Abstract
The estuarine plume front is a typical dynamic interface in estuaries, where plume and oceanic waters converge at the surface, subducting ocean water beneath the plume and forming a barrier layer between the surface and bottom water. Phytoplankton blooms, fueled by nutrient-rich freshwater in the plume, generate substantial amounts of sinking organic matter, enhancing bacterial consumption of dissolved oxygen (DO) in the bottom water, but the effects of the frontal convergence on DO remain insufficiently understood. In this study, field surveys were conducted in the Pearl River estuary (PRE) during August 27 and September 7 in 2016, including a 24-h time series at an anchored station (ED) near the front. The results showed that the occurrence of bottom hypoxia at ED coincided with surface phytoplankton blooms (high DO and chlorophyll fluorescence). Three distinct water masses were identified: (1) the Pearl River estuarine water (PRW), where nutrients stimulated blooms, (2) the hypoxic bottom water (BW), receiving sinking organic matter, and (3) offshore surface water (OSW), subducted beneath the plume as a middle layer, preventing PRW–BW mixing. The hypoxic layer average DO correlated negatively with equivalent thickness (ET) of PRW (p < 0.05), which negatively correlated with tidal level (p < 0.01) and OSW ET (p < 0.01), indicating that tidal-dynamics-induced front modulated the OSW barrier effect, exacerbating bottom hypoxia. These findings highlight the critical role of frontal processes in regulating oxygen depletion in stratified estuaries.
1 Introduction
Coastal hypoxia has become a global ocean issue exacerbated by anthropogenic pressures and climate change (). Hypoxia appears to occur more frequently than the last decades in the Pearl River estuary (PRE) (). The formation of hypoxia depends on two main factors: sufficient supply of organic matter for microbial organisms to decompose and consume DO and physical barriers such as halocline, thermocline, or pycnocline to block the exchange between bottom water and surface water.
Chl a in the PRE was predominantly composed of particles that were >5 μm in size fraction (exceeding 50%) (Yin et al., 2004a). The spatial variation of phytoplankton biomass in the PRE exhibited typical characteristics: in the turbid estuarine waters, biomass and productivity were relatively low (less than 20 mg Chl a m-² and less than 100 mg C m-² day-¹, respectively) due to rapid dilution and light limitation; in the region with intermediate salinity within the plume, biomass and productivity reached local maxima (for example, at the edge of the nearshore plume south of Hong Kong, it reached 70 mg Chl a m-² and 4.2 g C m-² day-¹, respectively); in contrast, in the marine waters, biomass was relatively low owing to nutrient limitation (Yin et al., 2004a). A substantial amount of phytoplankton biomass in the estuarine plume (EP) may lead to significant quantities of sinking organic matter, which serves as substrate for microbial respiration and consequently contributes to oxygen consumption in the lower water layers (Rabalais et al., 2002; Rabouille et al., 2008).
Estuarine plume (EP) refers to the low-density surface water mass formed by the continuous freshwater discharge from a river beyond the estuary, whereas the plume front denotes the narrow transition zone between the EP and the ambient high-salinity seawater (Simpson and Sharples, 2012). Convergence between the spreading EP and offshore surface water (OSW) occurs within the continental shelf transition zone, driving the subduction of high-density OSW beneath the low-salinity plume at the frontal interface (). This process forms a stable three-layered structure, which facilitates the accumulation of organic matter and the progression of hypoxia (). found that a mid-depth transitional layer, serving as a barrier layer, hindered the vertical DO exchange between river plume and the shelf salt wedge, which resulted in hypoxia in the bottom layer in the outer range of the PRE. Similarly, Wang and Yin (2021) found that PRW outflow induces a barrier effect on the wind-induced coastal upwelling of nutrients. Tidal fluctuations intensify the pulsatile advance of the EP, leading to the overlying of the new plume water onto the older water layer (). In summer, the coastal current and cyclonic eddies, driven by the southwest monsoon, can extend water residence time, facilitate organic matter accumulation, and sustain hypoxia in the PRE (). Intense wind events (e.g., typhoons) can induce transient reaeration via the disruption of water column stratification, yet hypoxia can be rapidly re-established in the post-wind period (Zhao et al., 2021).
A considerable body of research has investigated the driving factors of hypoxia in the PRE; however, most studies have focused on seasonal and monthly scales, neglecting the impacts of interactions among tides, front convergence, and biological activities at the diurnal scale. This has resulted in an insufficient elucidation of the instantaneous mechanisms underlying hypoxia formation and variation. Furthermore, diel fluctuations in DO may lead traditional observations to miss hypoxia events, thereby misestimating the spatial extent and duration of hypoxia. The objective of this study is to investigate the effects of the front convergence on the hypoxia formation in coastal bottom waters using data from a 2016 summer cruise along the PRE and a 24-h time series at an anchored station.
2 Materials and methods
2.1 Cruise and sampling
We conducted a cruise along the PRE on R/V Haidiao VI during August 27 and September 7 in summer 2016 to investigate the spatial distribution of DO and other fundamental physical parameters in the PRE. Measurements of temperature, salinity, density, chlorophyll fluorescence, and DO were taken at 13 stations along one transect spanning the coordinates of 21.8°–23.1° N, 113.2°–114.5° E (Figure 1). Vertical profile sampling was conducted at all of the 13 stations and at the station ED; the sampling was conducted every 3 h for 19 h (Figure 1). Sampling was conducted along the Z1–Z19 transect during August 27 (8:20–23:50) and during September 7 (6:20–15:20). ED01 was observed on August 28 and DZ3 was observed at 2:20 on September 7. Because the CTD used was not equipped with the probes of DO and fluorescence during August 27, there were no DO data along Z1–Z19. The depth contours of temperature, salinity, and density from the first observation of the Z1–Z19 transect during August 27 are provided in the Supplementary Materials (Supplementary Figure 1). The data during September 7 were used and presented instead for Stns Z1–Z19 in the manuscript as the T–S diagram showed similar vertical mixing dynamics (Figures 2b, c).
Figure 1
Figure 2
A Sea-Bird SBE 9 plus CTD equipped with sensors recorded the temperature, salinity, DO, chlorophyll fluorescence, pH, turbidity, and depth at approximately 0.1-s intervals. Seawater sample was collected and gently stirred to achieve DO saturation via air–water gas exchange, and a subsample was then analyzed using the Winkler titration (Winkler, 1888), while another subsample was measured with CTD. The DO sensor of CTD was calibrated using the result of the Winkler titration (initial accuracy: ± 2% of saturation, Sea-Bird Scientific, 2012). CTD’s chlorophyll fluorescence data were calibrated using spectrophotometric measurements of chlorophyll a, which was collected by filtering seawater samples through GF/F membranes and extracted according to standard solvent extraction procedures (Ritchie, 2006).
2.2 The three-end-member model and water masses
In estuarine waters, temperature and salinity often exhibit a triangular distribution in a T-S diagram, reflecting the mixing of three characteristic water masses. This pattern can be analyzed using a three-end-member mixing model, a well-established method in estuarine studies (Yin et al., 1995; Wang and Yin, 2021; Xiong et al., 2024). In this study, we established a three-end-member model using time series temperature and salinity data from station ED to calculate the three typical water masses, which were defined as Pearl River estuarine water (PRW) with the highest temperature and lowest salinity and bottom water (BW) with the lowest temperature and highest salinity, with offshore surface water (OSW) in the middle. The temperature and the salinity of the data points nearest to the three vertices of the triangle were utilized as representative values for the corresponding typical water masses, respectively. Assuming conservative mixing, the proportional contributions of PRW, BW, and OSW to any point within the triangle can be calculated using the following system of equations:
where P1, P2, and P3 represent the mixing proportion of PRW, BW, and OSW in this point, respectively.
For the time series sampling at Stn ED, we constructed a T–S diagram using vertical profiles of temperature and salinity, which revealed a clear triangular distribution (Figure 2a). The characteristic salinity and temperature (S, T) of the three typical water masses are as follows: PRW (S1, T1) = (28.25, 29.26), BW (S2, T2) = (34.53, 22.83), and OSW (S3, T3) = (33.71, 29.05), respectively. The T–S diagrams at other stations landward of Stn ED along the estuary mostly showed one line or one water mass, indicating two-end-member mixing between the surface and the bottom and the absence of a middle layer (Figures 2b, c).
2.3 Calculation of equivalent thickness and DO integration
Assuming that the water at each depth in the water column was conservatively mixed by three typical water masses in varying proportions, the proportional contributions (P1, P2, P3) of these water masses can be calculated using the three-end-member mixing model. By vertically integrating these proportions throughout the entire water column, we obtain the equivalent thickness (ET) for each water mass, representing the hypothetical thickness that each water mass would occupy if compressed into a single layer.
Given the intense biological activity within the water column at station ED, its in situ DO cannot be treated as the initial DO under the conservative mixing assumption. Following the principles employed by and in their comparisons of eastern boundary upwelling ecosystems, end-members were defined based on the source water properties or preformed values of water masses. We adopted the DO characteristic values of three source water masses that are spatially most proximate to the three end-member water masses at station ED as surrogates, assuming that these values represent the initial characteristic DO of the three water masses at station ED prior to the onset of biological processes. It should be noted that during the actual mixing process, the DO of these three surrogate water masses will also undergo changes as they mix toward station ED. Thus, the calculation presented herein is an approximation. In addition, the selection of DO surrogate values exhibits low sensitivity and high robustness—for example, the DO surrogate value of the OSW endmember adopted in this study is 6.62 mg/L. When alternative DO surrogate values are employed—such as those with a ±5% fluctuation—the resulting deviation in the theoretical DO value is less than 4.63%.
Similar to the three-end-member model, a measured DO has a deviation from a theoretical DO value as a result of biological production or consumption. This theoretical DO value is a result of the conservative mixing of the three water types. The three DO values matching the three water types are as follows: we identified the surface DO at Stn Z19 as DOPRW (5.59 mg/L, 83.96%), the bottom DO at Stn DZ3 as DOBW (3.01 mg/L, 44.12%), and the middle water DO at Stn DZ3 as DOOSW (6.62 mg/L, 101.58%). Based on the conservative mixing of the three DO sources, we calculated the theoretical DO at any depth, which was determined as: DOtheoretical = P1×DOPRW + P2×DOBW + P3×DOOSW. The deviation between measured DO and theoretical DO represents biological DO production or consumption. It should be noted that the DO calculated using this method assumes that the water conditions did not change dramatically during the investigation period or the changes are not large enough to affect the estimate. This approach was employed to qualitatively assess biological production or consumption of DO, as actual mixing processes involve other non-conservative behavior of DO, primarily due to air–sea exchange. The reason why we used in situ DO as the characteristic DO value of typical water mass is that, during the research process, the DO of the two typical water masses, PRW and OSW, was observed to be not saturated and slightly supersaturated, respectively, under their corresponding temperature and salinity. This implies that employing the in situ DO as the characteristic DO value for the typical surface water mass might be more suitable than utilizing the saturation concentration.
2.4 Calculation of hypoxic layer thickness and hypoxic layer average DO0
Hypoxia has been defined using various DO thresholds including 2 mg/L (Rabalais et al., 2001b) and 3 mg/L (Ritter and Montagna, 1999; Vaquer-Sunyer and Duarte, 2008). In this study, we adopted 3 mg/L as the hypoxia threshold. At Stn ED, we quantified bottom water hypoxia using two key metrics: hypoxic layer thickness—the vertical extent of the water column where bottom DO concentrations fell below 3 mg/L and hypoxic layer average DO—calculated by vertically integrating DO concentrations across the hypoxic layer and dividing by its thickness. These metrics enabled us to assess the severity of bottom water hypoxia and to examine correlations between hypoxia intensity and other environmental factors.
2.4.1 Sources of tidal data
The data set of tidal water level was sourced from Hong Kong Observatory1 (https://www.hko.gov.hk/), and tidal water levels at Quarry Bay (22.29, 114.21) were used as it is near Stn ED.
2.4.2 Data plot
For the plotting of the depth–distance contours and depth–time contours, Ocean Data View2 (Schlitzer, R. 2023. Ocean Data View, Version 5.6.2) was employed. The interpolation was conducted using the weighted-average gridding method available in the ODV.
Scatter plots and bar charts were created using SigmaPlot3 (Systat Software Inc. 2006. SigmaPlot, Version 10.0).
2.4.3 Correlation analysis
We used the R package CORRPLOT (v.4.2.2; R Foundation for Statistical Computing, Vienna, Austria) to establish the correlation matrix and carried out bivariate correlations for each of the two parameters. The correlation coefficient that we used was Pearson coefficient, and the significance test was the bilateral test.
3 Results
3.1 Distribution of hydrological and biogeochemical parameters along the PRE
The depth–distance contours of temperature (T) showed that surface T varied little (27.84 °C–29.85 °C) along the estuary, with a slight decrease with depth in the upper-80-km section, while at ED01 and 80 km downstream a dramatic thermocline occurred (surface ~30 °C to bottom ~22 °C), indicating strong thermal stratification (Figure 3a). Surface salinity increased from 2.11 upstream to 34.7 downstream (Figure 3b), with sharp frontal gradients at approximately 50 km (riverine front) and 90 km near Stn ED (estuarine front), where the vertical salinity gradient was stronger in the riverine frontal section. Density patterns reflected salinity variations, with ΔDensity showing significant gradients at ~6.5- and 17-m depths at Stn ED (Figure 3d). Chlorophyll fluorescence increased from near-zero upstream to 11.6 μg/L at Stn ED, indicating phytoplankton blooms, with elevated bottom values at adjacent Stn Z19 and DZ3 (Figure 3e). DO was low (<4 mg/L) at Z1–Z5, was 4 to 5 mg/L at Z7–Z17, and increased downstream of Z17, reaching the maximum of 11.4 mg/L at ED (Figure 3f). At ED, DO was reduced to <2 mg/L in the bottom (Figure 3f).
Figure 3
3.2 Diurnal variations in hydrological/biological factors and water mixing at Stn ED
During the 24-h anchored observation at Stn ED, surface water temperature showed minimal diurnal variation (28.99 °C–29.85 °C) while exhibiting a strong vertical gradient (6.38 °C ± 0.25 °C diurnal variation, Figure 4a). A distinct thermocline fluctuated near the bottom layer (~20 ± 3-m depth), likely tidally influenced, with sub-thermocline temperatures consistently below 27 °C (minimum 22.82 °C), indicating that shelf water forms a halocline with the overlying layer due to the pronounced thermal gradient. Surface salinity was the lowest (28.25, Figure 4b) and confirmed the PRW influence, with significant vertical variation (5.57 ± 0.72 diurnally) reaching 34.65 at the bottom. The upper water column (<1 diurnal salinity variation above the surface halocline) contrasted sharply with the intermediate layers (31–34 salinity), suggesting convergence where offshore surface water was subducted beneath the plume, creating a strong halocline. The ΔDensity contour (Figure 4f) revealed this surface halocline and bottom thermocline as dual pycnoclines dividing the water column into three distinct layers.
Figure 4
Chlorophyll maxima (11.49 μg/L peak, Figure 4d) occurred in surface waters above the pycnocline (8.88 ± 2.57 mg/m³), indicating phytoplankton blooms, while sub-surface concentrations were low at <1 μg/L below 10 m). DO exhibited strong vertical contrast: >9 mg/L (9.47 ± 1.06 mg/L) above the upper pycnocline versus hypoxic conditions (mostly <3 mg/L) below the deeper pycnocline (Figure 4e). The high chlorophyll level coincided with low salinity and high DO temporally and spatially (Figures 4b, d, e), suggesting PRW nutrient stimulation of phytoplankton growth, whereas bottom hypoxia with minimal chlorophyll implied microbial respiration of organic matter as a key hypoxia driver.
Three water masses were identified: Pearl River estuarine water (PRW, ≥40% in the upper 7 m, equivalent thickness 6.26–9.48 m, Figures 5a, d), offshore surface water (OSW, ≥40% at 7–18 m, Figure 5b), and bottom water (BW, ≥40% below 18 m, Figure 5c). PRW dominance was correlated with low salinity and high DO and chlorophyll (Figures 5a, 4b, d, e), while BW≥80% coincided with DO ≤3 mg/L (Figure 5c, 4e). Water mass equivalent thickness varied diurnally (PRW: 6.26–9.48 m; OSW: 5.84–12.83 m; BW: 5.79–9.68m, Figure 5d) following tidal cycles (0.86–2.4-m range, Figure 5f): flood tide (01:02–08:00) decreased PRW/BW but increased OSW; ebb tide (08:00–16:00) boosted PRW/BW while reducing OSW; a subsequent flood (16:00–20:15) reversed these trends. The depth-integrated DO varied (142.96–165.62 g/m², Figure 5e) and exhibited an analogous tidal cycle.
Figure 5
3.3 Dynamics of DO under conserved and non-conserved mixing
Based on the T–S diagrams, three end-members (S, T) were identified. Similarly, three end-members of DO were determined. This allowed for the establishment of the DO–S conservative (theoretic) mixing diagrams (Figure 6b). When a measured DO has a deviation from the conservative value, it indicates biological production or consumption of DO—for example, the measured DO values indicated net DO production by phytoplankton when they exceeded the conservative DO at salinities below 32.2 and net DO consumption at salinity >32.5 (Figure 6b). In a vertical profile, the measured DO was higher than the theoretical DO in the upper layer but was generally lower in the middle and deeper layers (Figure 6a). This pattern remained consistent in all of the vertical profiles throughout the 24-h observation period at Stn ED.
Figure 6
3.4 Quantification of hypoxia intensity at the ED station and analysis of its influencing factors
During the 24-h observation at station ED, the hypoxic layer thickness ranged from 1.77 to 6.67 m, with hypoxic layer average DO concentrations of 1.53–2.31 mg/L (Table 1). The correlation analysis revealed that tidal level negatively correlated with PRW ET (Figures 7, 8a, p < 0.01) and BW ET (Figures 7, 8c, p < 0.05) but positively correlated with OSW ET (Figures 7, 8b, p < 0.01). The depth-integrated DO showed positive correlations with tidal level (Figures 7, 8d, p < 0.01) and OSW thickness (Figure 7) but a negative correlation with BW thickness (Figure 7). The BW and OSW thicknesses were inversely correlated (Figures 7, 8e, p < 0.01), while the hypoxic layer average DO was negatively correlated with PRW thickness (Figures 7, 8f, p < 0.05).
Table 1
| Sampling time | Hypoxic layer thickness (m) | Hypoxic layer average DO (mg/L) |
|---|---|---|
| 1:02 | 6.04 | 1.96 |
| 4:00 | 2.19 | 2.10 |
| 7:00 | 1.77 | 1.77 |
| 10:00 | 3.01 | 2.31 |
| 13:18 | 3.20 | 1.74 |
| 16:00 | 6.30 | 1.53 |
| 17:51 | 6.67 | 1.94 |
| 20:15 | 6.60 | 2.05 |
Hypoxic layer thickness and hypoxic layer average DO at different sampling times.
The layer with DO concentration <3 mg/L was defined as the hypoxic layer.
Figure 7
Figure 8
4 Discussion
4.1 Role of the front-convergence-induced OSW barrier in hypoxia formation
The stability of the water column is critical in driving the formation of hypoxia under an abundant supply of organic matter from surface phytoplankton blooms to the bottom layer. When the Pearl River freshwater outflows, it forms the EP which covers a large area of the adjacent South China Sea in summer (Yin et al., 2001, 2002). There is a seasonal progression of the EP and front in the region (Yin et al., 2001). From spring to summer, increased river outflow drives the EP seaward, where it entrains underlying seawater, as seen in the salinity depth–distance structure (Figure 3b; Supplementary Figure 1b). The estuarine plume front advances offshore and converges with more oceanic waters, termed offshore surface water (OSW) (Wang and Yin, 2021; ). During this study’s cruise, the plume front was located near Stn ED (Figure 3b; Supplementary Figure 1b). Tidal fluctuations in the estuary further drive pulsed seaward movements of the plume, causing newer plume waters to override older ones. This results in a stratified water column with three distinct layers: PRW in the surface layer, OSW in the mid-layer, and bottom water (BW) below (Figures 5a–c).
The three water masses form the dynamic structure of the water column. The OSW and the BW formed a pronounced thermocline and pycnocline near the 25 °C isotherm at their interface (Figures 3a, 4a, 5b, c). The halocline between PRW and OSW further reinforced stratification, creating sharp density gradients between layers. The DO variations reflected this layering: PRW (>9 mg/L) above the upper pycnocline, OSW (5–7 mg/L) between the two pycnoclines, and BW (<4 mg/L, often ≤3 mg/L) below the deeper pycnocline (Figures 3f, 4e, 5a–c). This vertical structure demonstrates that OSW acts as a barrier, inhibiting vertical DO exchange and preventing surface oxygen from replenishing BW, thereby promoting hypoxia. In contrast, other stations along the transect lacked this three-layer structure, exhibiting two-end-member mixing in T–S diagrams (Figures 2b, c) and no hypoxia occurrences (Figures 2d, 3f). This clearly underscores the critical role of OSW as a barrier in hypoxia formation. The middle OSW layer strengthened the stratification and prevents vertical mixing and re-oxygenation to the bottom water, eventually resulting in the DO consumption to hypoxia (Figures 3f, 4e) under bloom-derived organic detritus to the seabed (; Rabalais and Turner, 2001a).
Hypoxia in the bottom layer is primarily driven by the biological consumption of O2, which is fueled by algal organic matter sinking from the surface layer (). However, the consumption of O2 to the hypoxic level requires a period of stable time or residence time of the bottom water mass layer (). Basically, it is the balance between the rate of O2 consumption and the rate of aeration of the bottom layer that determines the formation of the hypoxia (). Aeration of the bottom layer is predominantly regulated by the dilution of lateral advection and vertical mixing ()—for instance, advection of water masses with high DO concentrations into the bottom layer inhibits hypoxia development (Zhang and Li, 2010). Persistent wind directions can cause low−oxygen water to be advected onto shoals, where mixing and aeration occur (Scully, 2010). Enhanced vertical mixing weakens stratification and provides vertical mixing of higher O2, which prevents hypoxia occurrences (). Accordingly, this mechanism is the most likely explanation for the lack of hypoxia at the upstream stations of the Pearl River Estuary in this study (Figures 2b, c). During the formation of bottom hypoxia, residence time plays a critical role in stabilizing the water column when the supply of organic matter is sufficient. In this context, the OSW may act as a barrier that limits vertical aeration, favoring conditions conducive to DO depletion in the bottom layer following phytoplankton-bloom-derived organic matter inputs. This interpretation is consistent with previous studies conducted during summer in the Pearl River estuarine waters (Su et al., 2017; Wang et al., 2017; ). While this interpretation is supported by the observed correlations, alternative processes such as lateral advection of oxygenated waters and variability in mixing intensity may also contribute. In this framework, the water-column-integrated DO shows a significant positive correlation with OSW equivalent thickness, which is, in turn, negatively correlated with BW equivalent thickness.
Estuarine plume dynamics are widely reported to play a key role in the development of hypoxia in global estuarine systems. The Chesapeake Bay represents a classic example, where frontal convergence zones exert critical controls on hypoxia formation (). The same fundamental mechanism operates in the Gulf of Mexico, where the pycnocline and stratification that act as a vertical mixing barrier are direct products of plume dynamics (Rabalais et al., 2002; ). In the Changjiang estuarine system, large-scale coastal hypoxia was first documented by and has since become a well-recognized seasonal feature in numerous subsequent studies (; Zhu et al., 2011; Wang et al., 2012; ). In particular, Zhu et al. (2017) identified the frontal zone as a critical barrier to vertical mixing and downward oxygen supply. The Changjiang Diluted Water (CDW) forms an extensive low-salinity surface plume that extends offshore, and its seaward edge forms a sharp estuarine front against the warmer, more saline Taiwan Warm Current (TWC) water (Zhu et al., 2017). The denser oceanic (TWC) water subducts beneath the lighter CDW at this front, creating a three-layered system: an upper layer composed of nutrient-rich, fresh CDW that frequently supports intense phytoplankton blooms and eutrophication; a middle layer consisting of intruding saline TWC water, which is hydraulically isolated from both surface and bottom layers; and a bottom layer characterized by stagnant, low-oxygen water that accumulates organic matter settling from the productive upper layer, leading to the development and persistence of seasonal hypoxia. This three-layer configuration generates a barrier layer that favors the formation of bottom hypoxia. Although the Pearl River freshwater discharge is considerably smaller than that of the Changjiang, leading to a more spatially constrained frontal system, the underlying dynamical processes are highly similar (; Wang and Yin, 2021). Our study provides new observational evidence for the tidal regulation mechanism of bottom hypoxia coupled with estuarine frontal dynamics.
The findings of the OSW as the barrier agree with observations by , who identified a mid-depth barrier layer limiting the DO exchange between river plumes and shelf salt wedges. However, their study primarily utilized the regional ocean modeling system (ROMS) to estimate the DO budget, while a three-end-member mixing model was specifically adopted in this paper, considering the special three-layer water structure in the hypoxic area.
The distinct hydrographic features observed along the transect indicate that, in contrast to the river water–seawater binary mixing regime in the mid-estuarine region (e.g., stations upstream of ED, Figures 3a–f; Supplementary Figures 1a–c), the development of the three-layer water structure at the estuarine outer margin—driven by the presence of bottom water with low temperature and low basal value of DO and affected by the barrier effect of OSW—may be more conducive to hypoxia formation, as exemplified by the station ED in this study (Figures 3a, d, f, 5b; Supplementary Figure 1a). The comparative investigation conducted by Wei et al. (2015) and Tian et al. (2022) on the binary mixing regime in the middle reach of the Changjiang Estuary (122°00′ E–122°30′ E) and the three-layer water column structure in its outer reach (east of 122°30′ E) further corroborates this perspective. Moreover, this gives rise to another concern. If climate activities, such as global warming, lead to an increase in precipitation and a wider expansion of the river plume, the water column, which was originally composed of OSW and low temperature and low DO base values BW (e.g., the ED01 station in this study; Figures 3a, f; Supplementary Figure 1a), might form a three-layer water structure under the influence of EP and thereby experience hypoxia.
Tidal dynamics further modulated hypoxia at Stn ED. The T–S diagram revealed time-varying mixing patterns, with some profiles resembling two-end-member mixing (arc-shaped) and others reflecting three-end-member mixing (Figure 2a) (Wang and Yin, 2021). The analysis of water mass fractions (ET) and depth-integrated DO during the 24-h anchor station showed that OSW ET followed tidal trends, while PRW and BW ET exhibited inverse relationships (Figures 5d–f). The correlation analysis (Figure 7) confirmed significant tidal linkages: negative with PRW ET (Figures 7, 8a, p < 0.01), positive with OSW ET (Figures 7, 8b, p < 0.01), and negative with BW ET (Figures 7, 8c, p < 0.05), suggesting that tidal level was closely linked to the diurnal variations in water mass mixing. Water-column-depth-integrated DO correlated positively with tides (Figures 7, 8d, p < 0.01) and OSW ET (Figure 7, p < 0.05) but negatively with BW ET (Figure 7, p < 0.05), suggesting that tidal variability was associated with diurnal changes of the overall DO level within the water column through modulating water mass mixing. OSW ET correlated negatively with PRW ET (Figures 7, 8e, p < 0.01) and BW ET (Figures 7, 8e, p < 0.01), while hypoxic layer average DO (Table 1) correlated negatively with PRW ET (Figures 7, 8f, p < 0.05), suggesting that tidal variability may be indirectly linked to bottom hypoxia through modulating the barrier effect of OSW and the supply of sinking organic matter associated with PRW. Accordingly, tidal forcing contributes to diurnal variations in stratification and oxygen dynamics and thus represents an important contributing factor to summer hypoxia at Stn ED.
Tidal flows during floods cause the bottom water to flow from offshore toward the nearshore, which often leads to higher turbulent flow and hence vertical mixing, whereas tidal ebbs push the estuarine plume seaward at the surface while the bottom water retreats. The tidal floods usually dilute hypoxia, while tidal ebbs enhance the formation of hypoxia.
In general, there are coupled physical–biogeochemical effects of tidal asymmetry. The foundational physics is best explained by and Simpson et al. (1990). gave a comprehensive review that covers estuarine circulation theory, including tidal straining and asymmetries, providing context for how physical processes control properties like oxygen. The direct link to hypoxia formation is reviewed by and demonstrated in modeling and observational studies by Wang et al. (2017).
4.2 Organic matter supply fueling bottom hypoxia: phytoplankton blooms in the PRW
The OSW in the middle water column serves as a barrier layer that causes vertical mixing to slow down to the degree to which DO consumption driven by organic matter supply is fast enough to deplete DO to hypoxic levels. The observed bottom hypoxia at station ED is strongly driven by phytoplankton blooms in the PRW, as evidenced by elevated surface chlorophyll-a fluorescence and very high DO >10 mg/L near the station (Figures 3e, f). The comparison between theoretical DO calculated with three-end-member model and in situ-measured DO further confirms that, at station ED, substantial DO was released in the surface layer via photosynthesis, whereas DO in the bottom layer was depleted to hypoxic levels through respiratory processes (Figures 6a, b). The plot of in situ-measured DO and the theoretical DO against salinity (Figure 6) shows that the in situ DO in the middle layer of OSW has little deviations from the theoretical DO. However, in situ DO increases at salinity <32.2, giving higher deviations from the theoretical DO at Stn ED, indicating high photosynthetic DO production above the middle layer, whereas in situ DO at salinity >32.5 decreases to lower values than the theoretical DO, indicating that DO in the bottom layer was consumed to hypoxic levels through respiratory processes (Figures 6a, b). During biological production and consumption of DO, the middle layer plays the role of a barrier to stabilize the water column and prevent vertical mixing so that the bottom layer DO could be consumed without vertical aeration. The PRE, enriched by riverine nutrient inputs, sustains recurrent phytoplankton blooms, consistent with findings from Yin et al. (2004b); , and other recent studies (Qiu et al., 2010; ).
Phytoplankton distribution in the estuary is regulated by estuarine front dynamics: high turbidity and strong currents near the river mouth inhibit algal accumulation (), explaining the absence of hypoxia upstream of Stn ED (Figures 3e, f), while reduced suspended loads and increased light penetration offshore—enhanced by front-induced convergence—promote blooms that propagate seaward with river discharge (Figures 3b, e; Supplementary Figure 1b; ; ; ). The resulting high phytoplankton biomass generates substantial organic matter flux to bottom waters, while sinking particulate organic carbon (POC) fuels microbial respiration that depletes bottom oxygen (Rabalais et al., 2002; Rabouille et al., 2008) as observed at Stn ED (Figures 6a, b).
5 Conclusions
This study elucidates the mechanism of bottom hypoxia formation driven by offshore surface water (OSW) subduction at the estuarine frontal convergence. The OSW acts as a hydrological barrier that inhibits vertical mixing and stabilizes the bottom layer, creating conditions conducive to hypoxia through oxygen consumption during decomposition of organic matter derived from surface phytoplankton blooms under the EP. These findings align with global patterns in typical hypoxic estuary where eutrophication and stratification synergistically drive coastal hypoxia (). However, the three-layer water structure in our study appears to be more conducive to the formation of hypoxia. For future research, two key priorities emerge: (1) quantitative assessment of the relative contributions of sinking particulate organic carbon versus resuspended sediments to oxygen depletion across tidal and seasonal cycles and (2) evaluation of climate change impacts, particularly whether increased precipitation will expand the hypoxic area or how increased rainfall may intensify stratification and exacerbate hypoxic conditions. Such investigations will provide critical insights for effective coastal zone management.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
JL: Formal analysis, Writing – original draft, Visualization, Data curation. LH: Project administration, Methodology, Writing – review & editing, Investigation. CX: Validation, Conceptualization, Writing – review & editing. ZL: Supervision, Writing – review & editing. KY: Resources, Writing – review & editing, Conceptualization, Supervision.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This study was part of the Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai) projects (SML2021SP204, SML2024SP022, 40010102, and SML2024SP004), Guangdong-NSF China Joint Scheme Key Project (U1701247). CX acknowledges the support by the National Natural Science Foundation of China (42306221). HL acknowledges the support by the National Natural Science Foundation of China (41406133), the Natural Science Foundation of Guangdong Province (2015A030313071), and the Sun Yat-sen University (15lgpy08). The cruise was supported by the South China Sea Multi-disciplinary Investigation cruise of Sun Yat-sen University. This study contributes to the global Ocean Negative Carbon Emissions (ONCE) Program.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmars.2026.1768899/full#supplementary-material
References
1
BianchiT. S. (2007). Biogeochemistry of Estuaries (New York: Oxford University Press).
2
BianchiT. S.DiMarcoS. F.CowanJ. H.HetlandR. D.ChapmanP.DayJ. W.et al. (2010). The science of hypoxia in the Northern Gulf of Mexico: a review. Sci. Total. Environ.408, 1471–1484. doi: 10.1016/j.scitotenv.2009.11.047
3
BreitburgD.LevinL. A.OschliesA.GrégoireM.ChavezF. P.ConleyD. J.et al. (2018). Declining oxygen in the global ocean and coastal waters. Science359, eaam7240. doi: 10.1126/science.aam7240
4
BroeckerW. S.PengT. H. (1982). Tracers in the Sea Vol. 690 (Palisades, New York: Lamont-Doherty Geological Observatory, Columbia University).
5
ChavezF. P.MessiéM. (2009). A comparison of eastern boundary upwelling ecosystems. Prog. Oceanogr.83, 80–96. doi: 10.1016/j.pocean.2009.07.032
6
ChenC. C.GongG. C.ShiahF. K. (2007). Hypoxia in the East China Sea: one of the largest coastal low-oxygen areas in the world. Mar. Environ. Res.64, 399–408. doi: 10.1016/j.marenvres.2007.01.007
7
CloernJ. E. (1987). Turbidity as a control on phytoplankton biomass and productivity in estuaries. Cont. Shelf Res.7, 1367–1381. doi: 10.1016/0278-4343(87)90042-2
8
CuiY.WuJ.RenJ.XuJ. (2019). Physical dynamics structures and oxygen budget of summer hypoxia in the Pearl River Estuary. Limnol. Oceanogr.64, 131–148. doi: 10.1002/lno.11025
9
CuiY.WuJ.TanE.KaoS. J. (2022). Role of particle resuspension in maintaining hypoxic level in the Pearl River Estuary. J. Geophys. Res.: Oceans.127, e2021JC018166. doi: 10.1029/2021jc018166
10
DiazR. J.RosenbergR. (2008). Spreading dead zones and consequences for marine ecosystems. Science321, 926–929. doi: 10.1126/science.1156401
11
FennelK.HetlandR.FengY.DiMarcoS. (2011). A coupled physical-biological model of the Northern Gulf of Mexico shelf: model description, validation and analysis of phytoplankton variability. Biogeosciences8, 1881–1899. doi: 10.5194/bg-8-1881-2011
12
GeyerW. R.SmithJ. D. (1987). Shear instability in a highly stratified estuary. J. Phys. Oceanogr.17, 1668–1679. doi: 10.1175/1520-0485(1987)017<1668:siiahs>2.0.co;2
13
GuoX.SuJ.GuoL.LiuZ.YangW.LiY.et al. (2023). Coupling of carbon and oxygen in the Pearl River plume in summer: Upwelling, hypoxia, reoxygenation and enhanced acidification. J. Geophys. Res.: Oceans.128, e2022JC019326. doi: 10.1029/2022jc019326
14
HarrisonP. J.YinK.LeeJ. H. W.GanJ.LiuH. (2008). Physical–biological coupling in the pearl river estuary. Cont. Shelf Res.28, 1405–1415. doi: 10.1016/j.csr.2007.02.011
15
LaiH.BiS.YiH.LiH.WeiX.WangG.et al. (2024). Characteristics of demersal fish community structure during summer hypoxia in the Pearl River Estuary, China. Ecol. Evol.14, e11722. doi: 10.1002/ece3.11722
16
LiD.GanJ.HuiR.LiuZ.YuL.LuZ.et al. (2020). Vortex and biogeochemical dynamics for the hypoxia formation within the coastal transition zone off the Pearl River Estuary. J. Geophys. Res.: Oceans.125, e2020JC016178. doi: 10.1029/2020jc016178
17
LiD. J.ZhangJ.HuangD. J.WuY.LiangJ. (2002). Oxygen depletion off the Changjiang (Yangtze River) estuary. Sci. China Ser. D. Earth Sci.45, 1137–1146. doi: 10.1360/02yd9110
18
LiM.ZhongL. (2009). Flood–ebb and spring–neap variations of mixing, stratification and circulation in Chesapeake Bay. Continental Shelf Research29(1):4–14.
19
LuZ.GanJ. (2015). Controls of seasonal variability of phytoplankton blooms in the Pearl River Estuary. Deep. Sea. Res. Part. II.: Topical. Stud. Oceanogr.117, 86–96. doi: 10.1016/j.dsr2.2013.12.011
20
LuZ.GanJ.DaiM.LiuH.ZhaoX. (2018). Joint effects of extrinsic biophysical fluxes and intrinsic hydrodynamics on the formation of hypoxia west off the Pearl River Estuary. J. Geophys. Res.: Oceans.123, 6241–6259. doi: 10.1029/2018jc014199
21
MaX.LiuA.ZhaoQ.WangB.TianD.MengQ.et al. (2022). Temporal variation of summer hypoxia off Changjiang Estuary during 1997–2014 and its association with ENSO. Front. Mar. Sci.9. doi: 10.3389/fmars.2022.897063
22
MacCreadyP.GeyerW. R. (2010). Advances in estuarine physics. Annu. Rev. Mar. Sci.2, 35–58. doi: 10.1146/annurev-marine-120308-081015
23
MiddelburgJ. J.LevinL. A. (2009). Coastal hypoxia and sediment biogeochemistry. Biogeosciences6, 1273–1293. doi: 10.5194/bg-6-1273-2009
24
O’BoyleS.McDermottG.SilkeJ.CusackC. (2016). Potential impact of an exceptional bloom of Karenia mikimotoi on dissolved oxygen levels in waters off western Ireland. Harmful. Algae.53, 77–85. doi: 10.34074/thes.6741
25
PanJ.GuY.WangD. (2014). Observations and numerical modeling of the Pearl River plume in summer season. J. Geophys. Res.: Oceans.119, 2480–2500. doi: 10.1002/2013jc009042
26
PenaM. A.KatsevS.OguzT.GilbertD. (2010). Modeling dissolved oxygen dynamics and hypoxia. Biogeosciences7, 933–957. doi: 10.5194/bg-7-933-2010
27
QiuD.HuangL.ZhangJ.LinS. (2010). Phytoplankton dynamics in and near the highly eutrophic Pearl River Estuary, South China Sea. Cont. Shelf Res.30, 177–186. doi: 10.1016/j.csr.2009.10.015
28
RabalaisN. N.TurnerR. E. (2001a). Hypoxia in the northern Gulf of Mexico: description, causes and change. Coast. Hypoxia.: Consequences. For. Living. Resour. Ecosyst.58, 1–36. doi: 10.1029/ce058p0001
29
RabalaisN. N.TurnerR. E.WisemanW. J. (2001b). Hypoxia in the gulf of Mexico. J. Environ. Qual.30, 320–329. doi: 10.2134/jeq2001.302320x
30
RabalaisN. N.TurnerR. E.WisemanW. J. (2002). Gulf of Mexico hypoxia, aka “The dead zone. Annu. Rev. Ecol. Syst.33, 235–263. doi: 10.1146/annurev.ecolsys.33.010802.150513
31
RabouilleC.ConleyD. J.DaiM. H.CaiW. J.ChenC. T. A.LansardB.et al. (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
32
RitchieR. J. (2006). Consistent sets of spectrophotometric chlorophyll equations for acetone, methanol and ethanol solvents. Photosynth. Res.89, 27–41. doi: 10.1007/s11120-006-9065-9
33
RitterC.MontagnaP. A. (1999). Seasonal hypoxia and models of benthic response in a Texas bay. Estuaries22, 7–20. doi: 10.2307/1352922
34
SchlitzerR. (2023). Ocean Data View, Version 5.6.2. Zenodo. https://doi.org/10.5281/zenodo.7897029
35
ScullyM. E. (2010). Wind modulation of dissolved oxygen in Chesapeake Bay. Estuaries. Coasts.33, 1164–1175. doi: 10.1007/s12237-010-9319-9
36
Sea-Bird Scientific (2012). Application Note 64-2: SBE 43 Dissolved Oxygen Sensor Calibration and Data Corrections ( Sea-Bird Scientific). Available online at: https://calcofi.org/downloads/underway/2020/2007SR/data/ctd/docs/APPnotes/Appnote64-2Jun12.pdf.
37
SimpsonJ. H.BrownJ.MatthewsJ.AllenG. (1990). Tidal straining, density currents, and stirring in the control of estuarine stratification. Estuaries13, 125–132. doi: 10.2307/1351581
38
SimpsonJ. H.SharplesJ. (2012). “ Chapter 8: “Fronts and frontogenesis”,” in Introduction to the Physical and Biological Oceanography of Shelf Seas (Solomons, MD, USA: Cambridge University Press), ISBN: ISBN: 978-0521877626. (Hardback) / 978-0521701488 (Paperback).
39
Systat Software Inc (2006). SigmaPlot (Version 10.0) [Computer software]. Systat Software Inc.
40
SuJ.DaiM.HeB.WangL.GanJ.GuoX.et al. (2017). Tracing the origin of the oxygen-consuming organic matter in the hypoxic zone in a large eutrophic estuary: the lower reach of the Pearl River Estuary, China. Biogeosciences14, 4085–4099. doi: 10.5194/bg-14-4085-2017
41
TianD.ZhouF.ZhangW.ZhangH.MaX.GuoX. (2022). Effects of dissolved oxygen and nutrients from the Kuroshio on hypoxia off the Changjiang River estuary. J. Oceanol. Limnol.40, 515–529. doi: 10.1007/s00343-021-0440-3
42
Vaquer-SunyerR.DuarteC. M. (2008). Thresholds of hypoxia for marine biodiversity. Proc. Natl. Acad. Sci.105, 15452–15457. doi: 10.1073/pnas.0803833105
43
WangB.HuJ.LiS.LiuD. (2017). A numerical analysis of biogeochemical controls with physical modulation on hypoxia during summer in the Pearl River estuary. Biogeosciences14, 2979–2999. doi: 10.5194/bg-14-2979-2017
44
WangB.WeiQ.ChenJ.XieL. (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
45
WangK.YinK. (2021). Barrier effect of the Pearl River estuarine plume on wind‐induced coastal upwelling of nutrients. J. Geophys. Res.: Biogeosci.126, e2020JG006067. doi: 10.1029/2020jg006067
46
WeiQ.WangB.ChenJ.XiaC.QuD.XieL. (2015). Recognition on the forming-vanishing process and underlying mechanisms of the hypoxia off the Yangtze River estuary. Sci. China Earth Sci.58, 628–648. doi: 10.1007/s11430-014-5007-0
47
WinklerL. W. (1888). Die Bestimmung des im Wasser gelösten Sauerstoffes. Ber. Dtsch. Chem. Ges.21, 2843–2854. doi: 10.1002/cber.188802102122
48
XiongZ.ZhangY.LiangJ.ChenZ.HeL.YinK. (2024). A new method of estimating carbon sequestration and its efficiency in coastal waters. Prog. Oceanogr.226, 103289. doi: 10.1016/j.pocean.2024.103289
49
YinK. (2002). Monsoonal influence on seasonal variations in nutrients and phytoplankton biomass in coastal waters of Hong Kong in the vicinity of the Pearl River estuary. Mar. Ecol. Prog. Ser.245, 111–122. doi: 10.3354/meps245111
50
YinK.HarrisonP. J.PondS.BeamishR. J. (1995). Entrainment of nitrate in the Fraser River estuary and its biological implications. I. Effects of the salt wedge. Estuar. Coast. Shelf Sci.40, 505–528. doi: 10.1006/ecss.1995.0035
51
YinK.LinZ.KeZ. (2004b). Temporal and spatial distribution of dissolved oxygen in the Pearl River Estuary and adjacent coastal waters. Cont. Shelf Res.24, 1935–1948. doi: 10.1016/j.csr.2004.06.017
52
YinK.QianP. Y.WuM. C.ChenJ. C.HuangL.SongX.et al. (2001). Shift from P to N limitation of phytoplankton growth across the Pearl River estuarine plume during summer. Mar. Ecol. Prog. Ser.221, 17–28. doi: 10.3354/meps221017
53
YinK.ZhangJ.QianP. Y.JianW.HuangL.ChenJ.et al. (2004a). Effect of wind events on phytoplankton blooms in the Pearl River estuary during summer. Cont. Shelf Res.24, 1909–1923. doi: 10.1016/j.csr.2004.06.015
54
ZhangH.LiS. (2010). Effects of physical and biochemical processes on the dissolved oxygen budget for the Pearl River Estuary during summer. J. Mar. Syst.79, 65–88. doi: 10.1016/j.jmarsys.2009.07.002
55
ZhaoY.UthaipanK.LuZ.LiY.LiuJ.LiuH.et al. (2021). Destruction and reinstatement of coastal hypoxia in the South China Sea off the Pearl River estuary. Biogeosciences18, 2755–2775. doi: 10.5194/bg-18-2755-2021
56
ZhuZ.WuH.LiuS.WuY.HuangD. (2017). Hypoxia off the Changjiang (Yangtze River) Estuary: Oxygen depletion and organic matter decomposition. Mar. Chem.194, 71–82. doi: 10.1016/j.marchem.2011.03.005
57
ZhuZ. Y.ZhangJ.WuY.ZhangY. Y.LinJ.LiuS. M. (2011). Hypoxia off the Changjiang (Yangtze River) Estuary: Oxygen depletion and organic matter decomposition. Mar. Chem.125, 108–116. doi: 10.1016/j.marchem.2011.03.005
Summary
Keywords
barrier effect, bottom hypoxia, estuarine plume, frontal convergence, phytoplankton blooms, tidal regulation
Citation
Liang J, He L, Xiao C, Lai Z and Yin K (2026) Effects of Pearl River estuarine-front-induced convergence on formation of bottom hypoxia in summer. Front. Mar. Sci. 13:1768899. doi: 10.3389/fmars.2026.1768899
Received
16 December 2025
Revised
28 March 2026
Accepted
30 March 2026
Published
22 April 2026
Volume
13 - 2026
Edited by
Chung-Chi Chen, National Taiwan Normal University, Taiwan
Reviewed by
Borja Aguiar-González, University of Las Palmas de Gran Canaria, Spain
Yanyi Miao, Ministry of Natural Resources, China
Updates
Copyright
© 2026 Liang, He, Xiao, Lai and Yin.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Lei He, helei23@mail.sysu.edu.cn; Canbo Xiao, xiaocb@mail.sysu.edu.cn
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.