30 29 28

A1 A2 A3

ME1

C1

ME2 P11 T7 ME3 LS1 C2 C3 C4C5C6 ME4 ME5 T8 LS2 ME6 ME7

E2 E3 E4 E5 E6E7100 E1

100

100

D05 x1 x2 Z2 MC4 F12 x3 x4 MZ9 D03 D04 MT1 F11 D01 yz1 Z4 B2 F10 Z3 yz2 x5 x6 yz3 F08 F09 yz4 w01 MZ11 F03 F04 F05 F07 FP02 FP04 PN1F02 yz5 yz6w05a PN2 FP01 FP03 F06 PN3 w04 x7 yz7 PN4 FP05 w03 P05 P06 T5 w02 P07 ZS P08 T6 P09 FP06

27 27 120 121 122 123 124 125126E 120 121 122 123 124 125126E 120 121 122 123 124 125126E 34  (h) 34 (c) 30 30 N 201309 N 201210 I1 I2 I3 I4 I5 I1 I2 I3 I4 I5 33 33

27

32 31 30

K1K2 K3 K4 K5 M2 M3 M4 50 PN1 PN2 PN3 PN4

32 31 30

N1 N2 N3 N4

Q1 Q2 Q3 Q4

29

29

28

28 27

K1 K2K3K4K5K6 M1 M2 M3 M4 50 PN1 PN2 PN3 PN4

100

27

120 121 122 123 124 125126E

100

120 121 122 123 124 125126E

Figure 2. Maps of sampling stations off the Changjiang estuary (a-c: June, August and October in 2012; d-h: May to September in 2013). The layout is designed to ease comparison between 2012 and 2013 for June and August. Black, green and blue bold lines denote section at 122.5°E and

33

section K in Fig. 4, and section PN in Fig. 5, respectively. 34 N 33

(a) 201206

34 N 33

30

5 67

32

32

8

50

31

5

30

6

29

4

7 6

30

5

29

34 N 33

30

8

31

5 28

10

(e) 9 201306

32

5

2

4

100

9

9

8

6

8

30

5

5

4

4

3

3

28 100

50

6

2

2

100

27 27 1 1 10 10 120  121 122 123 124 125 126E 120 121 122 123 124 125 126E 120  121 122 123 124 125 126E 34 (f) 34 (g) 34 (b) 30 30 30 9 9 N 201307 N 201308 N 201208 50 33 33 33

27

8

5

2 32

32

3 4

50

31

31 30

30

29

29

28

28

2 3

6

3

4

5

8

2

7

32

7

31

6

50

30

4

29

5

3 2

5

28

2

100

3

4

2

30 29 28 27

31 30

6 4 5

50 3

6

50 6

7 5

4

29 28

5

4

3 100

100

2

27 1 120 121  123  124  125  126 E 126E 120 122 121 122 123 124  125

Dissolved Oxygen(mg L-1) at Bottom

31

7

6 5

2

100

31 30

4

29

3

28

2 1

50

5 4

1 10

9 8 7 6 5 4 3

100

27

120 121 122 123 124 125 126E

Figure 3. Dissolved oxygen concentration (mg L-1) of bottom water in the same layout as Fig. 2.

34

7

3

7

7

8

4

4

27 27 1 110 120 121 122 123 124 125 126E  121 122 123 124 125 126E 120 121 122 123 124 125 126E-1 120 Dissolved Oxygen(DO,mg L ) at Bottom 34 (h) 34 (c) 30 30 10 9 N 201309 N 201210 34 (h) 33 33 30 9 N 201309 8 6 54 2 7 3 33 4 6 8 32 32 2 7

6 5

9

5

27

32

1 10

6

8 5

3

3 100

7

31

29

2

8

7

3 28

6

10 30

7

50

3

10

(d) 9 201305

2 1

DO (mg L-1)

DO (mg L-1)

0

0

8

9

9

8

7

7

7

9

6

8

4

-20

5

Depth (m)

Depth (m)

-20

5 6

6

6 5

-40

10

7 8

7

4

3 2

6 5

-40

(a)201206

(b)201306

4 3

-60

-60 Q3

P1

M1

K1

I2

Q3

P1

PN1

K1

2

I2

-70 -70 1 28.5 29.0 29.5 30.0 30.5 31.0 -1 31.5 32.0 32.5 33.0 33.5 28.5 29.0 29.5 30.0 30.5 31.0 -131.5 32.0 32.5 33.0 33.5 DO (mg L( )N) DO (mg L( )N) Latitude Latitude 0 0 10 4 10

8

9

7

3

9 2

6

8

-20

5

Depth (m)

Depth (m)

-20 4 3

-40

7 6 2

5

-40

(c)201208

(d)201308

4 3

-60

-60 K2

-70 122.5

123.0

K3

123.5 Longitude ( E)

K4

124.0

K5

K1

-70 124.5 122.5

K2

123.0

K3

K4

123.5 124.0 Longitude ( E)

K5

124.5

2

K6

125.0

1

Figure 4. Dissolved oxygen concentration (mg L-1) along section at 122.5°E (a: June 2012; b: June 2013) and section K (c: August 2012; d: August 2013).

35

34 N 33

(a) 201206

34 N 33

30

-1.5

32

34 N 33

30

-2.5 -3.0 -3.5

-2.0

-3.0 -3.5

32 31

-2.0

30

-2.0

29 28

29 28

-2.0 100

-3.5

-1.0 30

-3.0 -2.0

29

-4.0

28

-4.5

100

-2.0

-2.0 -2.5

31 30

-1.5

-4.0-3.5 -2.5

-2.5

50

-1.5 -3.0

30

-1.0

(d) -1.5 201305

32

-2.5

50

-2.0 -2.5

31

-1.0

(e) -1.5 201306

-2.5

50

-3.0

-3.0

-3.5

-3.5

-4.0

-4.0

-4.5

-4.5

100

27

27 27 -5.0 -5.0 -1.0 -1.0 120 121 122 123 124 125 126E 120  121 122 123 124 125 126E 120  121 122 123 124 125 126E 34 (g) 34 (f) 34 (b) 30 30 30 N 201308 N 201307 N 201208 -1.5 -1.5 50 33 33 33 -2.0

-1.5

32

-2.0

-2.5

32

-2.5

50

31

-2.0 -3.0 -3.5

-2.5 -1.5

31

-2.5

-2.0 32

-2.5

50

30

29

29

28

28 100

-2.0

-2.0

-3.0

-2.5

-3.0 30

-2.0

-3.5

29

-4.0

28

-4.5

100

-3.0 -2.0

-3.5 -4.0 -4.5

-3.5

-2.5

-2.5 -4.0

-2.0

-4.5

100

27

27 27 -5.0 -5.0 -1.0 N2 126 (s-2 120 121 122 123 log 12410125 E) 120  121 122 123 124 125 126E 120 121 122 123 124 125 126E 34 (c) 34 (h) -1.0 30 30 N 201210 N 201309 -1.5 34 (h) 33 33  30

31

30

30

29

-2.5

-2.5

-3.550 -3.5 -2

-3.0

28

50

-3.0

-2.5

-3.5 -4.0

29

28 27

-4.0 -3.5

100

100 120 121 27 122 123 124 125 126E

N2 (s-2)

32

31

-2.0

-3

10

32

-1.5

log

N 201309 -4.0 -3.5 33

-2.0

32

-2.5

31

-3.0

30

-3.5

29

-4.0

28

-4.5

-4.5 -5.0

-5.0

120 121 122 123 124 125126E

-3.5 -3.5 -2.0

-3.0

50

-2.5

36

-1.5 -2.0 -2.5 -3.0

-4.0 100

27

-5.0 -1.0

-3.5

120 121 122 123 124 125 126E

Figure 5. Same as Fig. 3 except for maximum squared buoyancy frequency (N2: s-2) at logarithmic scale.

-1.5

31

-3.0

30

-2.0

-5.0 -1.0

-4.5 -5.0

34 N 33

(a) 201206

34 N 33

30

32

32

32 50

34 34.5

31

31

30

30

29

29

34 28

(e)>34.5 201306 34.0

34.5

28 100

27

34 N 33

30

32

33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

32

32

50

31 30

34

29

34.5

28 100

(d)>34.5 201305 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

30

32 50

32

26 30

100

27 27 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 34 (g) 34 (b) 34 (f)>34.5 >34.5 30 30 30 N 201308 N 201208 N 201307 50 34.0 34.0 33 33 33 30

32

32

32 50

34 34.5

31

31

30

30

29

29 28

28 100

32

33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

32 50

34

31 30 29 28

100

33.5 26 33.0 32.5 32.0 31.5 30 31.0 30.5 30.0 28.0 <26.0

32 32 34 34.5

100

27 27 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 34 (c) 34  (h) >34.5 34 (h) 30 30 N 201210 N >34.5 201309 30 34.0 N 201309 33 33 34.0

33

30

32

32

32 31 30

50

32

31 30

34

30

50

34.5

29

29

28

28

27

>34.5 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

Bottom Salinity

100

100

33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

Salinity at bottom

27

>34.5 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

120 27   121 122 123 124 125 126E 120 121 122 123 124 125126E

33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

32

32

34

31 30

50

34.5

29 28 100

27

120 121 122 123 124 125 126E

Figure 6. Same as Fig. 3 except for bottom salinity (psu). The contour of S = 30.0 in blue represents the extension of the Changjiang Diluted Water; S = 34.0 in magenta represents the Taiwan Warm Current; and S = 34.5 in red indicates the intrusion of Kuroshio Subsurface Water.

37

>34.5 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

34 N 33

(a) 201206

34 N 33

30

26

28 30 32

32

32 50

26

31 30

31 30

32 34

29

30

29

34.5

28

28 100

27

(e)>34.5 201306 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

34 N 33

30

28

30 32 50

31

26 28

30

30 29

32

28 100

(d)>34.5 201305 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

30

30 28 26 26

50

28

100

27 27 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 34 (b) 34 (f)>34.5 34 (g) >34.5 30 30 30 N 201208 N 201307 N 201308 50 34.0 34.0 30 32 33 33 33 30

26 28

32

32

34 34.5

31

50

31

30

30

29

29

28

28 100

27

30

33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

30 28 32

32

32 50

31

26 30 29 28 100

33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

28 26

28 30

26 32

32

100

27 27 Salinity 120 121 122 123Surface  124 125  126E 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 34 (c) 34 (h) >34.5 30 34 (h) 30 N 201210 N 201309 >34.5 34.0 32 N 201309 30 30 33 33

31 30 29 28 27

32

30

32 31

50 28

32

32

30 29 28 27

100

100

33.5 33.0 32.5 50 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

Salinity at depth=2m

33

32

120 121 122 123 124 125 126E

120 121 122 123 124 125126E

Figure 7. Same as Fig. 6 except for surface salinity (psu).

38

30

32

28

50

31

32 30 29 28 100

27

120 121 122 123 124 125 126E

>34.5 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0 >34.5 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

>34.5 34.0 33.5 33.0 32.5 32.0 31.5 31.0 30.5 30.0 28.0 <26.0

34 (a) N 201206 33

34 (e) >23 22 N 201306 33 21

30

13 15

17

32

32

19

50

31

31

21 30

30

19

29

29

18 28

28

21 100

27

20 19 18 17 16 15 14 13 12 11 10 9 8 <7

34 (d) >23 22 N 201305 33 21

30

13 17

15 19

32 50

31

21

30

19 29

19 18

28 100

20 19 18 17 16 15 14 13 12 11 10 9 8 <7

30

15 50

17 17 18

100

27 27 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 120 121 122 123 124 125 126E 34 (b) 34 (f)>23 34 (g) >23 30 30 30 22 22 N 201208 N 201307 N 201308 50 19 33 33 21 33 21 17 13

15

19

32

32

23

50

21 19

31

31

30

30

29

29

28

28 100

21

17

32

23 23

18

50

31

21 19

30 29 28

100

20 19 18 17 16 15 14 13 12 11 10 9 8 <7

13

23 21

21

25 15 17 19

19 21

19

100

27 27 Temperature 120 121 122 123 124 125 126E 120 121 122Bottom  123 124  125 126E 120 121 122 123 124 125 126E 34 (h) 34 (c) >23 30 30 22 N 201309 N 201210 >23 34 (h) 21 19 33 33 22 30 N 32 31 30 29 28 27

33

201309

21

19 21

50 23

32 31 30

25

23

23 21

23 19

29 28

100

100 12027  121 122 123 124 125 126E

20 19 18 17 50 16 15 14 13 12 11 10 9 8 <7

Temperature(  C) at bottom

27

20 19 18 17 16 15 14 13 12 11 10 9 8 <7

21 20 19 18 17 16 15 14 13 12 11 10 9 8 <7

32

23

25 23 21

31 30

21

50

19

29 28 100

27

120 121 122 123 124 125 126E

120 121 122 123 124 125126E

Figure 8. Same as Fig. 3 except for bottom temperature (°C). Shading area denotes frontal zone.

39

>23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 <7 >23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 <7 >23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 <7

Figure 9. Variations of DO (mg L-1) corresponding to bottom salinity (psu) and bottom temperature (°C) for (a) Region I and (b) Region II, and (c) corresponding to surface salinity (psu) and the logarithm of maximum squared buoyancy frequency (N2: s-2) for Region I. Dots stand for samplings collected in summer (June to August); diamonds in spring (May) and autumn (September and October). TWCW, KSW, SMW, YSW, CDW and CFW are abbreviations of Taiwan Warm Current Water, Kuroshio Subsurface Water, Shelf Mixed Water, Yellow Sea Water, Changjiang Diluted Water and Coastal Fresh Water.

40

Figure 10. The distribution of DO (mg L-1) versus log10 (N2) in Region I with water depth greater than 30 m for (a) all samplings in 2012 and 2013 and separately in (b) 2012 and (c) 2013. Color symbols denote samplings collected during different monthly cruises. Black solid line in (a) denotes the linear regression relationship. In (b) and (c), black arrows illustrate the seasonal evolution of DO in 2012 and 2013, respectively; colored dots denote the averaged values of DO and N2 in different months (also listed in Table 3). The vertical black dashed lines represent N2 =10-3.0 s-2, and the horizontal black dashed line in (a) represents [DO] = 2 mg L-1.

41

5 4 3 2 1

3 20 1

12 20

11 20

10 20

09 20

8

7

20 0

20 0

5

06 20

20 0

04 20

2

3 20 0

20 0

0 20 0

1

Year 0

20 0

Monthly Runoff(unit: 104 m3 s-1)

6

Figure 11. Monthly discharge of the Changjiang River during 2000–2013, averaged from daily monitoring data collected at the Datong Hydrologic Station (http://yu-zhu.vicp.net/).

Table 1. Records of locations and values of minimum dissolved oxygen off the Changjiang estuary. Time

Location

Min. DO

References

value (mg L-1) Longitude

Latitude

Aug., 1959

122°45′

31°15′

0.34

Liu et al., 2012

Aug., 1976–

123°00′

31°00′

0.80

Zhang, 1990

1985

42

Aug., 1981

123°00′

30°50′

2.0

Limeburner et al., 1983

Aug., 1988

123°00′

30°50′

1.96

Tian et al., 1993

Aug., 1998

124°00′

32°10′

1.44

Wang and Wang, 2007

Aug., 1999

122°59′

30°51′

1.0

Li et al., 2002

Aug., 2002

122°29′

32°00′

1.73

Shi et al., 2006a, 2006b

123°00′

31°00′

1.99

Jun., 2003

122°50′

30°50′

~1.0

Xu, 2005

Aug., 2003

123°30′

31°30′

2.0

Chen et al, 2007

Sep., 2003

122°56′

30°49′

0.8

Wei et al, 2007

122°45′

31°55′

<1.5

Aug., 2005

122°48′

32°18′

1.55

Zhu et al, 2007

Jul., 2006

122°23′

32°42′

1.36

Wei et al, 2010

Aug., 2006

122°55′

32°55′

1.0

Zhou et al, 2010

122°30′

29°12′

<2

Sep., 2006

123°06′

30°09′

1.96

Zou et al., 2008

Oct., 2006

123°21′

30°05′

1.8

Zhou et al, 2010

Aug., 2012

122°41

33°00′

1.0

this study

Jun., 2013

122°39′

29°50′

1.5

this study

122°52

29°48′

1.6

122°19′

32°24′

1.7

122°19′

32°03′

1.6

Jul., 2013

43

this study

Aug., 2013

Sep., 2013

122°36′

31°54′

1.8

123°25′

32°10′

1.9

123°44′

33°00′

1.8

123°01′

32°00′

1.9

123°32′

32°00′

1.8

124°28′

32°00′

1.3

124°46′

32°00′

1.2

123°58′

33°02′

1.1

44

this study

this study

Table 2. Information on cruises analyzed in this study. Cruise

Sampling Date

Number of Stations

Instrumentation

Research Vessel

54 stations

SBE 17plus

Beidou

in YECSa

RBR620

42 stations

SBE 17plus

in YECS

RBR620

42 stations

SBE 19plus

in YECS

RBR620

9 stations

RBR620

Suruyuyun288

15 stations off

SBE 911plus

Zhepuyu 23012

Zhejiang Coast

RBR620

48 stations

SBE 911plus

in YECS

RBR620

20 stations in the

RBR620

Runjiang1

Dongfanghong2

and Research Area 201206

201208

201210

201305-I

08-29 Jun.

11-20 Aug.

09-26 Oct.

11-15 May

Beidou

Beidou

in Subei Bank 201305-II

201306

201307-I

11-15 May

15-19 Jun.

16-22 Jul.

Beidou

Changjiang estuary 201307-II

201308

a

12 Jul. - 01

74 stations

SBE 911plus

Aug.

in YECS

RBR620

13 Aug. - 02

63 stations

SBE 17plus

Sep.

in YECS

RBR620

YECS: Yellow and East China Seas 45

Beidou

Table 3. Average bottom DO ± SD (mg L-1) and the logarithm of the maximum squared buoyancy frequency (s-2) in Region I from different cruises in 2012 and 2013. K denotes the number of samples. cruise

DO

N2

K

201206

6.5±1.6

-2.3±0.9

12

201208

3.9±1.1

-2.0±0.5

10

201210

7.1±0.4

-3.9±0.3

11

201306

4.3±0.7

-2.5±0.5

11

201307

2.6±0.6

-1.8±0.3

16

201308

2.3±0.7

-2.1±0.4

13

201309

4.8±1.9

-2.8±0.8

14

Highlights: 

Hypoxia evolution adjacent to Changjiang estuary is control by hydrodynamics



Hypoxia in the slope tends to happen without the Kuroshio Subsurface Water arrival



Hypoxia over the Changjiang Bank changes with the Changjiang Dilute Water extension

46