Coupling the Chemical Dynamics of Carbonate and Dissolved Inorganic Nitrogen Systems Printer-friendly Version Interactive Discussion Coupling the Chemical Dynamics of Carbonate and Dissolved Inorganic Nitrogen Systems in the Eutrophic and Turbid Inner Changjiang (yangtze River) Estuary Coupling the 

Introduction Conclusions References Tables Figures Back Close Full Screen / Esc Abstract Introduction Conclusions References Tables Figures Back Close Full Screen / Esc Abstract To better understand biogeochemical processes controlling CO 2 dynamics in those eu-trophic large-river estuaries and coastal lagoons, we investigated surface water carbon-ate system, nutrients, and relevant hydrochemical parameters in the inner Changjiang (Yangtze River) Estuary, covering its channel-like South Branch and the lagoon-like 5 North Branch, shortly after a spring-tide period in April 2010. In the North Branch, with a water residence time of more than 2 months, biogeochemical additions of ammonium (7.4 to 65.7 µmol kg −1) and alkalinity (196 to 695 µmol kg −1) were detected along with high salinity of 4.5 to 17.4. In the South Branch upper-reach, unusual salinity values of 0.20 to 0.67 were detected, indicating spillover waters from the North Branch. The 10 spillover waters enhanced the springtime Changjiang export fluxes of nutrients, dissolved inorganic carbon, and alkalinity. And they affected the biogeochemistry in the South Branch, by lowering water-to-air CO 2 flux and continuing the nitrification reaction. In the North Branch, pCO 2 was measured from 930 to 1518 µatm at the salin-ity range between 8 and 16, which was substantially higher than the South Branch 15 pCO 2 of 700 to 1100 µatm. Based on field data analyses and simplified stoichiometric equations, we suggest that the North Branch CO 2 productions were quantified by bio-geochemical processes combining organic matter decomposition, nitrification, CaCO 3 dissolution, and acid-base reactions in the estuarine mixing zone. Although our study is subject to limited temporal and spatial coverage of sampling, we have demonstrated 20 a procedure to quantificationally constrain net CO 2 productions in eutrophic estuaries and/or coastal lagoons, by coupling the chemical dynamics of carbonate and dissolved inorganic nitrogen systems. Abstract Introduction Conclusions References Tables Figures Back Close Full Screen / Esc


Figures Back Close
Full

Abstract
To better understand biogeochemical processes controlling CO 2 dynamics in those eutrophic large-river estuaries and coastal lagoons, we investigated surface water carbonate system, nutrients, and relevant hydrochemical parameters in the inner Changjiang (Yangtze River) Estuary, covering its channel-like South Branch and the lagoon-like North Branch, shortly after a spring-tide period in April 2010.In the North Branch, with a water residence time of more than 2 months, biogeochemical additions of ammonium (7.4 to 65.7 µmol kg −1 ) and alkalinity (196 to 695 µmol kg −1 ) were detected along with high salinity of 4.5 to 17.4.In the South Branch upper-reach, unusual salinity values of 0.20 to 0.67 were detected, indicating spillover waters from the North Branch.The spillover waters enhanced the springtime Changjiang export fluxes of nutrients, dissolved inorganic carbon, and alkalinity.And they affected the biogeochemistry in the South Branch, by lowering water-to-air CO 2 flux and continuing the nitrification reaction.In the North Branch, pCO 2 was measured from 930 to 1518 µatm at the salinity range between 8 and 16, which was substantially higher than the South Branch pCO 2 of 700 to 1100 µatm.Based on field data analyses and simplified stoichiometric equations, we suggest that the North Branch CO 2 productions were quantified by biogeochemical processes combining organic matter decomposition, nitrification, CaCO 3 dissolution, and acid-base reactions in the estuarine mixing zone.Although our study is subject to limited temporal and spatial coverage of sampling, we have demonstrated a procedure to quantificationally constrain net CO 2 productions in eutrophic estuaries and/or coastal lagoons, by coupling the chemical dynamics of carbonate and dissolved inorganic nitrogen systems.

Introduction
Large-river estuaries are important interfaces between continents and the oceans.They are biogeochemical hot spots since they receive large inputs of particulate matters, organic carbon, and nutrients from continents and oceans to support high rates of metabolism and/or chemical reactions.Significantly, CO 2 emission from estuaries, coastal lagoons, and salt marshes to the atmosphere has been proposed as an important component of the global carbon cycle (Frankignoulle et al., 1996(Frankignoulle et al., , 1998;;Cai and Wang, 1998).The global estuarine CO 2 emission rate has been estimated at 0.1 to 0.5 Gt C yr −1 (Borges, 2005;Borges et al., 2005;Cai, 2011;Chen et al., 2013).Also the chemical dynamics in large-river estuaries and their biogenic element fluxes have global and/or regional impacts on marine biogeochemistry (e.g., Justić et al., 1995;Bricker et al., 2008;Liu et al., 2015).So far, mechanisms supporting the estuarine CO 2 emission rates need to be better understood, especially in those eutrophic large-river estuaries.Cai (2011) suggested that respiration processes may not occur in the large-river estuaries due to short water transit or residence times.However, many medium to large rivers have forked estuaries.For example, the inner estuary of the Yangtze River (Changjiang) is divided into two primary branches by the Chong-ming Island (Fig. 1).The Channel-like South Branch is subject to very low respiration rates (Zhai et al., 2007), while the lagoon-like North Branch shows substantial additions of free CO 2 and dissolved organic matters (Zheng et al., 2011;Guo et al., 2014).We contend that, if the estuarine areas have a sufficient water residence time (such as the North Branch of the inner Changjiang Estuary), they may function as sites of terrestrial carbon incineration.
In addition, biogeochemistry of dissolved inorganic carbon and nutrients in the North Branch are rarely studied.How the North Branch biogeochemistry sustains the estuarine CO 2 additions reported by Zheng et al. (2011) remains unknown.Furthermore, the North Branch is occupied by saline water rather than freshwater, especially in relatively dryer seasons (He et al., 2006;Zheng et al., 2011;Guo et al., 2014).According Introduction

Conclusions References
Tables Figures

Back Close
Full to earlier results (e.g., Mao et al., 2001), the saltwater spillover from the North Branch usually occurs during spring-tide periods (with tidal ranges of more than 2.5 m) in winter and spring, with water discharge rates of lower than 25 000 m 3 s −1 (from the Datong Station, ∼ 400 km upstream of the study area).The saltwater spillover from the North Branch has received increasing attention since the occurrence of salty water in the South Branch threatens an important water source of Shanghai, the Qingcaosha Reservior (e.g., Shen et al., 1980;Mao et al., 2001;He et al., 2006;Wu et al., 2006;Xue et al., 2009;Qiu and Zhu, 2013).However, the potential impacts of spillover signals from the North Branch (Fig. 1a) on the budgets of carbon and nitrogen in the South Branch are poorly understood.
To evaluate the impacts of the North Branch saltwater spillover on the South Branch biogeochemistry, we investigated carbonate system and nutrients in the inner Changjiang Estuary in April 2010, covering both of the two primary branches (Fig. 1), shortly after a spring tide period.Together with simultaneous dissolved oxygen (DO) and partial pressure of CO 2 (pCO 2 ) data, this dataset also provides an opportunity to quantificationally examine how coupled dynamics of nitrogen and carbon elements affect CO 2 degassing fluxes from this important large-river estuary to the atmosphere, which may help to understand CO 2 dynamics and the controlling processes in those eutrophic large-river estuaries and coastal lagoons of the world.Introduction

Conclusions References
Tables Figures

Back Close
Full Changjiang is also one of the most important solid transporting rivers (Gaillardet et al., 1999), although its downstream solid content has declined from ∼ 600 mg L −1 in the 1960s to < 400 mg L −1 in the 2000s (Li and Zhang, 2003;Lin et al., 2007).Since its upstream water flows through the Yun-Gui Plateau and the Sichuan Basin, where the basement rocks are abundant in carbonates (Chen et al., 2002), the suspended sediment in the Changjiang River is rich of calcite (Chen et al., 2001).Also the Changjiang water is characterized by high alkalinity (Chen et al., 2002;Wu et al., 2007;Zhai et al., 2007) as compared with many other major rivers in the world.
The inner Changjiang Estuary is a large tidal estuary, complicated by bathymetry, islands, and deep channels (Xue et al., 2009).Nearly all of the river flow is transported via its South Branch into the ECS (Shen, 2001;Qu, 2010).The North Branch of the inner Changjiang Estuary is relatively shallow (water depth ∼ 3 m) and open to navigation only at high tides in a day.At low tide periods, however, the North Branch is isolated from the Changjiang main stream.
In the outer Changjiang Estuary, a coastal current (i.e., Yellow Sea Coastal Current, YSCC for short) affects the hydrology and hydrochemistry during the northeast monsoon period from October to April in the following year (Fig. 1a; Chen, 2009).This current has higher dissolved inorganic carbon (DIC) and total alkalinity (TAlk) as compared with most of ECS surface waters (Zhai et al., 2014).

Survey design
During 2 to 7 April 2010, a sampling cruise was carried out in the inner Changjiang Estuary.Shortly before the cruise, the water discharge rates (from the Datong Station) increased from 17 600 m 3 s −1 in 27 March to 20 800 m 3 s −1 in 2 April (Fig. 2a).
Therefore, this cruise represented a transitional period between the dry seasons (usually with the water discharge rate of less than 14 000 m 3 s −1 , from December of the last year to February) and the flood seasons (usually with the water discharge rate of more than 40 000 m 3 s −1 , from May to September).Overall, our cruise included three surveys/legs in the South Branch, one survey in the North Branch, and one survey in Introduction

References
Tables Figures

Back Close
Full the outer Changjiang Estuary.The three South Branch surveys occupied the 3rd day, 6th day, and 7th day after the spring-tide day in 1 April (Fig. 2b).In the North Branch, two anchored stations were repeatedly sampled for 12 to 20 h against tide height and salinity variations.

Sampling and analyses
Using an underway pumping system similar to Zhai et al. (2005Zhai et al. ( , 2007)), surface water (at a depth of ∼ 1 m) was pumped from a side intake for continuous measurements of hydrochemical parameters such as temperature, salinity, DO and pCO 2 .Via a side vent of our pumping system, discrete samples for Winkler DO, pH, DIC, TAlk, nutrients and particulate matters were collected at selected sites (Fig. 1).
Surface water temperature and salinity were continuously determined with a precalibrated YSI ® 6600 m, with the precisions of ±0.01 • C and ±0.01 salinity units.And the underway salinity data were validated by simultaneous discrete salinity data.Aqueous pCO 2 was continuously detected by a Li-Cor ® non-dispersive infrared spectrometer (Li-7000) together with a continuous flow and fully sealed cylinder-type equilibrator (Zhai et al., 2007;Jiang et al., 2008).For calibration purposes, four CO 2 gas standards with CO 2 molecular fractions from 400 to 1510 µmol mol −1 were applied.The uncertainty of these standards was ∼ 1 %, which represents the maximum level of uncertainty during the period of extensive measuring of pCO 2 and data processing (see details in Zhai et al., 2005).The field-measured atmospheric CO 2 data were corrected to survey-based barometric pressure at 10 m above the water surface and 100 % humidity at water surface temperature and salinity, following the procedure described in Zhai et al. (2007).
For discrete salinity measurements, water samples were sealed in 140 mL highdensity polyethylene bottles and kept at room temperature.They were determined in a week using a WTW TetraCon ® 325 probe based on conductivity measurements, with a precision of ±0.1 salinity units (Yan et al., 2012).Winkler DO and NIST (National In-Introduction

Conclusions References
Tables Figures

Back Close
Full stitute of Standards and Technology, USA)-traceable pH (at 15 • C) were sampled and determined according to Zhai et al. (2007Zhai et al. ( , 2012)).The possible nitrite interference in the DO titration was removed by adding 0.01 % NaN 3 during sample fixation (Wong, 2012).pH was measured on board using a precision pH meter and an Orion ® 8102BN Ross electrode, which were calibrated against three NIST-traceable buffers (pH = 4.00, 7.03, and 10.12 at 15.0 • C, Thermo Fisher Scientific Inc., USA).The precisions of DO and pH data were ±0.5 % DO and ±0.01 pH, respectively (Zhai et al., 2012).To express the oxygen consumption during water mixing, DO saturation was calculated from field-measured DO concentrations divided by DO concentration at equilibrium with the atmosphere.The latter was calculated as per the Benson and Krause (1984) equation.
For measurements of nutrient elements and TAlk, trinary water samples were filtered on board with 0.45 µm cellulose acetate membranes (Zhai et al., 2007;Yan et al., 2012).One of these was poisoned with 0.1 % chloroform and preserved at 4 • C for NH + 4 −N (ammonium) and silicate determination (within a time frame of 10 days upon sampling).The second one was deep-frozen and kept at −20 • C for NO − 3 −N (nitrate), NO − 2 −N (nitrite) and phosphate determination (within a time frame of 25 days upon sampling).The third one (for TAlk) was stored in a 140 mL high-density polyethylene bottle, immediately mixed with 50 µL of saturated HgCl 2 , and then sealed and preserved at room temperature until determination (within a time frame of 15 days upon sampling).Water samples for DIC were unfiltered but allowed to settle before measurement.They were stored in 60 mL borosilicate glass bottles (bubble free), and also preserved with 50 µL of saturated HgCl 2 and determined within a time frame of 7 days upon sampling.
Water samples for suspended particulate matter (SPM), particulate organic carbon (POC), and particulate inorganic carbon (PIC) were filtered on board with carbon-free 0.7 µm quartz microfiber (GFF) membranes.The SPM data were collected after drying in an oven at 50 • C.And then the membranes were divided into two equal parcels.One of them was prepared for total particulate carbon measurements (without acid fuming), and another one was fumed with concentrated HCl so as to remove carbonate before POC determination.Both total particulate carbon and POC were measured according Introduction

Conclusions References
Tables Figures

Back Close
Full to the JGOFS protocols (Knap et al., 1996), using a PE-2400 SERIES II CHNS/O analyzer.Finally, the PIC data were obtained from the difference between total particulate carbon and POC.NO  et al., 2001;Tzollas et al., 2010).
Following Zhai et al. (2007), DIC was measured by infrared detection after acid extraction of a 0.5 mL sample with a Kloehn ® digital syringe pump, while TAlk was determined by Gran acidimetric titration on a 25 mL sample with another Kloehn ® digital syringe pump, using a precision pH meter and an Orion ® 8102BN Ross electrode for detection.To ensure the measuring quality of DIC and TAlk, certificated reference materials from Andrew G. Dickson's laboratory (Scripps Institute of Oceanography) were regularly checked at a precision of ±2 µmol kg −1 .

Flux estimation
Riverine element export fluxes were estimated based on the element concentrations at the river end, i.e. west side of the Chongming Island (Fig. 1a).

Calculating carbonate system parameters from DIC and TAlk
For the purpose of modeling calculations, aqueous pCO 2 and concentrations of bicarbonate and carbonate ions were calculated from the DIC, TAlk, silicate, phosphate, seawater temperature, and salinity values using the program CO2SYS.xls (Pelletier et al., 2011), an updated version of the original CO2SYS.EXE (Lewis and Wallace, 1998).The dissociation constants for carbonic acid were those determined by Millero et al. (2006), and the dissociation constant for the HSO − 4 ion was determined as per Dickson (1990).

Hydrological and particle backgrounds
During the cruise, water temperature ranged from 10.6 to 13.8 • C. In the North Branch, water temperature was detected as low as 10.9±0.3 • C, lower than the nearby offshore sea surface temperature observed in early spring (Zhai et al., 2014).The relatively high temperature of > 13 • C was only detected in the South Branch.
Salinity in the South Branch varied day by day (Fig. 3a).In the 3rd day after the spring-tide day, it ranged from 0.13 to 0.67, with a salinity peak at 121 • 22 E (Fig. 3a), ∼ 20 km downstream the west side of the Chong-ming Island.In the 6th day after the spring-tide day, it changed from 0.23 to 0.54, with a peak at ∼ 40 km downstream the earlier salinity peak (Fig. 3a).These patterns were similar to results reported by He et al. (2006).In the 7th day after the spring-tide day, the salinity peak in the South Branch moved to downstream further, while salinity at the upstream and middle stations ranged from 0.14 to 0.16 (Fig. 3a).In the North Branch, however, very high salinity of 4.5 to 17.4 was determined in the surface water (Fig. 1b), showing the occupation of saline water (He et al., 2006;Zheng et al., 2011;Guo et al., 2014).East to the line of Introduction

Conclusions References
Tables Figures

Back Close
Full 122 • E, surface salinity ranged from 0.23 to 18.4, showing the combined influences of Changjiang dilution and YSCC waters (Fig. 1a; Zhai et al., 2014).The South Branch SPM ranged from 25 to 202 mg L −1 .It was lower than earlier SPM values of 240 to 720 mg L −1 between 1960 and 2001 (Li and Zhang, 2003), but similar to recent SPM data of 20 to 190 mg L −1 between 2003 and 2005 (Lin et al., 2007).
In the North Branch, however, surface water SPM was at very high levels of 390 to 455 mg L −1 , with POC of 259 to 417 µmol L −1 and PIC of 253 to 293 µmol L −1 .Particle related data are presented in the Appendix (Fig. A1) for public reference.

Distributions of hydrochemical parameters in the South Branch
In the South Branch, surface water DO saturation ranged from 86 to 93 % (Fig. 3b), similar to previous results reported by Zhai et al. (2007).In contrast, pH varied well and declined slightly from 3 April to 7 April (Fig. 3c).In 3rd day after the spring-tide day, pH showed a peak value of 8.06 at the site with the highest salinity.In 6th day after the spring-tide day, although the salinity varied from 0.23 to 0.54, pH ranged smoothly from 7.88 to 7.94.During the last survey in 7 April, pH ranged from 7.75 to 7.88.
NH + 4 −N ranged from 5.0 to 25.2 µmol kg −1 (Fig. 3d), while NO − 2 −N declined from 4.05-4.36µmol kg −1 at our upstream stations to 2.03 µmol kg −1 at the river mouth stations (Fig. 3e).As the products from chemical fertilizer application and soil erosion in the drainage basin of Changjiang, riverine NO − 3 −N ranged from 126.5 to 139.9 µmol kg −1 (Fig. 3f).The concentration of NO − 3 −N was one or two magnitudes higher than those of NH  3g).Except for a sampling site presumably influenced by a sewage outlet (Chai et al., 2006), the distributions of the two major DIN species (NH data showed an inverse trend (Fig. 3f).DIN in this water mass slightly declined during the three days from 3 April to 6 April (Fig. 3g).As a particle-concentrated element (e.g., Das et al., 2006;Leote et al., 2013), phosphate was measured at relatively low levels from 0.87 to 1.63 µmol kg −1 (Fig. 3h).However, silicate was at relatively high levels from 99.6 to 104.9 µmol kg −1 (Fig. 3i).
Except for the sampling site presumably influenced by sewage, the two carbonate system parameters (i.e.TAlk and DIC) also reflected the salinity variation patterns (Fig. 3l and m).TAlk ranged from 1495 to 1694 µmol kg −1 , while DIC ranged from 1520 to 1654 µmol kg −1 .
The South Branch pCO 2 ranged from 628 to 1148 µatm (Fig. 3n).Its distribution patterns roughly mirrored the distributions of pH (Fig. 3).Similar to earlier results reported by Zhai et al. (2007) and Chen et al. (2008), pCO 2 was always higher than the waterair equilibration level of 398 to 411 µatm (Fig. 3n).Hydrochemical parameters against salinity in the South Branch are plotted in the Appendix (Fig. A2) for public reference.

Water mixing behaviors of hydrochemical parameters in the North Branch and the outer estuary area
In the North Branch, surface water DO saturation ranged from 83 to 95 % (Fig. 4a).
In the outer estuary area, only two stations were sampled for nutrient analyses.Silicate data in both the outer area and the South Branch followed the water mixing line obtained from the North Branch (Fig. 4i), suggesting that water mixing in the outer area was controlled by nearly the same end-members as those in the North Branch.Therefore, we assumed conservative water mixing lines of dissolved nitrogen species in the Introduction

Conclusions References
Tables Figures

Back Close
Full where seawater end-members of NO − 3 −N (∼ 3 µmol kg −1 at a salinity of 31.5) and DIN (∼ 10 µmol kg −1 at a salinity of 30) referred to spring YSCC values reported by Chen et al. (2009) and Wang et al. (2003).
TAlk and DIC also showed additions in the North Branch, while they behaved like conservative elements in the outer area (Fig. 5).In the outer area, the simplified water mixing lines of TAlk and DIC, with seawater end-members of them from YSCC obtained in April 2007 (Zhai et al., 2014), were as follow: TAlk conservative (µmol kg −1 ) = 22.068 × Salinity + 1580 (5) DIC conservative (µmol kg −1 ) = 14.551 × Salinity + 1600 (6) Most of field-measured TAlk and DIC values in the outer area satisfactorily followed Eqs. ( 5) and (6) (Fig. 5), giving confidence in the water mixing line reduction.The intercepts of Eqs. ( 5) and ( 6) were consistent with TAlk and DIC values at the river mouth (1595 to 1615 µmol kg −1 ), and slightly higher than TAlk and DIC values at the river end (1495 to 1585 µmol kg −1 ) (Fig. 3).However, they were ∼ 200 µmol kg −1 lower than the freshwater end-member values of TAlk and DIC in spring of 2006 and 2007 (∼ 1800 µmol kg −1 ) (Zhai et al., 2007(Zhai et al., , 2014)).

Conclusions References
Tables Figures

Back Close
Full the relatively high-salinity value was consistent with the earlier reported YSCC pCO 2 range of 310 to 377 µatm obtained in April 2007 (Zhai et al., 2014).
To better evaluate the biogeochemical additions/removals of nitrogen and carbon elements, we calculated the conservative concentrations of DIN species and carbonate system parameters at salinity, according to Eqs. ( 1)-( 6).And then we obtained the differences between the measured concentrations and the conservative concentrations along the mixing lines, i.e., ∆NH + 4 −N, ∆NO − 2 −N, ∆NO − 3 −N, ∆DIN, ∆TAlk and ∆DIC in Eqs. ( 7)-( 12), with positive values indicating biogeochemical additions and negative values indicating removals.Note that both of the additions and removals are relative to the conservative mixing between the Changjiang freshwater end-member near the river mouth and the spring YSCC surface water end-member.

Residence time estimation
Estuaries are not only transport passages of terrestrial materials from the continents to the oceans, but also chemical reactors and/or buffers (Officer, 1979).Before discussing biogeochemical processes and element transport fluxes in the inner Changjiang Estuary, the residence times need to be examined.To estimate the residence times, we balanced water and salt budgets under the condition of the North Branch saltwater spillover shortly after the spring-tide period: where Q S , Q R and Q N are water discharges from the South Branch, from the upper river, and the mean water flux of the North Branch saltwater spillover during the springtide period in early April 2010 (Fig. 1a); S S , S R and S N are mean salinity of the South Branch water, the Changjiang River water and the North Branch water.In this study, Q R and S R were 20 000 m 3 s −1 and 0.14, respectively (Figs. 2a and 3a).S S and S N were estimated at 0.30 and 15.26, based on a compilation of our underway data obtained from 2 to 5 April (data partially reported in Figs.1b and 3a).Therefore, Q S and Q N were estimated at 20 214 and 214 m 3 s −1 , respectively (Table 1).
Based on Shen (2001), Meng and Cheng (2005), the navigation map (2010 edition), and our field records, mean water depths of the South Branch (d S ) and the North Branch (d N ) were estimated at 8 and 3 m, respectively, while water areas of the South Branch (A S ) and the North Branch (A N ) were 900 and 422 km 2 , respectively (Table 1).
Therefore, the mean residence times of the North Branch water (τ N ) and the South Introduction

Conclusions References
Tables Figures

Back Close
Full Branch water (τ S ) were estimated as bellow: We must point out that Eq. ( 16) ignored the tidal effect.That is, effluent water particles across the interface between inner and outer estuaries can return to the inner estuary (i.e., the South Branch) during the next tide cycle, leading to a longer exposure time of water particles than the steady-state residence time (Monsen et al., 2002).Based on Fig. 3a, the North Branch saltwater spillover (presumably in 1 April) induced salinity peak in the South Branch moved from 121 • 22 E in 3 April to 121 • 50 E in 6 April.Even in 7 April, the salinity peak was still observable inside the river mouth (west of 122 • 00 E) (Fig. 3a).Thus the South Branch water was exposed to the estuarine biogeochemical processes for ∼ 7 days in spring, more than 1.5 times of the residence time estimated under the steady-state assumption via Eq.( 16).Similar to the South Branch, the exposure time of the North Branch water estimated via the steady-state Eq. ( 15) is also subject to uncertainties.However, Eq. ( 15) suggested that the North Branch water had a very long residence time of more than 2 months in spring.It provided reaction times for many biogeochemical processes to function.

Maintaining mechanisms of the biogeochemical additions in the North Branch
In the North Branch, DIC and TAlk were linearly correlated (n = 13, r = 0.996), with a slope of 0.9888 that was very close to 1 (Fig. 7a).Since DIC is the sum of free CO 2 , HCO TAlk plot (with a slope lower than 0.5) as driven by water mixing between the Changjiang freshwater and ECS seawaters (Zhai et al., 2007).

BGD Introduction Conclusions References
Tables Figures

Back Close
Full ∆TAlk and ∆DIN were also linearly correlated in the North Branch (n = 13, r = 0.985), with a slope of 6.56 (Fig. 7b).This plot suggested that both the carbonate system and dissolved inorganic nitrogen dynamics in the North Branch might controlled by same biogeochemical processes.To examine these processes from a quantitative point of view, we analyzed the stoichiometric relationship of the Redfield respiration.Also resolved were the nitrification, the CaCO 3 dissolution, and the acid-base reaction during the oxidation of organic matters.
In a saline aquatic environment with abundant oxygen and the pH of ∼ 8, the Redfield respiration of biogenic organic matters is directly associated with a release of NH As compared with the decomposition of organic matters, the oxidation of NH + 4 −N (a key step of nitrification) is a time-expensive reaction (e.g., Dai et al., 2008), especially when the temperature is lower than 15 • C. The nitrification occurs as below: Combining Reactions (R1)-(R3), a complete Redfield respiration of biogenic organic matters is: In the North Branch, these processes were associated with the consumption of DO and the releases of free CO 2 and the three dissolved inorganic nitrogen species (Fig. 4).To reveal how many NH + 4 −N released through respiration were transformed into NO − 2 −N and/or NO  R4) based on this ratio, ignoring the incomplete nitrite oxidation, the local status of Redfield respiration in the North Branch was better characterized as below: Furthermore, the North Branch was an environment with abundant supplies of PIC (mostly calcite or CaCO 3 ) and seawater (having the CO 2− 3 ion).Besides CO 2 degassing, most of the above-mentioned estuarine CO 2 productions are potentially removed by CaCO 3 dissolution (Abril et al., 2003) and/or CO 2− 3 titration: Except for the CO 2− 3 titration (Reaction R7), all of respiration (Reaction R5), nitrification (Reaction R2) and CaCO 3 dissolution (Reaction R6) affected the ratio of ∆DIC to ∆TAlk (Fig. 7d).In the North Branch, both ∆DIC and ∆TAlk were positive, suggesting net biogeochemical additions.∆DIC = 1.09 × ∆TAlk (n = 13, r = 0.992).The slope of 1.09 (higher than the CaCO 3 dissolution induced slope of 0.5) suggested that respiration and CaCO 3 dissolution controlled the North Branch carbonate system as below: In the South Branch, however, most ∆DIC and ∆TAlk were negative, suggesting biogeochemical removals.∆DIC = 0.66×∆TAlk−21 µmol kg −1 (Fig. 7d).The slope of 0.66 together with the negative ∆DIC and ∆TAlk values suggested that both CO 2 degassing Introduction

Conclusions References
Tables Figures

Back Close
Full and nitrification controlled the South Branch carbonate system (Fig. 7d).The later process was also evidenced by the above-mentioned dissolved nitrogen system dynamics in Fig. 3.Although CaCO 3 dissolution may also occur in the South Branch, as indicated by the facts that several ∆DIC and ∆TAlk plots in the South Branch were positive, and the slope of 0.66 was similar to the CaCO 3 dissolution induced slope of 0.5 (Fig. 7c), its effects on the South Branch carbonate system were negligible (Zhai et al., 2007).In summary, the biogeochemical additions of NH + 4 −N and carbonate observed in the North Branch were driven by the biogeochemical processes combining organic matter decomposition, nitrification, and CaCO 3 dissolution, as described by Reaction (R8).Abundant organic matters and CaCO 3 particles, together with the quite long residence time and suitable environmental conditions such as temperature and oxygen, had made the North Branch into a natural reactor.This big reactor affected both element export fluxes (via its water spillover, see Sect.4.4) and the water-to-air CO 2 flux in this important estuary (via increasing aqueous pCO 2 , see Sect.4.3).

Estuarine CO 2 dynamics in the North Branch
Reaction (R8) suggested that, if most of estuarine CO 2 products from Redfield respirations are removed by CaCO 3 dissolution, the ratio of ∆TAlk to ∆DIN should be 176/16 = 11.125.Alternatively, if none of the CO 2 products is removed by the CaCO 3 dissolution, the ratio of ∆TAlk to ∆DIN should be 2/16 = 0.125 based on Reaction (R5).Therefore, the real ratio of ∆TAlk to ∆DIN of 6.56 (Fig. 7b) suggested that only 6.56/(11.125 − 0.125) × 100 % = 60 % of respiration-induced free CO 2 was removed via CaCO 3 dissolution.Ignoring the minor impacts of CO 2 degassing fluxes (due to the much slower air-water exchanging rates as compared with the acid-base reactions), the other 40 % of estuarine CO 2 products were potentially titrated by CO 2− 3 ion supplied by the seawater end-member, as indicated by Reaction (R7).
To discuss this effect, we calculated pCO addition , where K H is the solubility coefficient of CO 2 , calculated via the Weiss (1974) equation.Figure 8a showed that [CO 2 * ] addition ranged from 2.7 to 53 µmol kg −1 in the North Branch, usually lower than those predicted values based on Reaction (R8) and our ∆DIN data.The differences were likely caused by acid-base titration between respiration-induced free CO 2 and seawater-introduced CO 2− 3 ion.Figure 8b showed that conservative concentrations of non-carbonate alkalinity (such as borate ions) ranged from 7 µmol kg −1 at salinity 4 to 30 µmol kg −1 at salinity 16.These non-carbonate alkalinity values primarily accounted for the observed differences between TAlk and DIC in the North Branch, i.e. 22 to 29 µmol kg −1 as suggested by Fig. 7a.This comparison provided another evidence supporting the idea that seawaterintroduced [CO 2− 3 ] conservative was mostly titrated by respiration-induced CO 2 , and transferred into HCO − 3 ions.Furthermore, we regarded the conservative concentrations of CO 2− 3 ion (Fig. 8b) as the maximum removals via CO 2− 3 titration.Figure 8a showed that the predicted val- conservative were highly consistent with the field-measured values of [CO 2 * ] addition at salinity of 5 to 13, indicating that seawater-introduced [CO 2− 3 ] conservative was fully titrated by respiration-induced CO 2 .However, the predicted values were lower than the real at salinity of > 13, presumably due to the overestimate of CO 2− 3 titration reactions there.
In summary, we have quantificationally demonstrated that the observed aqueous pCO 2 in the North Branch was determined by water mixing and several biogeochemical processes.Although 80 to 85 % of estuarine CO 2 production from organic matter decomposition and nitrification had been removed by CaCO 3 dissolution (∼ 60 %) and CO 2− 3 titration (50 to 60 % of the residuals, Fig. 8), the North Branch acted as a significant source area of the atmospheric CO 2 with higher pCO 2 than the South Branch (Fig. 4c).To determine the seasonality and magnitude of the North Branch water-to-air CO 2 flux, more investigations are needed.

Conclusions References
Tables Figures

Back Close
Full The North Branch biogeochemical signals intrude to Changjiang mainstream (the South Branch) via the tide induced saltwater spillover process in dry seasons (Fig. 1a), potentially increasing element export fluxes to the ECS.Considering the steady state spillover flux of saline water (∼ 214 m 3 s −1 , Table 1) from the North Branch into the Changjiang mainstream, the spillover flux of salt was estimated at 3265 kg s −1 , slightly higher than the riverine transport salt flux (Table 2).Similar to the salt flux estimation, riverine transport fluxes and spillover fluxes of TAlk, DIC, and nutrients were calculated (Table 2).
Generally, spillover fluxes of TAlk, DIC, and nutrients from the North Branch into the Changjiang mainstream were minor, as compared with the Changjiang transport fluxes.However, the springtime areal yield rates of TAlk, DIC, and DIN species from the North Branch were 34 to 133 times higher than those from the Changjiang drainage basin, as Changjiang has a drainage area of 1.8×10 6 km 2 (Chen et al., 2002), roughly 4000 times of the North Branch water area.Note that the organic matters and CaCO 3 particles supporting the springtime TAlk, DIC, and DIN production in the North Branch were also supplied by Changjiang during its flood seasons.Therefore, the North Branch may serve as an estuarine chemical buffer, to some extent smoothing seasonal variations of Changjiang element export fluxes.
During our sampling period, the spillover of TAlk, DIC, and DIN species from the North Branch affected the biogeochemistry in the South Branch.Firstly, the spillover water transported more TAlk than DIC into the South Branch (Table 2), leading to the South Branch pCO 2 decline from ∼ 1000 µatm at salinity of ∼ 0.14 (with the water-to-air pCO 2 difference of ∼ 600 µatm) to ∼ 700 µatm at salinity of > 0.4 (with the water-to-air pCO 2 difference of ∼ 300 µatm) (Fig. 3n).If the air-water gas transfer velocity was changeless, the 50 % decline of water-to-air pCO 2 difference means that the waterto-air CO 2 degassing flux had decreased by 50 % due to the spillover of salty water.

BGD Figures Back Close
Full Secondly, the spillover water likely contained active nitrifiers, which continued the nitrification reaction in the South Branch, as indicated by the NH + 4 −N decline and the NO − 3 −N increase of the salty water mass from 3 to 6 April (Fig. 3).This was also evidenced by the pH decline and the pCO 2 increase signals (at salinity of > 0.4) during the same period (Fig. 3), as shown by Reaction (R2).

Concluding remarks
This study showed that the spillover waters from the North Branch had minor contributions to the Changjiang transport fluxes of nutrients, dissolved inorganic carbon, and alkalinity.They affected the biogeochemistry in the South Branch, by lowering water-to-air CO 2 fluxes and continuing the nitrification.Significantly, several primary biogeochemical processes such as organic matter decomposition, nitrification, CaCO 3 dissolution, and acid-base reaction occurred in the North Branch of the Changjiang Estuary, leading to the unusual high pCO 2 values at middle salinity and the biogeochemical additions of dissolved inorganic nitrogen species and carbonate system parameters.Similar biogeochemical processes may occur in many eutrophic estuaries and/or coastal lagoons of the world.This study demonstrated a procedure to quantificationally analyze the coupled dynamics of dissolved inorganic nitrogen and carbonate systems, which may help to better understand the combined effects of metabolic processes and chemical reactions on CO 2 fluxes in estuaries and/or coastal lagoons with similar physical and biogeochemical conditions.

Conclusions References
Tables Figures

Back Close
Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | and phosphate were determined using an AA3 Auto-Analyzer (Bran + Luebbe Co., Germany), while NH + 4 −N and silicate were measured using a Tri-223 continuous Auto-Analyzer (see details inYan et al., 2012).Briefly, NO − 3 −N and NO − 2 −N were measured by reducing NO − 3 to NO − 2 with a Cd column, and then determining NO − 2 using the standard pink azo dye spectrophotometric method.Phosphate, silicate and NH + 4 −N were measured based on the standard phospho-molybdenum blue, silicon molybdenum blue and indophenol blue spectrophotometric procedures.Note that an appropriate quantity of more reagent NaOH was added during the NH + 4 −N measurement so as to make the final pH within the optimum range of 10.5 ± 0.1 (Pai Water discharge data needed were measured at the Datong Station, as released by the China Bureau of Hydrology (http://xxfb.hydroinfo.gov.cn/).The North Branch saltwater spillover induced element fluxes were estimated based on a mass balance approach.See Sects.4.1 and 4.4 for details. Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |
Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Figure 1 .Figure 2 .
Figure 1.Area map (a) and surface distributions of salinity, total alkalinity and ammonia during the April 2010 cruise (b-d).Following Zhai et al. (2007), we regarded the line of 122 • E as the boundary between inner sub-estuaries and the outer estuary area.ECS = East China Sea; CM = Chong-ming Island; YSCC = Yellow Sea Coastal Current.As sketched in (a), the inner Changjiang Estuary is geographically divided into three sub-estuaries by the Chong-ming Island and several sandbanks, i.e. the North Branch (I), the north channel of the South Branch (II), and the south channel of the South Branch (III).Four neighboring sampling sites in April 2007 (Zhai et al., 2014) were sketched in (a) by "+" symbols.The pentacle shows a reference station for the tidal cycle.

Figure 3 .
Figure 3. Evolution of surface water hydrochemical parameters in the South Branch of the inner Changjaing Estuary in early April 2010, with arrows showing signals likely from a sewage outlet, as noticed by Chai et al. (2006).Dashed lines in (d-m) indicate the assumed element concentrations at the river end.

Figure 4 .Figure 5 .Figure 6 .Figure 7 .Figure A2 .
Figure 4. Surface water distributions of dissolved oxygen, pH, partial pressure of CO 2 , and dissolved inorganic nutrients against salinity during the survey.Grey line in (c) is calculated based on conservative mixing lines of TAlk, DIC, and silicate.