Distribution and Affecting Factors of Aragonite Saturation in the Northern South China Sea in Summer

Abstract

Based on the carbonate and hydrological parameters of a survey made in August–September 2011, we investigated the distribution and affecting factors of aragonite saturation (Ω_arag) in the northern South China Sea. The levels of Ω_arag were found to gradually decrease with depth in the northern South China Sea. Surface-water Ω_arag values ranged from 2.56 to 3.68, with the highest value occurring in the region of Pearl River-diluted water near the northern coast. The increase in Ω_arag due to primary production, stimulated by the Pearl River freshwater input, exceeded the decrease in Ω_arag due to the direct input of low-Ω_arag fresh water, resulting in high Ω_arag in that area. In contrast, Ω_arag levels below 2 generally appeared in subsurface water below 50 m in depth. Intense community respiration was the main reason for the low Ω_arag. By 2100, bottom-water Ω_arag levels could be lower than 1.7, and even the undersaturation of aragonite could appear, due to the oceanic absorption of atmospheric CO₂.

1. Introduction

Over the past century, oceans have absorbed about 1/3 of anthropogenic atmospheric carbon dioxide (CO₂) emissions, slowing down the pace of global warming [1]. However, this results in a decrease in pH, a process commonly referred to as “ocean acidification” [2,3]. Ocean acidification has a significant impact on the biochemical characteristics of seawater, such as biochemical reactions, equilibrium processes, and biotoxicity shifts. A typical example is that a decrease in seawater pH will lead to a decrease in CO₃²⁻ and carbonate saturation (Ω), which will weaken or even stagnate the calcification process of shellfish and corals [4,5], thereby threatening the health of marine ecosystems [6,7].

As the most abundant form of CaCO₃ in shallow waters [8], aragonite is usually about 50% more soluble than calcite at 25 °C [9]. Thus, an aragonite saturation (Ω_arag) may better reflect the impact of ocean acidification on biologically driven calcification [10,11]. In open waters, ocean acidification has caused a unit decline in Ω_arag of about 0.5 due to the absorption of atmospheric CO₂ after the Industrial Revolution [12], with the current global ocean average Ω_arag being ~3.0 [13]. However, in coastal waters, Ω_arag shows a larger variation range and the level is lower than that in open waters in most cases. For example, on the eastern shelf of the Gulf of Cádiz, the average Ω_arag in surface water is ~2.68, while it is ~2.05 in deep water [14]. In the Bering Sea, the recorded surface Ω_arag was between 1.43 and 3.17, while the bottom aragonite reading was observed to be strongly undersaturated, with the lowest Ω_arag value of 0.45 [15]. Although Ω_arag = 1 is generally considered to be the threshold for calcium carbonate dissolution, the calcification rates still declined to Ω_arag > 1 in waters where biologically driven calcium carbonate deposition occurred [10,16].

The reason why coastal Ω_arag levels are significantly different from those in open waters is that offshore areas are not only affected by the absorption of atmospheric CO₂ but are also affected by various human activities and natural processes, resulting in a more complex and intense control mechanism for Ω_arag [14,17,18]. There are several typical processes such as primary production [19], organic matter degradation [18], freshwater input [20], and high CO₂ water upwelling [21]. In addition, these processes are often synergistic or antagonistic to each other. Therefore, it is extremely important to carry out research on the distribution and control mechanism of Ω_arag in coastal areas.

The South China Sea is the largest marginal sea in China, with diverse coral reef ecosystems and rich fishery resources. The northern South China Sea is adjacent to the developed Pearl River Delta, an economic circle of China. Thus, large amounts of waste materials generated by human activities have been discharged into the sea along the Pearl River [22]. As a result, the variability and controls of Ω_arag are quite complex. Based on field surveys conducted in summer, the distribution and variability of Ω_arag were investigated in this region. Furthermore, we elucidated the control mechanisms of Ω_arag in the northern South China Sea.

2. Materials and Methods

2.1. Study Area

The South China Sea is the largest and deepest sea in China and is also the largest marginal sea in the Pacific Ocean, located to the south of the Chinese mainland (Figure 1). Consequently, there is active material exchange between the northern South China Sea and the northwestern Pacific Ocean. The northern South China Sea is also affected by coastal currents, as well as mesoscale processes such as upwelling, eddies, and typhoons. The northern South China Sea has a vast continental shelf, with a water depth of ~150 m and a shelf width ranging from 300 to 484 km. The Pearl River, the second-largest river in China, flows from the north into the South China Sea, thereby bringing large amounts of nutrients to the South China Sea [22]. In this study, we reference 4 transect field investigations conducted in the northeastern South China Sea from 29 August to 1 September 2011, ranging from 113° E to 116° E and 19° N to 23° N (Figure 1).

2.2. Sampling and Analytical Methods

Underwater temperature (T) and salinity (S) were measured using the SBE 45 MicroTSG (SeaBird Inc., Bellevue, WA, USA). The nominal readings were 0.002 °C for temperature and 0.005 for salinity.

For the analysis of dissolved inorganic carbon (DIC) and total alkalinity (TAlk), the DIC samples were directly collected from the sampler using a disposable syringe filter, treated with saturated mercury chloride (final concentration ~0.02% by volume), and preserved at 4 °C.

The DIC samples were measured using coulometric titration, and the variation between duplicate samples was <1%. The TAlk samples were measured using single-point titration, and the variation between duplicate samples was <1%.

Ω_arag and pCO₂ levels were calculated from TAlk and DIC using the CO2SYS _v2.1 program (K1, K2 from Mehrbach et al.; 1973 refit by Dickson and Millero, 1987) [23,24].

Dissolved oxygen (DO) was measured using the Winkler titration method according to the specifications for marine monitoring–Part 4: Seawater analysis.

The survey collected a total of 65 data points, including 12 from the surface layer and 53 from the deeper layers.

3. Results

3.1. Hydrological Data and DO%

The distribution of salinity in the northern South China Sea is shown in Figure 2a,b. The surface salinity ranges from 31.77 to 33.60, with an average of 33.24 (Figure 2a). In most of the sea area, the salinity was relatively constant, with a range of 33.3–33.6. Values lower than 32 only occurred in the nearshore region of Transect C, due to freshwater input from the Pearl River. Bottom salinity was significantly higher than that at the surface, ranging from 34.26 to 34.57, with a relatively uniform regional distribution (Figure 2b). Vertically, the salinity of the four transects all gradually increased with depth (Figure 3a–d). Transects A, B, and D had small vertical variations, with difference values of ~1.5 (Figure 3a,b,d). However, Transect C had a greater salinity difference of 2.77 (Figure 3c), indicating the influence of Pearl River freshwater input.

Surface water temperature was generally high and evenly distributed, ranging from 29.72 to 31.00 °C (Figure 2c). The bottom temperature ranged from 13.43 to 23.12 °C, lower than that at the surface (Figure 2d). Spatially, the bottom temperature generally decreased from the nearshore to the offshore area. Vertically, the temperature of all the transects gradually decreased with depth (Figure 3e–h). Temperature differences were all above 10 °C with the maximum value in Transect A being as high as 16.62 °C.

The surface-water DO samples were all in an oversaturated state, with a DO% between 101% and 107% in most areas. A high value of 122% was observed in the Pearl River-affected area (Figure 2e). The bottom-water DO samples were all in an undersaturated state, ranging from 47% to 90% (Figure 2f). Vertically, with the increase in depth, DO% gradually decreased, and the shift from DO saturation to undersaturation occurred at a depth of ~40 m (Figure 3i–l).

3.2. Carbonate System Parameters

The surface-water DIC values ranged from 1842 μmol kg⁻¹ to 1931 μmol kg⁻¹, with a mean value of 1887 μmol kg⁻¹ (Figure 4a). Spatially, surface DIC gradually increased from the nearshore area to the offshore area; low values were found in the relatively low-salt waters affected by the Pearl River. The bottom-water DIC values exhibited strong spatial variation compared to the surface values, ranging from 1952 μmol kg⁻¹ to 2089 μmol kg⁻¹, with a mean value of 2017 μmol kg⁻¹ (Figure 4b). Bottom-water DIC values showed a similar distribution to those of the surface, while low values were mainly distributed on both sides of the area that was most affected by the Pearl River-diluted water. Vertically, DIC values in all the Transects gradually increased with depth, with the largest difference reaching 229 μmol kg⁻¹ in Transect A and the smallest difference being 162 μmol kg⁻¹ in Transect B (Figure 5a–d).

Surface-water TAlk values in the northeastern South China Sea ranged from 2128 μmol kg⁻¹ to 2177 μmol kg⁻¹, with a mean value of 2150 μmol kg⁻¹ (Figure 4c). Bottom-water TAlk values ranged from 2158 μmol kg⁻¹ to 2216 μmol kg⁻¹, with a mean value of 2183 μmol kg⁻¹ (Figure 4d). Vertically, TAlk values gradually increased with depths (Figure 5e–h). The profile distribution of TAlk values was consistent with DIC, while the variation was significantly smaller than that of DIC (Figure 5a–h).

Surface-water Ω_arag values ranged from 2.56 to 3.68 with a mean value of 3.10 (Figure 4e). Spatially, surface Ω_arag decreased from the nearshore area to the offshore area, with the highest values mainly occurring in the relatively low-salinity area affected by the Pearl River. The bottom-water Ω_arag values were generally low, with a range of 1.06–2.50 and a mean of 1.96 (Figure 4f). As the distance from the shore increased, Ω_arag significantly decreased. At the station farthest from the shore in Transect A, Ω_arag was only maintained in a state of saturation. Vertically, Ω_arag gradually decreased with depth (Figure 5). Ω_arag values below 2 generally appeared at depths greater than 50 m.

4. Discussion

4.1. Effects of Pearl River Freshwater Input on High Ω_arag Values in the Nearshore Area of Northern South China Sea

It can be seen from Figure 2a that surface salinity values lower than 32 occurred in the nearshore region of Transect C. Considering that the nearshore area of Transect C is close to the Pearl River Estuary, the decrease in salinity may be attributed to the freshwater input from the Pearl River. Additionally, the Ω_arag value in this area is the highest in the entire survey area, reaching 3.68, which may be related to the freshwater input from the Pearl River.

Generally, it is known that rivers have relatively lower values of Ca²⁺, DIC, and TAlk [25]; therefore, the mixing process due to direct freshwater input usually results in lower Ω_arag values [10]. It has been reported that the value for DIC is ~1100 μmol kg⁻¹ and for TAlk is ~1100 μmol kg⁻¹ in the Pearl River in summer [26]. We used the value of 1100 μmol kg⁻¹ as the freshwater endmember and the average values of DIC and TAlk in the surface water at the two farthest offshore stations (1910 μmol kg⁻¹, 2152 μmol kg⁻¹) as the ocean endmember, respectively. Using a two-end-member mixing model, it was calculated (through CO2SYS) that the direct input of freshwater from the Pearl River may lead to a maximum decrease of 0.14 units in Ω_arag in the offshore area of the northern South China Sea.

The Ω_arag value in the affected area of the Pearl River was the highest in the entire survey area (Figure 4e), but the direct input of the Pearl River played a role in reducing Ω_arag. It is clear that there were other processes that increased Ω_arag, making the Ω_arag in this area show the highest value in the survey area. We noticed that the DO% in the nearshore area was high, reaching 122% (Figure 2e). Additionally, the DIC value was relatively low, with a minimum of only 1842 μmol kg⁻¹ (Figure 4a). In addition, in summer, dissolved inorganic nitrogen (DIN) in the Pearl River is very high, at not less than 75 μmol L⁻¹ [27]. These phenomena all indicate that there might be strong primary production.

Based on the Redfield equation (Equation (1)) [28], when primary production absorbs CO₂, both HCO₃⁻ and CO₃²⁻ in the carbonate system will move toward the direction of supplementing CO₂ (Equation (2)① and (2)② to the left). In addition, the supplement of HCO₃⁻ to CO₂ is dominant and the supplement of CO₃²⁻ to CO₂ is minimal since HCO₃⁻ accounts for more than 90% of the seawater carbonate system. At the same time, the shift of HCO₃⁻ to CO₂ consumes H⁺ (Equation (2)①). In order to avoid a large increase in pH, the carbonate equilibrium system will inevitably decompose another part of HCO₃⁻ into CO₃²⁻ and H⁺ to keep the pH value stable, resulting in an increase in CO₃²⁻. As a result, Ω_arag will eventually increase, due to the increase in CO₃²⁻ (Equation (3), where K′sp is the apparent solubility product constant for aragonite). Therefore, the nutrients brought by the freshwater of the Pearl River could promote primary production, resulting in higher Ω_arag values in this area.

106CO₂ + 16HNO₃ + H₃PO₄ + 122H₂O ⇋ (CH₂O)₁₀₆(NH₃)₁₆H₃PO₄ + 138O₂

(1)

$${\mathrm{C}\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}\stackrel{\mathrm{①}}{\iff }{\mathrm{H}\mathrm{C}\mathrm{O}}_3^{-}+{\mathrm{H}}^{+}\stackrel{\mathrm{②}}{\iff }{\mathrm{C}\mathrm{O}}_3^{2-}+{2\mathrm{H}}^{+}$$

(2)

Ω_arag = [Ca²⁺][CO₃²⁻]/K′sp

(3)

In conclusion, the decrease in Ω_arag from direct low-Ω_arag river water input and the increase in Ω_arag from primary production, promoted by nutrient inputs, are two opposing processes. The level of Ω_arag depends on the relative strength of these two processes when influenced by freshwater input from rivers. In the nearshore area of Transect C, which is most affected by the Pearl River freshwater, the increase in Ω_arag by primary production exceeded the decrease from the direct input of fresh water, resulting in high Ω_arag values there. This phenomenon of high Ω_arag in the northern South China Sea was also reported by Cao [29]. Similar phenomena were also observed in some other seas. For example, in the southern Yellow Sea and the eastern China Sea, affected by the Yangtze River, nutrient inputs promoted primary production, resulting in high Ω_arag values in these two areas [30].

4.2. Dynamic Mechanism of Low Ω_arag in Subsurface Water

Subsurface Ω_arag in the northern South China Sea was generally low, with the values at depths below 50 m mostly at <2 (Figure 5i–l). Corresponding to the low Ω_arag, the DO in all samples was in an unsaturated state (Figure 3i–l) and the DIC value was significantly high (Figure 5a–d). Moreover, DO% and DIC gradually decreased and increased with depth, respectively, with the lowest DO% value being only 45% and the highest DIC value reaching 2089 μmol kg⁻¹ (Figure 3i–l and Figure 5a–d).

There might be various reasons for the low Ω_arag, such as terrestrial low-Ω_arag freshwater input, community respiration, the calcification of shellfish, and fish bones [31]. It is obvious that the influence of terrestrial low Ω_arag freshwater input is very small and could be ignored in the subsurface water of the northern South China Sea. Considering that the subsurface water Ω_arag value was mostly lower than 2 (Figure 4e,f), calcium carbonate precipitation was unlikely to occur. Therefore, the low subsurface Ω_arag values in the northern South China Sea may be caused by the introduction of CO₂ by community respiration/organic matter degradation.

The reason why Ω_arag decreases after CO₂ enters seawater is that CO₂ reacts with CO₃²⁻, resulting in a decrease in CO₃²⁻ (the numerator in Equation (3)), while K′sp and Ca²⁺ are both constants under the same temperature and salinity levels, as shown in Equation (3). According to the Redfield equation (Equation (1)), when 138 mol O₂ is consumed by community respiration, 106 mol CO₂ is produced at the same time. Thus, the ratio of CO₂ and O₂ in seawater will change, in a proportion of 106:138 (i.e., 0.77). Compared with the change ratio of CO₂ and O₂ in subsurface water to 0.77, it could be determined whether community respiration is the main control mechanism. Since CO₂ is one of the components of DIC, it would be converted into the other two components of DIC (HCO₃⁻ and CO₃²⁻) after entering seawater. Therefore, CO₂ variation cannot completely represent inorganic carbon variation, and DIC variation may be a more accurate parameter. Based on this assumption, we calculated the relative variations of DIC and O₂ in the subsurface water of the northern South China Sea, namely, ΔDIC and apparent oxygen utilization (AOU) (Equations (4) and (5)):

ΔDIC = DIC − DIC_pre (T, S, TAlk, pCO_{2 air})

(4)

where DIC and TAlk represent the in situ DIC and TAlk concentrations, T and S represent the in situ temperature and salinity, pCO_{2 air} represents atmospheric pCO₂, and DIC_pre is the DIC concentration when seawater pCO₂ is in balance with the atmosphere, which is calculated by CO2SYS:

AOU = ΔO₂ = [O₂] − [O₂]_eq,

(5)

where [O₂] is the in situ DO concentration, and [O₂]_eq is the DO concentration when seawater O₂ is in balance with the atmosphere.

AOU and ΔDIC were plotted, as shown in Figure 6. The results showed that there was a positive correlation between subsurface AOU and ΔDIC, and the slope was basically consistent with the Redfield ratio, indicating the controlling of community respiration on the subsurface water in the northern South China Sea. However, the scattering of ΔDIC and AOU lies above the Redfield line as a whole (Figure 6). This pattern might be related to the calculation methods used for AOU and ΔDIC. Both AOU and ΔDIC were calculated based on atmospheric O₂ and CO₂ levels. However, the exchange rates of seawater O₂ and CO₂ with the atmosphere are different. Usually, the exchange cycle of O₂ between the sea and atmosphere is very short, lasting about one week, while the exchange cycle of CO₂ could last up to several months [32]. The exchange rate of CO₂ produced by the subsurface community respiration was significantly lower than that of O₂, resulting in a certain surplus of DIC. Ultimately, the scatter points of ΔDIC and AOU were located above the Redfield line.

In the subsurface waters of the northern South China Sea, AOU and Ω_arag, ΔDIC, and Ω_arag all showed significant negative correlations (Figure 7a,b). The low values of Ω_arag corresponded to high values of AOU (that is, low values of DO%) and high values of ΔDIC (that is, high values of pCO₂), indicating the controlling of community respiration on subsurface Ω_arag.

4.3. Aragonite Saturation States of the Northern South China Sea in 2100

Since the Industrial Revolution, the absorption of anthropogenic CO₂ by the ocean has led to a decrease in its surface pH of ~0.1 unit, and, by the end of this century, another 0.3~0.4 pH decrease will have occurred [1]. Due to the influence of various human activities and natural processes, the pH value of coastal areas varies dramatically, which might have a greater impact on Ω_arag. Therefore, it is necessary to calculate future Ω_arag values in the nearshore area to assess the impact of future ocean acidification on corals, shellfish, etc.

To assess the impact of atmospheric CO₂ absorption on seawater Ω_arag in the northern South China Sea, we simulated Ω_arag at the end of this century using CO2SYS. According to the prediction in the Intergovernmental Panel on Climate Change (IPCC) IS92a ‘sustained increasing’ scenario, by 2100, atmospheric pCO₂ will rise to 788 ppm. The seawater pCO₂ will increase accordingly, and the increment is the difference between the atmospheric pCO₂ in 2100 and the current atmosphere. Since the absorption of atmospheric CO₂ by the ocean will not change the TAlk level, the TAlk level in the northern part of the South China Sea in 2100 will still be the same as the current level. Substituting the current TAlk level and the seawater pCO₂ in 2100 into CO2SYS, the Ω_arag value in 2100 in the northern South China Sea can be calculated. The results are shown in Figure 8 and Figure 9.

Based on the simulations, a notable decline of Ω_arag in the northern South China Sea by 2100 could be observed, according to Figure 9a–d. Specifically, compared with the current Ω_arag of between 1.06 and 3.68 (with a mean of 2.56; see Figure 4e,f and Figure 5i–l), the findings indicate a reduction in Ω_arag values, which are predicted to be from 0.81 to 2.40, with an average of 1.80. A Ω_arag value above 2 only occurred in the surface water (Figure 8a). In the bottom water, the Ω_arag value was consistently below 1.7 (Figure 8b). For the farthest and deepest offshore station in Transect A, there was even an unsaturated area of Ω_arag that appeared (Figure 9a).

Compared with previous and existing studies, this decrease is consistent with the predictions made by Yang et al. in a similar region (the western sea area of this study) that the increase in atmospheric CO₂ concentration and seawater temperature would cause a decrease in Ω_arag to 2.12 ± 0.22 by 2100 [10]. Similarly, studies in other regions such as Tokyo Bay, the southern Yellow Sea, and the BoHai and Yellow Seas have predicted the decline of Ω_arag [31,33,34]. Furthermore, using a future scenario to predict the impacts of rising atmospheric CO₂ levels on seasonal variations in Ω_arag values, Li and Zhai showed that very low Ω_arag values of <1.5 might exist all year round in the northern Yellow Sea subsurface waters by 2050, while aragonite-undersaturated water (Ω_arag < 1) might appear in autumn months, causing great stress to the area’s benthic faunal community [18].

Although, theoretically, calcium carbonate will only dissolve when Ω_arag is unsaturated (Ω_arag < 1), some experimental studies have found that larval bivalves could not synthesize aragonite, even if the Ω_arag value was 1.6 [35]. At an oyster hatchery on the Oregon coast, both larval production and the mid-stage growth rate of the oyster Crassostrea gigas were significantly negatively correlated with Ω_arag in the range of 0.8–3.2 [16]. The survival rate and skeletal density of Caribbean corals have declined to varying degrees in the submarine spring environment, with an average Ω_arag value of 2.1 over two years [36].

The northern South China Sea is rich in coral resources. In recent years, the coral reefs there have declined dramatically under the influence of a large number of human activities and global warming [37]. With the continuous increase in atmospheric CO₂ and the contribution of community respiration to subsurface water CO₂ [38], Ω_arag in the northern South China Sea will continue to decrease by 2100, which may cause serious damage to the coral reef systems in the northern South China Sea.

5. Conclusions and Implications

The distribution and dynamics of Ω_arag in the northern South China Sea were investigated and distinct spatial variations and control mechanisms were found. Surface-water Ω_arag values were all above 2.5 and decreased from the relatively low-salinity area affected by the Pearl River to the offshore area. The intense primary production stimulated by the Pearl River freshwater input resulted in high Ω_arag values near the Pearl River. Bottom-water Ω_arag values were all below 2.5. Ω_arag values below 2 generally appeared at depths below 50 m. Community respiration dominated the low Ω_arag process. As atmospheric CO₂ continues to rise, by 2100, Ω_arag is predicted to decrease by an average of 0.76 units. In the bottom water, Ω_arag is predicted to be lower than 1.7, and even an area of Ω_arag unsaturation may appear. The synergy of further increases in atmospheric CO₂ and community respiration will lead to lower Ω_arag levels in subsurface waters, which may have an important influence on the marine ecosystem.