The increase in population in the Pearl River delta region has increased domestic sewage and industrial discharges, which influences biogeochemical (e.g. carbon dioxide release) and environmental (e.g. oxygen depletion) conditions. Dissolved oxygen and dissolved inorganic carbon concentrations along with salinity, temperature, pH, and chlorophyll a at the surface were collected during seven cruises at eight stations in different months in 2005–2006 to document the seasonal influence of anthropogenic inputs on dissolved oxygen in different areas of Hong Kong waters. Long-term data during 1997–2006 were used to assess air-sea exchange of dissolved oxygen. Near the Pearl River estuary, dissolved oxygen was undersaturated, while partial pressure carbon dioxide was oversaturated, which indicated that Pearl River estuary influenced waters were heterotrophic and represented net sources of carbon dioxide to the atmosphere. In Victoria Harbour, where there was greater sewage effluent influence, however, the degree of dissolved oxygen undersaturation was even more than the Pearl River estuary influenced waters throughout the year. In contrast, the eastern waters, where there was less influence of anthropogenic inputs, showed seasonal variability: undersaturation of dissolved oxygen (∼90%) in the dry season shifting to slight oversaturation (∼105%) in the wet season. The monthly average air-sea influx of oxygen decreased by ∼50 to 200 mmol m−2 d−1 in the wet season relative to that in the dry season in Victoria Harbour, which was coupled with an increase in chlorophyll a from the dry to wet season. These findings are important in understanding why the eutrophication impact from nutrient enrichments in Hong Kong waters is not as severe as one would expect, and also how sewage effluent influences biogeochemical processes of dissolved oxygen and carbon in coastal waters.
Urbanization and anthropogenic nutrients are known to result in eutrophication in many estuarine and coastal waters (Ver et al., 1999; Diaz, 2001), which may change both biogeochemical and environmental processes (such as CO2 and O2 flux) in many coastal waters, such as the Yangtze River, York, Satilla, Scheldt, inner PRE and the Gulf of Mexico (Raymond et al., 1997; Cai and Wang, 1998; Li et al., 2002; Rabalais, 2002; Cai et al., 2004).
In southern China, as the population continues to increase, the inorganic nutrient loading is also increasing through domestic and industrial sewage discharge in the Pearl River estuary (PRE) and adjacent waters. Hong Kong, as an interface where the land, sewage discharge outfall, PRE and ocean meet and interact, is potential zones for transferring terrestrial carbon to the atmosphere and deeper ocean. Previous studies in these areas suggested that these anthropogenic inputs of inorganic nutrients and organic matter have exerted important impacts on variations of phytoplankton biomass, oxygen and CO2 in the PRE and adjacent sediments (Hu et al., 2008; Dai et al., 2009), but only few studies have investigated the influence of these anthropogenic inputs on oxygen and CO2 flux based upon long time series of observations.
CO2 and O2 release and consumption couple with biological activities (e.g. primary production and respiration) in natural seawater, but O2 does not involve in the buffering system like CO2. These similarities and differences provide detailed insight into biological processes as well as gas exchange dynamics (DeGrandpre et al., 1998; Kuss et al., 2006). Therefore, combined examination of O2 and CO2 data can not only help to elucidate the processes of gaseous exchange, but also complex physical and biochemical process (e.g. water mixing, respiration and calcification) in estuarine and coastal waters (DeGrandpre et al., 1998; Kuss et al., 2006). In addition, identifying changes in the annual O2 cycle has important implications of understanding the variability of climate (Sarmiento et al., 1998). For example, due to the variations of temperature, O2 uptake primarily occurs at high latitudes, while both O2 and CO2 may outgas in the tropics (Nevison et al., 2007).
Coupling with CO2 release and O2 drawdown, DOC could be oxidized by micro-organisms and returned to the atmosphere, especially in shallow coastal waters (Kirchman et al., 1991), as well as exported to the deeper ocean by the continental shelf pump. However, little information was available to estimate how much DOC was discharged by the PRE and sewage discharge, and how much DOC that is retained in the Hong Kong waters are respired. By examining the air-sea flux of O2 and CO2, dark community respiration (DCR) and DOC distribution in the Hong Kong waters, we hope to estimate how much DOC retained in the coastal ocean may be returned to the atmosphere or export to the shelf.
In comparison with the open ocean, coastal waters are usually considerably more variable and heterogeneous. Hence, we use long-term data to estimate the variations of DO. The objectives of this study are to: (1) study seasonal variations of O2 and the impacts of anthropogenic inputs on the DO variations in Hong Kong coastal waters; (2) estimate the contributions of air-sea flux to the ambient concentrations of O2 and CO2 in waters; (3) study the DO consumption rats and how much DOC loading may be respired.
Materials and methods
Hong Kong is situated on the eastern side of the PRE, and its marine water area covers about 1,700 km2. The annual average Pearl River discharge is 10,524 m3 s−1, with 20% occurring during the dry season in October to March and 80% during the wet season in April–September, which carries heavy pollution and high nutrient loadings due to local sewage and agricultural inputs (Cai et al., 2004). In addition, Hong Kong waters receive >2 million tons of sewage effluent daily constantly from the local sewage discharge in Victoria Harbor (VH). In 2001, the Hong Kong government implemented the Harbor Area Treatment Scheme (HATS), which collects and treats 70% of the sewage previously discharged into VH. There was strong horizontal advection, as the water residence time is ∼1.5 to 2.5 days in the wet season near the VH (Kuang and Lee, 2004) and ∼3 to 5 days in the PRE (Yin et al., 2003).
Long-term temperature, salinity, wind, chl a and DO data from 1997 to 2007 was obtained from the Hong Kong Environmental Protection Department's (EPD, http://www.epd.gov.hk) monitoring program. During 2005 and 2006, sampling was conducted at 3 stations (Figure 1) and 2 depths: (1) 1 to 4 m below the surface; (2) 2 m above the bottom during 4 cruises in June, July and November 2005 and March 2006. Data from the middle and bottom was not shown in this article due to shallow depths, but could be referred to (Ho et al., 2008). The depths ranged from 11 to 30 m.
Stn 1 represents the waters in the PRE. Stn 5 in the VH are near the sewage discharge site, and eastern waters (EW, Stn 8) is considered to be a reference station since it received little anthropogenic input from the PRE or the local sewage discharge site (Yuan et al., 2010). Water samples were taken using a custom-made 5 L Plexiglas sampler. The vertical profiles of salinity, temperature and chl a were measured with a YSI® 6600 sensor, and chl a was also measured on discrete water samples using the in vitro fluorometric method with acetone extraction and measured on a Turner Designs TD700 fluorometer (Knap et al., 1996).
Dissolved oxygen and apparent oxygen utilization
Dissolved oxygen (DO) was determined by the Winkler titration, as outlined in the JGOFS protocols (Knap et al., 1996). After 4 to 5 volumes of water were allowed to overflow from the 60 ml BOD bottles, Winkler reagents were added. Winkler titrations were carried out in the laboratory with an automated titration apparatus (716 DMS Titrino, Metrohm®) that analyzed the samples with a potentiometric detector to determine the endpoint. DO solubility was calculated according to Benson and Krause (1984). Apparent oxygen utilization (AOU) was calculated as the difference between the saturation oxygen concentration and the observed in situ oxygen concentration. DO solubility was calculated according to Benson and Krause (1984).
Water samples (50 ml) were preserved with saturated HgCl2 (20 µl) for DIC analysis, and measured with a DIC analyzer (AS-C2, Apollo SciTech) (Dickson and Goyet, 1994). pH was measured with an Orion Ross combination glass electrode (Dickson and Goyet, 1994). In order to remove the temperature effect, pCO2 was normalized to an annual average temperature, using the equation of Takahashi et al. (2002).
Air-sea fluxes of CO2 and O2
The CO2 and O2 fluxes across the air-sea interface are calculated by following the one-dimensional stagnant-film model (Wanninkhof, 1992).
Dissolved organic carbon
DOC concentrations were determined after water samples were filtered through 0.7 mm GF/F pre-combusted filters and were acidified with ∼50 μl of 50% H3PO4 acid to pH <2 to drive off the inorganic carbon. DOC concentrations were measured by high-temperature combustion using a Shimadzu TOC analyzer (Knap et al., 1994).
The significance of difference (e.g. seasonal and spatial variations of DO) was assessed by using an analysis of variance followed by a means comparison (t-test). The error bars for the bioassay represent a pooled sample standard deviation of the means. The Pearson test was used to obtain the correlation coefficient and the significance of the correlation. A significance level of 0.05 was used to determine statistical differences. All statistical analyses were performed using SPSS 15.0 for Windows (SPSS Inc.).
Surface salinity, temperature and wind
In the dry season between November and March, salinity was usually ∼33 with no significant difference between Stns 1, 5 and 8 (Figure 2). However, in the wet season between May and October, there was a strong spatial gradient in salinity, indicating the influence of freshwaters from the PRE (Figure 2). Salinity was not significantly different between Stns 5 and 8 (p > 0.05), but much lower at Stn 1 in the wet season (p < 0.05) (Figure 2). The annual water temperature of HK waters varied from 16 to 29°C. The water column temperatures were the highest in summer (23–28°C) and fall (25–27°C), relatively lower in spring (21–25°C) and the lowest in winter (17–19°C).
Wind speed averaged ∼6 m s−1, varying from 4 to 9 m s−1 during the last 10 years, with relatively higher wind speed in the dry season (usually >6 m s−1) than the wet season, especially in December (usually >7 m s−1) (Figure 2). Lower or higher wind speeds have been recorded by Hong Kong Observatory during past 10 years, but was not during the sampling period (data not shown).
Surface chl a
Chl a concentration was low (<5 mg m−3) near the PRE at Stn 1 over all year (Figure 3) and increased away from PRE at Stn 5 and Stn 8, reaching highest chl a (∼15 mg m−3) at summertime. Chl a concentrations were relatively low (<3 mg m−3) at these three stations during the dry season and reached a maximum in July at all stations (Figure 3).
Dissolved oxygen and apparent oxygen utilization in the last 10 years
At Stn 1, DO was not significantly different between before and after 2001 (the period of implementation of the Harbor Area Treatment Scheme by Hong Kong government) (Table 1). However, this sewage treatment scheme reduced sewage discharge and enhanced DO levels at Stn 5 and 8 after 2001 (Table 1).
Monthly, DO at Stn 1 decrease from ∼220 µM in the spring (wet season) to ∼160 µM in October (dry season) (Figure 3). The lowest oxygen often occurred at Stn 5, especially in May and in October (Figure 3), when there was low chl a (∼3 mg m−3) (Figure 3). There was no significant monthly variability in surface DO concentrations at Stn 8 (p < 0.05), where DO was relatively high (averaged ∼220 µM) all year round.
AOU was usually >0 µM at Stn 1 and Stn 5 all year round (Figure 3), indicating that oxygen was generally undersaturated. AOU was much lower at Stn 8 (∼2 µM) than Stn 5 (∼50 µM) (Figure 3), and the average AOU was ∼–9 µM (∼105% oversaturated and a slight source of oxygen to the atmosphere) at Stn 8 in the wet season (Figure 3), which is probably a combined result of air injection and photosynthesis. The average AOU over the whole year was >0 µM at Stn 8 (Figure 3).
Air-sea flux of oxygen in the last 10 years
Influx of O2 at Stn 1 was ∼−102 ± 120 mmol m−2 d−1 and −107 ± 100 mmol m−2 d−1 in the wet and dry season, respectively (Figure 3), showing no significant difference between the wet and dry season. However, influx of O2 at Stn 5 was the highest in the wet and dry season among all stations (Figure 3a, b), and exhibited higher influx in the dry season than the wet season. Average air-sea flux of O2 over the last 10 years was the lowest (∼−14 ± 90 mmol m−2 d−1) at Stn 8 (Figure 3a, b), and showed different direction of monthly average flux between the dry and wet seasons. Average air-sea efflux of O2 was higher after than before 2001 at Stn 5, because of the implementation of the Harbor Area Treatment Scheme by Hong Kong government in 2001.
Parameters during 2005–2006
DOC was relatively high near the PRE and the VH in comparison with the EW in the wet season (Figure 4). DOC reached to ∼300 µM at Stn 1, which was approximately equal to that at Stn 5 in the wet season. In the dry season, DOC was lower at Stn 1 (∼150 µM) than Stn 5 (∼200 µM) (Figure 4).
pCO2 showed clear horizontal variations in the dry season, and exhibited the lowest levels (∼300 to 600 μatm) in the EW and the highest pCO2 (∼500 to 800 μatm) near the VH (Figure 4). In the wet season, surface pCO2 could reach as high as 1,000 μatm at Stn 1 and decreased in areas away from the estuary with lowest (∼300 μatm) pCO2 found at Stn 8 (Figure 4). CO2 was released from waters into the atmosphere in most of the time, since pCO2 usually >400 µatm at Stn 1 and 5.
Regulating factors on long-term oxygen variability
Changes in oxygen content (ΔO2) in the mixed waters is related to biological production, the air-sea exchange of oxygen, and other physical regulating factors (Skjelvan et al., 2001). The relative contribution of each factor has been a widely debated topic (Emerson et al., 1999), which might be variable in different regions.
Biological regulating factors
Since 200 µM DO and ∼10 µM d−1 DCR were common conditions in the water column of Hong Kong, it would take about 20 days to deplete all DO. However, there was strong horizontal advection, as the water residence time is ∼1.5 to 2.5 days in the wet season near the VH (Kuang and Lee 2004) and ∼3 to 5 days in the PRE (Yin et al., 2004). Hence abnormally low or high oxygen waters caused by high planktonic production or respiration could be mixed and replaced quickly by water mass from PRE and offshore waters due to short residence time. Rabouille et al. (2008) also summarized that the lower PRE and Rhône River with a short residence time displays no hypoxia, while the Mississippi system, with the longest bottom water residence time, shows the strongest and largest hypoxia.
Physical regulating factors
The oxygen flux across the air-sea interface is an important physical factor regulating DO concentrations in the ocean. There have been some observations and modeling for air-sea fluxes of O2 in other marine system, which also suggested oxygen inventories significantly increased as a result of air-sea gas exchange of O2 concentrations. For example, in the Ross Sea, the daily air-sea fluxes of O2 and CO2 ranged from −300 to 200 mmol m−2 and −25 to 10 mmol m−2, respectively.
High wind speed could accelerate the air-sea flux of oxygen. For example, in the VH where AOU ∼55 µM was common and average wind speed was ∼6 m s−1 in the last 10 years, air-sea flux of oxygen would increase by ∼50 mmol m−2 d−1 if wind speed increased by 1 m s−1. Hence, relatively high wind speed in the dry season accelerates the influx of oxygen from the atmosphere into waters in comparison with the wet season. In Hong Kong, wind speed rarely dropped to <4 m s−1.
In addition the wind direction also affect the DO levels in Hongkong Waters. In the winter dry season, prevailing northeastern monsoonal winds cause downwelling due to the Ekman transport, and the downwelling results in the shoreward movement of surface offshore waters (Yin, 2002, 2003). Hence the Pearl River estuary and the adjacent coastal waters are dominated by offshore waters which results in low temperature, nutrients, chl a and high DO (Yin, 2003). In the summer wet season, surface waters flow offshore (southwards) due to the southwest monsoon-induced winds which draw the bottom oceanic waters from the continental shelf to the nearshore, and hence high salinity and low DO were present at the bottom (Yin, 2003).
Therefore, although there was active biological respiration in Hong Kong waters, bottom DO in HK waters was generally >3 mg L−1 due to strong physical mixing. This suggested that the eutrophication impacts in HK waters are not as severe as expected for waters with high sewage discharge.
Seasonal variations of air-sea flux of oxygen and carbon dioxide
In temperate marine ecosystems, the seasonal cycle of photosynthesis generally results in a net release of O2 during the late spring and summer, which will be returned to the seawater throughout the remainder of the year, particularly in the fall and early spring (Keeling, 1993). Such seasonal cycle may be less pronounced in tropical region where seasonal cycle of air-sea flux of O2 was rarely examined using long-term data. In our study, it was only in the EW where there was a slight net release of O2 during the late spring and summer. Zhai et al. (2005) recorded that CO2 in the surface water overall was a source to the atmosphere in the coastal waters of the northern South China Sea in 2001 and 2002. However, the long-term observation of CO2 was still lack in coastal waters of the South China Sea.
In contrast, there was a consistent sink of O2 from the atmosphere into seawaters near the PRE all year round. Oxygen influx slightly increased in the wet season relative to the dry season near the PRE, which might be due to that: (1) light limitation for phytoplankton photosynthesis occurred due to high turbidity of the freshwaters in the wet season (Ho et al., 2008), which resulted in community respiration might exceed primary production and hence lower ambient DO concentrations; (2) DO in the freshwaters from the PRE was lower in the wet season than the dry season. The decreased DO concentrations in seawaters accelerated the influx rates of O2 from the atmosphere. In addition to water column process, DIC flux from sediment could be another O2 sink and CO2 source; (3) the increase in temperature may enhance the pCO2 (∼4%°C−1) and reduce DO solubility (∼2%°C−1), while salinity reduces solubility by only <0.7% per unit salinity (Colt, 1984).
Near the VH, where oceanic waters were more dominant and sewage discharge influence was present all year round, the phytoplankton photosynthesis and oxygen release was higher in the wet season relative to the dry season (Ho et al., 2008). Hence, coupling with the increase of chl a, AOU was lower and O2 influx decreased near the VH.
Previous study estimated new production (new input of nitrogen) using the sum of the positive seasonal fluxes (or efflux) of oxygen (Keeling, 1993), which might not be directly applied for Hong Kong waters, since air-sea flux of oxygen was usually negative (or influx) in Hong Kong waters except in the EW in the wet season (Figure 3). This is probably a result of benthic O2 consumption of external OC input from sewage effluent. However, we can roughly estimate the seasonal increase of oxygen production using the difference of seasonal oxygen fluxes between the wet and dry season near the VH and EW. If ratios of CO2: O2 (∼1.1 mol CO2 per 1 mol O2) was used to convert moles of oxygen to moles of carbon produced, oxygen production in the wet season would be equivalent to an OC production of ∼50 to 200 mmol m−2 d−1 higher than the dry season near the VH and EW, which was lower but in the same magnitude as the 14C based on the measurement of primary production by Ho et al. (2008).
The DO dynamics in HK waters are driven by a combination of various factors including PRE estuarine waters, local sewage effluent, coastal/shelf seawater, monsoon winds, and biological utilization. Low DO and high air-sea DO influx in summer was due to the high rainfall and river discharge and prevailing monsoon winds which blows the surface waters of the PRE into Hong Kong waters. Downwelling occurs in winter due to the northeastern monsoon winds and results in the intrusion of high-DO coastal/shelf seawater. Continuous year round discharge of sewage effluent resulted in low DO at VH and its vicinity. Lower DO was observed in summer due to the period of maximal biological utilization due to high temperatures and respiration. In winter, strong tidal and wind mixing, rapid flushing and low temperatures resulted in high DO in Hong Kong waters. Due to the implementation of the sewage treatment scheme, DO increased at Stns 5 and 8 because of the reduced discharge of sewage effluent after 2001.
This research was supported by grants from NSFC Project (31370499, 41106107, 31370500, 40676074), National Key Technology Support Program (2014BAC01B03), and Science and Technology Planning Project of Guangdong Province, China(2014B030301064). We appreciate the contributions of K. Yin and P. J. Harrison to this work.