Variations in nitrate isotope composition of ...

MPB-07893; No of Pages 10 Marine Pollution Bulletin xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Variations in nitrate isotope composition of wastewater effluents by treatment type in Hong Kong A. Archana a,b, Luo Li c, Kao Shuh-Ji c, Benoit Thibodeau d, David M. Baker a,b,⁎ a

School of Biological Sciences, University of Hong Kong, Pok Fu Lam, Hong Kong, China Swire Institute of Marine Science, University of Hong Kong, Cape D'Aguilar, Hong Kong, China State Key Laboratory of Marine Environmental Science, Xiamen University, PR China d Department of Earth Sciences, University of Hong Kong, Pok Fu Lam, Hong Kong, China b c

a r t i c l e

i n f o

Article history: Received 2 December 2015 Received in revised form 11 July 2016 Accepted 13 July 2016 Available online xxxx Keywords: Hong Kong Nitrogen pollution Sewage Mass balance Wastewater treatment Stable isotope

a b s t r a c t Stable isotopes (δ15N, δ18O) can serve as tracers for sources of nitrogen in the receiving environment. Hong Kong discharges ~3 × 106 m3 d−1 of treated wastewater into the ocean from 68 facilities implementing preliminary to tertiary treatment. We sampled treated sewage from 18 plants across 5 treatment types and examined receiving seawater from northeast Hong Kong. We analyzed nitrate and nitrite (NO3− + NO− 2 , hereafter NOx) ammonium + 15 18 (NH+ 4 ), phosphate (PO4 ) concentrations and δ NNOx, δ ONOx. Sewage effluents contained high mean nutrient −1 −1 , PO+ ) with some indication of concentrations (NO3− = 260 μmol L−1, NH+ 4 = 1400 μmol L 4 = 50 μmol L nitrogen removal in advanced treatment types. Mean δ15NNOx of sewage effluents from all plants and treatment types (12‰) was higher than natural sources and varied spatially and seasonally. There was no overall effect of sewage treatment type on δ15NNOx. A mass balance model indicated that sewage (N 68%) remains a dominant source of nitrate pollution in seawater in Tolo Harbor. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Sewage is one of the largest contributors of anthropogenic pollution and sewage treatment plants are important for modulating nutrient concentrations in receiving marine and estuarine waters. Elevated nutrients (nitrogen and phosphorous) from wastewater induce eutrophication, a condition that alters the coastal ocean balance by changing food web structure, increasing primary production and the prevalence and severity of disease (Howarth et al., 2000; Baker et al., 2007; Wear and Vega Thurber, 2015). Numerous studies have documented the indirect effects of sewage pollution on the growth and survival of biologically and structurally complex ecosystems such as coral reefs, with consistent negative impacts on calcification, community structure and biodiversity (Walker and Ormond, 1982; Rodriguez, 1981; Baker et al., 2010; Smith et al., 1981; Baker et al., 2013). Over the last fifty years, several studies have reported a concomitant decline in water quality and macrofauna, coupled with the frequent incidence of red tides and elevated levels of heavy metals, antibiotics and faecal bacteria in Hong Kong's marine environment (Morton, 1988; Cope and Morton, 1988; Scott, 1990; Ng and Morton, 2003; Fabricus

⁎ Corresponding author at: 3S-13, Kadoorie Biological Sciences Building, School of Biological Sciences, University of Hong Kong, Pok Fu Lam, Hong Kong, China. E-mail address: [email protected] (D.M. Baker).

and McCorry, 2006; Connell et al., 1998; Yeung, 2000; Yin and Harrison, 2007). A fine example of this is Tolo Harbor in northeast Hong Kong (Fig. 1; Yim et al., 1982; Scott and Cope, 1982). Hard coral species richness at the mouth of the Tolo Channel decreased by 27% from 1986 to 2000 (McCorry and Blackmore, 1998). The progressive decline of water quality and marine life in the harbor owing to residential development has led scientists to refer to it as “Hong Kong's first marine disaster” (Morton, 1988). As a measure to improve water quality, the Hong Kong Government (HKSAR) re-routed all wastewater discharge outfalls from Tolo Harbor to Victoria Harbor in 1996. However, a recent study by Luo et al. (2014) confirmed that severe nutrient pollution persists in Tolo Harbor despite the removal of point-source inputs of wastewater in the region. Therefore, the source of this pollution is unclear and warrants further investigation. Another example is the historical and chronic discharges of sewage effluents into Hong Kong's urbanized Victoria Harbor at the rate of over 2 million m3 of sewage per day. Close to 75% of this heavily polluted water was discharged without any treatment until 2001 when the Harbor Area Treatment Scheme (HATS) came into effect. Despite these efforts, the effects of eutrophication are ever-present (Xu et al., 2011, 2014; HKSAR DSD, 2014a), signaling that stormwater, untreated wastewater from septic tanks and ferries, improper or damaged sewerage connections, and other nonpoint sources such as surface run-off are inhibiting mitigation efforts. To date, the sources of nutrients that contribute to persistent eutrophic conditions remain unknown.

http://dx.doi.org/10.1016/j.marpolbul.2016.07.019 0025-326X/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article as: Archana, A., et al., Variations in nitrate isotope composition of wastewater effluents by treatment type in Hong Kong, Marine Pollution Bulletin (2016), http://dx.doi.org/10.1016/j.marpolbul.2016.07.019

2

A. Archana et al. / Marine Pollution Bulletin xxx (2016) xxx–xxx

Fig. 1. Study site – HKSAR and sampling locations. 18 sewage treatment plants and 3 seawater sampling locations were selected for this study. Sewage treatment plants were classified into 5 treatment types according to the process employed.

Today, there is a well-defined gradient in water quality from the west to the east of Hong Kong, particularly obvious for total inorganic − nitrogen (TIN), ammonium (NH+ 4 ) and nitrate (NO3 ) owing to the influence of the Pearl River estuary carrying wastewater, agricultural inputs, and industrial pollutants from the heavily developed Guangdong Province. Urbanized population centers correspond with high TIN loadings. The HKSAR has several management practices involving water quality, including set water quality objectives, beach and open-water quality monitoring programs, long-term phytoplankton monitoring, zoned mariculture areas, three marine parks, one marine reserve, and intensive investment in wastewater infrastructure, including HATS (HK EPD, 2014). However, these management efforts rely on monitoring the concentrations of nutrients in seawater and wastewater, respectively, but cannot differentiate natural sources of nitrogen (e.g. fixation and mineralization) from anthropogenic inputs (e.g. fertilizers and sewage) originating from the Pearl River Delta region. To our knowledge, there has been no study that has successfully identified the sources of the near-shore water quality problems in Hong Kong. Stable isotope analysis of nitrogen (δ15N) fills this gap, serving as a tracer for sources and sinks of nitrogen in the environment (Heaton, 1986; Kendall et al., 2010) Although nitrogen is rapidly diluted, diffused and transported by ocean currents, source δ15N values are preserved in primary producers which can define an isotopic baseline for the marine food web. Moreover, δ15N measurements are especially useful in pollution monitoring studies because each source of nitrogen has a unique δ15N composition associated with it. For example, sewage has very different δ15N signatures relative to natural sources. δ15N of raw sewage is usually ~6‰ (Kendall et al., 2007), while that of treated sewage is estimated to be between 10 and 20‰ (McClelland and Valiela, 1998). Such enrichment in δ15N values arises from microbially mediated nitrification and denitrification reactions with fractionation factors (ε) of 10‰ to 40‰ that are dependent on the type of wastewater treatment employed (Mariotti et al., 1984; Risk et al., 2009). Other examples of

nitrogen sources that have unique δ15N values are synthetic fertilizer (~ 3‰), combustion (~ 1‰), and precipitation (~− 7‰; Table 1; Fry, 2006). Several studies have successfully used stable isotopes to characterize nitrogen pollution from anthropogenic inputs in sensitive coastal marine habitats around the world (Risk, 2014). For example, Magni et al. (2013) studied δ15N and δ13C variations in two dominant bivalve species along the European coast, demonstrating their effectiveness as bio-indicators of sewage pollution. In the tropics, studies of corals and gorgonians have shown that benthic invertebrates take up sewagederived nitrogen (Baker et al., 2010, 2013). Although the majority of studies focus on biological integrators, only a few have measured the sources of pollution directly, such as the δ15N of dissolved inorganic nitrogen (DIN) in wastewater effluents. Moreover, there is relatively little information available on the influence of sewage treatment type on δ15N values of wastewater effluents (Costanzo et al., 2005; Sebilo et al., 2006; Gaston et al., 2004; Table 2). Of the studies that exist, many focus on effluent from a single specific treatment type – preliminary, primary, secondary or tertiary. To our knowledge, there has been no study that provides δ15N values for wastewater effluent from five different treatment methodologies (CEPT-Chemically Enhanced Primary Treatment, preliminary, primary, secondary, tertiary). Moreover, there has been no study done yet on nitrogen source Table 1 δ15N values of different sources of nitrogen (Fry, 2006; Kendall et al., 2007). Sources

δ15N (‰)

N fixation Fertilizer Atmospheric deposition Marine nitrate Sewage

−2 to +2 −7 to +4 −15 to +3 +4 to +6 +6 to +9


A. Archana et al. / Marine Pollution Bulletin xxx (2016) xxx–xxx Table 2 δ15N values of wastewater effluents by treatment type from various studies (Rogers, 1999, Leavitt et al., 2006; Onodera et al., 2015; Piola et al., 2006; Ochoa-Izaguirre and SotoJimenez, 2015, Tucker et al., 1999; Sebilo et al., 2006). Literature

Our study 15

15

Treatment type

DIN (‰)

δ NNH4 (‰)

δ NNO3 (‰)

δ15NNOx (‰)

Preliminary Primary CEPT Secondary Tertiary

2.0 to 2.6 10.3 to 15.8 ? 11.7 to 14.2 12.5 to 16.2

? 7.2 to 16.4 ? 6.5 to 14.7 ?

? 6.3 to 10.1 ? ? ?

11.5 ± 3.1 14.8 ± 3.9 10.6 ± 4.9 12.5 ± 4.3 92.7 ± 83.9

determinations in Hong Kong seawaters – a gap that is important to fill for effective environmental policy decisions in the region. To address this gap, we examined the stable isotope ratios of wastewater effluents and receiving marine waters to test the hypotheses that δ15N varies with treatment type and sewage-derived nitrogen is the dominant source of nutrient pollution in the coastal marine environment in Hong Kong. We ask the following: What is the spatial and temporal variability of nutrient concentrations and δ15N of seawater and wastewater effluents? How do nutrient concentrations and δ15N vary across wastewater treatment type? What is the dominant source of nutrient pollution in seawater? The answers to these questions are crucial for developing efficient and cost-effective marine environmental monitoring methods in densely populated coastal cities.

2. Materials & methods 2.1. Study site and sampling methodology A megacity such as Hong Kong with a population density of ~7000 people per km2 is an ideal study site because it houses five sewage treatment types from preliminary to tertiary treatment in an area of only ~1650 km2. Multiple environmental stressors threaten the biodiversity and ecosystem resilience of the coastal marine environment and issues regarding water pollution from local and regional, point and non-point sources are repeatedly raised for investigation. HKSAR Drainage Services Department (DSD) currently operates 239 sewage treatment facilities including 68 sewage treatment plants (STP) and 172 sewage pumping stations in Hong Kong, Kowloon, New Territories and Lantau and Outlying Islands (HKSAR DSD, 2014a). The STPs can be divided into five categories based on their pollutant removal efficiencies, annual volume and treatment type – preliminary, primary, CEPT, secondary, and tertiary. Through a 12-month sampling effort across two seasons (dry, wet), wastewater effluents were sampled from 18 sewage treatment plants across five treatment types. Effluents from five preliminary, two primary, three CEPT, seven secondary and one tertiary treatment plant were sampled (Fig. 1). Grab samples (n = 3 per treatment facility) were collected in 500 ml acid-washed Nalgene bottles. All samples were collected between 1100 and 1200 h (Refer to Appendix for sampling dates). Water samples were filtered through GF/F filter (0.7 μm pore size) and immediately stored at −20 °C for nutrient concentration measurements and stable isotope analysis. Rapid urbanization and coastal development in Victoria Harbor (central) and Tolo Harbor (northeast) has resulted in a localized effect of nutrient pollution in surrounding waters. We collected seawater samples from Tolo Channel which includes receiving waters from STP outfalls. In total, 3 sampling sites were chosen (Centre Island-CT, Che Lei Pai-CLP, Port Island-TWP; Fig. 1) and samples (n = 3 per site) were collected across two seasons (dry, wet). We also analyzed freshwater samples from the Hoi Ha River, a small river that flows into Tolo channel, and precipitation from two extreme rainfall events. Seawater reference data for dissolved oxygen (DO), salinity, and chlorophyll-a data were used from the Environmental Protection Department's (HK EPD) water quality database (1986–2012).

3

2.2. Nutrient concentration Inorganic nutrients [ammonium, nitrate and phosphate] were analyzed using a Lachat Flow Injection Analyser QuikChem® 3000 according to QuickChem Method (Ammonia: 10-107-06-1-A; Nitrate and Nitrite: 31-107-04-1-A; Phosphate: 31-115-01-3-A). Deionized water was used as a blank. Ammonium chloride, potassium nitrate and sodium nitrite were used to prepare standards of known concentrations (0.25 μmol L−1, 1.5 μmol L−1, 2.5 μmol L−1, 5 μmol L−1). All samples were diluted prior to concentration measurements. Resulting measurements had a precision of ± 0.1 μmol L− 1 for ammonium and ±0.3 μmol L−1 for nitrate, nitrite and phosphate. 2.3. Stable isotope analysis δ15NNOx, δ18ONOx of all water samples was determined using the bacterial denitrifier method (Sigman et al., 2001; Casciotti et al., 2002). The principle of the method is the measurement of the isotopic composition of nitrous oxide (N2O) following the conversion of nitrate and nitrite (NOx) to N2O by bacteria that lack N2O reductase. Online purification, cryogenic trapping and chromatographic separation were performed online with a PreCon system. We used two international standards USGS34: δ15NNO3 = −1.8‰ ± 0.2, δ18ONO3 = −28‰ ± 0.2 and IAEAN3: δ15NNO3 = + 4.7‰ ± 0.1, δ18ONO3 = + 26‰ ± 0.7 (Gonfiantini et al., 1995) and two in-house laboratory working standards – NaNO3 (δ15NNO3 = + 5.1‰ ± 0.1, δ18ONO3 = + 26‰ ± 0.5) and NaNO3 (δ15NNO3 = +14‰ ± 0.2). Nitrogen isotope ratio measurements were made at the State Key Laboratory of Marine Environmental Science, Xiamen University by a GasBench II coupled to a continuous flow isotope ratio mass spectrometer (IRMS, Thermo Delta V Advantage) with an overall analytical precision of b 0.2‰ for δ15NNO3 and b0.4‰ for δ18ONO3. 2.4. Statistical analysis Data were screened for normality using Shapiro-Wilk W test and homoscedasticity using Levene's test. One-way ANOVA was used to evaluate the effect of treatment type on δ15N. Post hoc means comparisons to test for significant differences in δ15N values between treatment types were conducted using Student's t-test and Tukey's test. To understand variability and enrichment in δ15NNOx, δ18ONOx values and nutrient concentrations, Pearson's correlation (α = 0.05) was conducted. To calculate the proportion of sources in the seawater samples, we assumed that the three main sources of nitrogen in Hong Kong's northeastern marine waters are sewage, atmospheric deposition and fertilizer runoff from rivers. A dual isotope, three-end-member linear mixing model was developed from the following mass balance equations: δ18 ONOx ¼ f sewage δ18 ONOx sewage þ f atmosphericdeposition δ18 ONOx atmosphericdeposition þ f run−off δ18 ONOx run−off δ15 N NOx ¼ f sewage δ15 NNOx sewage þ f atmosphericdeposition δ15 NNOx atmosphericdeposition þ f run−off δ15 NNOx run−off 1 ¼ f sewage þf atmosphericdeposition þf run−off where the subscript refers to the isotopic composition of nitrate from three different sources (sewage, atmospheric deposition and run-off) and f refers to the respective partitioning coefficients. All calculations were done according to Phillips and Gregg (2001). All statistical analyses were conducted in JMP 10.0 (SAS Institute) and results are reported with ±standard deviation.


4


3. Results All wastewater samples were characterized with high mean concentrations of nitrate and nitrite (260 μmol L− 1), ammonium (1400 μmol L− 1), and phosphate (50 μmol L−1; Table 3). Overall, mean inorganic nitrogen concentrations in the dry season (NOx = −1 390 μmol L−1; NH+ ) were significantly higher (two 4 = 1900 μmol L sample t-test; n = 74; p b 0.05; Fig. 2) than values measured in the −1 wet season (NOx = 150 μmol L−1; NH+ ). 4 = 1000 μmol L 15 Mean δ NNOx values of wastewater effluents ranged from 6.7‰ to 22.9‰ and δ18ONOx values ranged from −5.4‰ to 16.1‰ across the 18 sampled sewage treatment facilities with the exception of Ngong Ping Treatment Works, a tertiary wastewater plant (δ15NNOx = 169.2 ± 4.6‰; δ18ONOx = 170.3 ± 5.2‰). Significant seasonal variability in δ15NNOx and δ18ONOx was observed (two sample t-test; n = 78; p b 0.05). Median δ15NNOx of effluents sampled during the dry and wet seasons were 12.7‰ and 10.9‰ respectively, while median δ18ONOx during dry and wet seasons were 4.8‰ and 3.9‰ respectively. Wastewater effluent mean nitrate, nitrite and ammonium concentrations varied significantly by treatment type (ammonium - one-way ANOVA; n = 79; F(4,75) = 15.4 p b 0.001; nitrate and nitrite - one-way ANOVA; n = 79; F(4,75) = 3.5 p b 0.001; Fig. 3). Wastewater δ15N also varied significantly with treatment type but only in the dry season (Table 3; Fig. 4) with least enrichment in CEPT (δ15NNOx = 10.7 ± 4.9‰ and δ18ONOx = − 2.1 ± 3.7‰) and most enrichment in tertiary treatment (δ15NNOx = 169.2 ± 4.6‰; δ18ONOx = 170.3 ± 5.2‰; oneway ANOVA of δ15NNOx, n = 34 F(4,30) = 1235.7, p b 0.001). When tertiary treatment was excluded from the dataset, there was no variability by treatment type (one-way ANOVA of δ15NNOx, n = 34; F(3,31) = 0.9844, p = 0.413; Fig. 5). In the wet season, although there was variability observed between treatment types, there was no overall significant effect of treatment type on δ15NNOx (one-way ANOVA of δ15NNOx, n = 42, F(4,38) = 2.6, p = 0.052; Fig. 6). Samples from secondary treatment plants showed the most variability (7.3‰ to 19.9‰ across 7 treatment facilities; one-way ANOVA of δ15NNOx; n = 21, F(6,15) = 34.9, p b 0.001) relative to other treatment types. The dataset was also characterized by a significant positive correlation between δ15NNOx and δ18ONOx (R2 = 0.60, p b 0.0001; Fig. 7) with m = 0.6. All seawater samples were characterized by mean nitrate and nitrite (3.3 μmol L− 1), ammonium (1.6 μmol L− 1), and phosphate (0.7 μmol L−1) concentrations. Inorganic nitrogen concentrations decreased with distance from inshore (Centre Island) to outside Tolo Channel (Port Island; Fig. 8). Conversely, δ15NNOx of seawater increased with distance from inshore to the open ocean (Fig. 8). Mean seawater δ15NNOx (11.7‰) closely resembled that of treated wastewater (~ 12.2‰) with no significant seasonal variability (13‰ and 11‰ in dry and wet seasons, respectively; p = 0.76). Mean seawater δ18ONOx (23‰) showed spatial variability with no obvious trend as in the case of δ15NNOx. However, significant seasonal variability was observed in seawater δ18ONOx values (two sample t-test; n = 17; p b 0.05). There was a significant negative correlation between δ15NNOx and [NO− 3 ] (n = 22, R = − 0.46, R2 = 0.21, p b 0.05; Fig. 9) and a significant but weak positive correlation between δ15NNOx and δ18ONOx (n = 22, R = 0.11, R2 = 0.01, p b 0.0001, m = 0.4; Fig. 9). Seawater δ15NNOx showed a strong negative correlation (n = 18, r = −0.69, R2 = 0.48, p b 0.05), with DO and no significant correlation with salinity and chlorophyll-a

+ Fig. 2. NO− 3 , NH4 seasonal variability for wastewater effluents. Mean inorganic nitrogen concentrations were relatively higher in the dry season than in the wet season.

content. Mean seawater NO3− + NO− 2 concentrations revealed strong correlation with DO (n = 18, R = 0.71, R2 = 0.51, p b 0.05), salinity (n = 18, R = −0.80, R2 = 0.64, p b 0.05), and chlorophyll-a (n = 18, R = 0.68, R2 = 0.46, p b 0.05). Freshwater samples from the Hoi Ha River had mean concentrations of nitrate and nitrite = 19 μmol L−1; ammonium = 2.3 μmol L− 1; phosphate = 0.2 μmol L− 1; δ15NNOx = 1.8‰; δ18ONO3 = 13.7‰, while precipitation from two extreme rainfall events had mean concentrations of nitrate and nitrite = 23 μmol L−1; ammonium = 21 μmol L− 1; phosphate = 0.6 μmol L− 1; δ15NNOx = 3.0‰; δ18ONO3 = 55.9‰. A mass balance model using dual nitrate isotopes and three nitrate sources (sewage, atmospheric deposition, runoff) showed that Tolo Harbor is dominated by sewage (N 68%; Table 4). 4. Discussion Depending on the level of treatment employed in a wastewater treatment facility, organic nitrogen from sewage advances through one or all of these steps; hydrolysis and remineralization to ammonia, oxidation to nitrite by Nitrosomonas and then to nitrate and nitrogen gas by Nitrobacter. Until the late 1990s only 10% of wastewater received secondary treatment that incorporated biological nitrogen removal, 25% received preliminary treatment (screening only) and the majority of Hong Kong's sewage was discharged through numerous outfalls as untreated effluent (Xu et al., 2014). In 2001, the HKSAR successfully commissioned the Harbor Area Treatment Scheme (HATS), where a ~ 21 km long pipeline now transports sewage from all preliminary STPs and a few secondary STPs to a centralized CEPT plant located at Stonecutter's Island. Stonecutters Island handles ~ 2 × 106 m3 d−1 of wastewater (HATS Stage 1A, 1B, 2A; HKSAR DSD, 2014b). As part of HATS Stages 1A, 1B and 2A, the majority of Hong Kong's sewage is now discharged after treatment into Victoria Harbor through multiple outfalls from Stonecutters Island. Since 1996, the HKSAR re-routed all sewage outfalls in Tolo Harbor to Victoria Harbor leaving only an emergency bypass that can actively flush secondary treated sewage into Tolo Channel during periods of heavy rainfall. However, long-term water quality data from HK EPD (1986–2012) shows that northeastern

Table 3 − + + 15 18 Mean NO3− + NO− 2 , NO2 , NH4 , PO4 concentrations and δ NNOx, δ ONOx signatures of wastewater effluents by treatment type in Hong Kong. Treatment type

−1 ) NO3− + NO− 2 (μmol L

−1 NO− ) 2 (μmol L

−1 NH+ ) 4 (μmol L

−1 PO− ) 4 (μmol L

NH4/(NO3 + NO− 2 )

δ15NNOx (‰)

δ18ONOx (‰)

Preliminary Primary CEPT Secondary Tertiary

47 ± 44 3±2 890 ± 500 310 ± 250 290 ± 280

5±9 1±1 2±2 5±9 15 ± 19

1800 ± 1000 1600 ± 900 2500 ± 1300 250 ± 500 1700 ± 1800

62 ± 28 40 ± 20 22 ± 19 35 ± 20 100 ± 35

39 550 2.8 0.9 5.8

11.5 ± 3.1 14.8 ± 3.9 10.6 ± 4.9 12.5 ± 4.3 90.7 ± 83.9

4.9 ± 4.2 8.6 ± 3.4 −2.1 ± 3.6 3.8 ± 2.4 87.7 ± 90.6



Fig. 3. Mean NO3−, NH4+ (±standard deviations) concentrations in μmol L−1 from wastewater effluents according to treatment type. Connecting letters indicate significant difference between treatment types (α = 0.05).

seawaters of Hong Kong still have high concentrations of nutrients despite minimal to no sewage being discharged in the region (mean Tolo −1 Harbor inshore readings: NO3− = 2.8 μmol L−1; NH+ ; 4 = 15 μmol L −1 + PO4 = 0.4 μmol L ). This near-shore pollution has been attributed to discharges originating from various sources such as stormwater, untreated wastewater from septic tanks and ferries, improper or damaged sewerage connections, and other non-point sources such as surface runoff and submarine groundwater discharges (Luo et al., 2014).

4.1. Spatial and temporal variability of wastewater effluents Spatial and temporal differences were observed in nutrient concentrations and nitrate isotopic compositions of treated wastewater effluent samples (Table 3). Several studies have reported varied effects of seasonality and flow rate on effluent quality from municipal wastewater treatment plants (Hashimoto et al., 2014; van Vliet and Zwolsman, 2008). Dry seasons often result in high nutrient concentrations because there is minimal to no dilution of point source effluents by precipitation (Adonadaga, 2014). We studied δ15NNOx because nitrate is relatively more abundant, long-lived and less volatile among other forms of inorganic nitrogen in the receiving marine environment. In our study, a possible explanation for the lower inorganic nitrogen concentrations and δ15NNOx of wastewater effluents in the wet season is the effect of dilution from the mixing of nitrate from various sources including rainwater nitrate and terrestrial runoff with sewage.

5

To further explain the variability seen in wastewater effluent samples, we analyzed the relationship between δ15NNOx and δ18ONOx. Thibodeau et al. (2013) demonstrated the use of an N and O isotope model to calculate the proportion of N loading in the St. Lawrence River. Moreover, Kendall et al. (2007) observed that, when nitrogen removal by denitrification occurs, the slope of the linear fit between δ18ONOx and δ15NNOx ranges from 0.5 to 0.7. We applied this model to our data, which showed a significant positive linear relationship between δ15NNOx and δ18ONOx (Fig. 7). As the slope in our study falls within the specified range (m = 0.6) this suggests increasing denitrification with increasing level of treatment type. One of the objectives of wastewater treatment is nitrogen removal and our values affirm that this occurs in the sewage treatment plants that were sampled for this study. Therefore, it can be concluded that both δ15NNOx and δ18ONOx values may serve as potential indicators for nitrogen removal efficiency in a wastewater treatment process. 4.2. Variability across wastewater treatment type We propose that the variation in ammonium, nitrate and nitrite and δ15NNOx across treatment type (Figs. 3–6) results from several factors that may affect effluent nitrogen source composition independently or cumulatively – (a) microbial activity in the activated sludge process (will convert ammonium to nitrate and enrich effluent δ15NNOx by 10 to 40‰), (b) type of wastewater influent - domestic sewage, livestock waste, landfill leachate, etc., (c) flow rate or volume handled and (d) ammonia volatilization. We observed that mean NO− 3 concentrations (Fig. 3) increased while mean ammonium concentrations decreased with increasing level of treatment type. This suggests that with the incorporation of biological nitrogen removal processes within a treatment type (such as secondary and tertiary), the resulting treated + wastewater effluent is likely to contain more NO− 3 than NH4 . Mean δ15NNOx values also showed variability by treatment type (Figs. 4–6). δ15NNOx values of wastewater effluents from primary STPs were higher than what has been reported in previous studies (Tables 2 and 3). During preliminary and primary treatment, organic nitrogen is converted to inorganic nitrogen (i.e. ammonium mainly) and in the absence of an activated sludge process, which is responsible for further conversion to nitrate, the nitrogen in the resulting effluent is likely to remain as ammonium (Sebilo et al., 2006) with low δ15NNOx values closely resembling raw sewage δ15NNO3. Contrary to our predictions we observed an increase in δ15NNOx values of wastewater effluents from primary STPs, which we suspect to be associated with the ongoing upgrade of the two primary treatment plants to secondary treatment through sand filtration. Sand filtration incorporates biological, physical and

Fig. 4. Mean δ15NNOx (±standard deviations) variability by treatment type for wastewater effluents in the dry season. Connecting letters indicate significant difference between treatment types (α = 0.05).


6


Fig. 5. Mean δ15NNOx (±standard deviations) variability by treatment type for wastewater effluents in the dry season excluding tertiary treatment. Connecting letters indicate no significant difference between treatment types.

chemical processes to act as a pathogen and particle filter (Lagenbach et al., 2009). Straining and adsorption in filters are responsible for bacteria retention, which result in microbial nitrification and denitrification of the sewage that passes through. Microbial nitrification (conversion of ammonium to nitrate) and denitrification (nitrate to nitrous oxide to nitrogen) have different fractionation factors owing to the involvement of various types of aerobic bacteria in the reaction pathways. The product of nitrification is depleted in 15N relative to the reactant ammonium. Similarly the product of denitrification is depleted in 15N relative to the reactant nitrate (Flanagan et al., 2004). The products are isotopically lighter than the reactants because of mass-dependent fractionation where lighter nitrogen isotopes react preferentially leaving behind a pool of isotopically heavy nitrogen (enriched δ15NNOx wastewater effluents). Treated effluent collected from Stonecutters Island (CEPT plant, part of HATS) recorded the lowest δ15NNOx signal (8.1‰). The influent is primarily domestic sewage and the measured effluent δ15NNOx closely resembles effluent that goes through no nitrogen removal processes. Recently, HKSAR Government proposed an upgrade of the CEPT plant to secondary treatment, which is currently under review. Secondary treatment is important because it provides an opportunity for increased duration of microbial nitrification and denitrification favoring the conversion of ammonium to nitrogen. Choi et al. (2009) quantified the marine environmental risk using an integrated stochastic risk assessment

model of sewage treatment upgrades. With increase in pollution loads, the environmental risk went from 1.18 in preliminary treatment, and 1.41 in CEPT to below 0.14 with the incorporation of biological treatment. As data from our study suggest that CEPT shows little isotopic evidence for nitrogen removal, this treatment type warrants priority for future secondary treatment upgrades. Secondary and tertiary treatment processes incorporate a biological nitrogen removal step with the addition of an aeration tank and a secondary clarifier with air diffusers to the primary treatment process. Ammonia oxidizing bacteria (AOB), nitrite oxidizers and denitrifiers convert ammonium to nitrate and nitrate to nitrogen as a result of which enriched δ15NNO3 is expected (Khayat et al., 2006). The DO provided helps maintain a viable population of microorganisms that are needed for the conversion of organic nitrogen into its inorganic byproducts. Sebilo et al. (2006) characterized completely nitrified wastewater effluent (transformation of the entire pool of ammonium to nitrate) having δ15NNO3 values between 7 and 12‰, and partially nitrified effluents (incomplete transformation of ammonium to nitrate) by δ15NNO3 b 7‰. In our study, most variability was observed within secondary treatment plants. A possible explanation for this is the role played by ammonium in influencing δ15NNOx of the treated effluent when there is only a partial conversion of ammonium to nitrate. Other factors are also likely to affect the isotopic composition. For example, there is a large variation in annual volume handled by the sampled

Fig. 6. Mean δ15NNOx (±standard deviations) variability by treatment type for wastewater effluents in the wet season. Connecting letters indicate significant difference between treatment types (α = 0.05).



7

Fig. 7. Mean wastewater δ15NNOx vs. δ18ONOx. δ15NNOx enrichment has been plotted as a function of δ18ONOx in wastewater effluent samples across 5 treatment types between the wet and dry seasons in Hong Kong. S* indicates recent upgrade made from primary to secondary treatment.

sites, ranging from 2.5 × 106 m3 to 31.5 × 106 m3. There is also variation in the type of influent handled. For instance, influent in Tai Po sewage treatment plant is dominated by landfill leachate while influent in Shek Wu Hui sewage treatment plant is dominated by livestock waste. Tertiary treatment presented the largest δ15NNOx = 170 ± 4.6‰ and δ18ONOx = 170 ± 5.2‰. However, this enrichment was noted only in the dry season and not in the wet season, which had δ15NNOx = 16 ± 0.1‰ and δ18ONOx = 5.1 ± 0.2‰. We hypothesize that these extreme enrichments are likely to be due to the treatment process itself and the use of several tertiary filters that gradually enriches the nitrate isotopic signatures of the treated sewage. However, as this hypothesis only applies to the values observed in the dry season and not in the wet season, and given that there is only one tertiary treatment facility in Hong Kong (Ngong Ping Sewage Treatment Plant), we propose that long-term sampling will help us better understand the nitrogen dynamics from this treatment type.

18 15 Fig. 9. Mean Seawater δ15NNOx vs. NO− 3 and δ ONOx (± standard deviations). δ NNOx 18 enrichment has been plotted as a function of NO− 3 concentration and δ O in seawater samples in Tolo Harbor. Dashed lines indicate typical expected slopes for data resulting from denitrification of nitrate. Solid line represents a linear function fit to all data. Weak positive slope between δ15NNOx and δ18ONOx (0.4) suggests mixing of other nitrate sources.

4.3. Nitrogen source partitioning in seawater To understand the effect of nitrogen source partitioning on the receiving marine environment, we analyzed seawater samples from northeastern Hong Kong. We observed elevated nutrient concentrations and nitrate isotopic compositions of seawater samples in Tolo Channel with spatial and temporal differences (Fig. 8). Mean δ15NNOx values of seawater were over three times higher (12‰) than naturally occurring marine nitrate δ15NNOx (4‰ to 6‰; Somes et al., 2010). We contend that treated sewage is the cause. However, since 1996, there are no longer government operated sewage discharges into Tolo Harbor. These findings led us to question the source of nutrients, particularly nitrogen that is still found in the seawater in Tolo. Luo et al. (2014) in a study on Tolo Harbor confirmed that there is a persistent source of nutrient pollution in the region and correlated submarine groundwater discharges (SGD) with eutrophication. Data from our study (significant negative correlation between δ15NNOx and NO− 3 ; Fig. 9) illustrate that there is a strong gradient of decreasing nutrient concentrations and increasing isotope enrichment towards Port Island and Mirs Bay. Such patterns are consistent with microbial transformations along this spatial gradient. However, the significant but weak positive linear relationship between δ15NNOx and δ18ONOx (m = 0.4, Fig. 9) of seawater samples from Tolo Harbor is not in complete agreement with the hypothesis of

Table 4 Nitrogen source partitioning of seawater in Hong Kong.

CT CLP TWP Tolo Channela Fig. 8. Mean nutrient concentrations, δ15NNOx and δ18ONOx (±standard deviations) from seawater samples collected in Tolo Harbor.

Sewage

Precipitation

Freshwater/run-off

39% to 81% 76% to 100% 100% to 101% 68% to 100%

20% to 41% 20% to 50% 32% to 60% 24% to 48%

0% to 35% −3% to 0% −41% to 0% 0% to 2%

a The Tolo channel value was obtained by resolving the mass balance equation for the mean value of all samples (CT, CLP and TWP).


8


nitrogen removal by denitrification (0.5 b m b 0.7) as we observed in the case of the wastewater samples (m = 0.6). So, we explored this variability in greater detail by analyzing the relationship of seawater nitrate concentrations and isotopic compositions with DO, salinity, and chlorophyll-a content obtained from HK EPD (2014). The relationship of δ15NNOx (n = 18, r = −0.69, R2 = 0.48, p b 0.05) 2 and NO− 3 (n = 18, R = 0.71, R = 0.51, p b 0.05) with DO suggests that there is likely to be more denitrification towards the open ocean (Port Island) than inshore (Centre Island). Another possibility is phytoplankton assimilation driven nitrogen fractionation in seawater, which could result in 15N and 18O enrichment in the remaining nitrate in seawater (Sigman et al., 2009). However, we found no correlation between − δ15NNOx, NO− 3 +NO2 and chlorophyll-a content (p N 0.05), which refutes this alternative hypothesis. Moreover, nitrate and nitrite concentrations exhibited a strong negative correlation with salinity (n = 18, R = −0.80, R2 = 0.64, p b 0.05), indicating that there are more wastewater inputs that could be diluted with freshwater or groundwater at Centre Island than at Port Island. Therefore, we propose that the source of nitrogen in Tolo Harbor may not be entirely wastewater, or perhaps is a mixture of sources (Kellman and Hillaire-Marcel, 1998; Mayer et al., 2002). In other words, there is mixing of different sources of nitrate from sewage, atmospheric deposition, and run-off etc. (Fig. 10) and this is responsible for the differences in isotopic composition both spatially and between seasons in northeastern marine waters of Hong Kong (Thibodeau et al., 2013; Kendall et al., 2007). By applying the mixing model to every station in the Tolo Channel, we observed that the model was not able to resolve the equation for two of the three stations (Table 4), yielding negative values for the freshwater component and over 100% for the sewage component. This is likely to be due to the isotope fractionation linked to local denitrification, which tends to enrich both isotopes and then artificially boost our estimate of the contribution of nitrate originating from sewage for these stations. The resulting mass balance average for the entire Tolo Channel (Table 4) suggests that seawater in northeast Hong Kong is dominated by sewage in most cases (68%–100%, 95% CI). While this is probably an overestimation because of the denitrification effect, it is still a robust first attempt to estimate the relative importance of the different Nsources in Hong Kong seawaters. We hypothesize that this sewagederived nitrogen can originate from the SGD in Tolo Harbor as there are no direct inputs of wastewater in the harbor. Further investigation

of samples of SGD and seawater from subsurface layers in the area is needed to support this hypothesis. Contribution from atmospheric deposition can explain the variability between the dry and wet seasons. Although the above-mentioned study revealed valuable information regarding nitrogen source identification, the current dataset is restricted to northeastern marine waters. It is therefore necessary to carry out more extensive spatial and temporal sampling to identify and map point and non-point sources of nitrogen pollution using δ15NNOx in Hong Kong's entire marine environment. Meanwhile, the spatial and temporal distributions of isotopic compositions of ammonium and dissolved organic nitrogen should be measured in future to close the nitrogen budget in Tolo Harbor and adjacent coastal systems and to have better understanding of the nitrogen pollution history. 5. Conclusion Our results are important for understanding the impact of coastal urbanization in Hong Kong's marine waters. This study also demonstrates the usefulness of the multi-tracer approach in understanding nitrogen dynamics of wastewater treatment types. Moreover, the study demonstrates that nitrate isotopic composition of seawater in Tolo harbor can be best explained by mixing of variable external sources of nitrate, affected by seasonality. Although the current dataset is restricted to northeast Hong Kong, the study can be refined by including a mass balance model that can predict the proportion and composition of nitrogen pollution at any given sampling location in all of Hong Kong's marine waters. Going forward, the present study will serve as a baseline for long-term environmental monitoring of wastewater, seawater, rainwater and freshwater from different parts of Hong Kong that will help understand the dynamics of the nitrogen cycle and physical-biological marine processes and interactions. Acknowledgement The work reported is supported by an Early Career Scheme award from the Hong Kong Research Grants Council (RGC Ref No. 789913) of the Hong Kong Special Administrative Region, China. The assistance of the Drainage Services Department in providing access to the sewage treatment facilities is greatly appreciated. We are grateful to the Environmental Protection Department and the Agriculture Fisheries and

Fig. 10. Values of δ15NNOx and δ15NNOx of nitrate derived from various nitrogen sources in Hong Kong. Every box indicates the typical isotopic composition for that water type. All seawater samples from our study lie outside the typical isotopic composition for marine nitrate in South China Sea, indicating mixing of other sources. Dashed lines indicate typical expected slopes for data resulting from denitrification of nitrate with initial δ15N = +6‰ and δ18O = −9‰ (typical for sewage).



Conservation Department for giving us permission to sample seawater from northeast Hong Kong. Appendix A. Sampling dates

Dry season

Wet season

Sewage treatment plant (STP) Aberdeen STP Ap Lei Chau STP Wah Fu STP Shek O STP Sandy Bay STP Cheung Chau STP Tai O Imhoff Tank Shatin STP Sai Kung STP Tai Po STP Shek Wu Hui STP Peng Chau STP Mui Wo STP Stanley STP Cyberport STP Stonecutters Island STP Siu Ho Wan STP Ngong Ping STP

21.02.2014 24.02.2014 25.02.2014 26.02.2014 27.02.2014 12.03.2014 10.03.2014 13.01.2014 20.01.2014 14.01.2014 21.01.2014 17.03.2014 11.03.2014 26.02.2014 21.02.2014 21.03.2014 18.03.2014 19.03.2014

21.05.2014 22.05.2014 23.05.2014 26.05.2014 21.05.2014 31.07.2014 24.07.2014 25.06.2014 26.06.2014 27.06.2014 24.06.2014 29.07.2014 24.07.2014 20.05.2014 20.05.2014 23.04.2014 22.07.2014 22.07.2014

Sampling locations Che Lei Pai Tai Wan Port - Port Island Centre Island

04.04.2014, 04.11.2014 06.04.2014, 04.11.2014 04.04.2014

04.07.2014 04.07.2014 05.08.2014

References Adonadaga, M.G., 2014. Nutrient removal efficiency of activated sludge plants treating industrial and municipal wastewater in Ghana. Journal of Environmental Pollution and Human Health 2 (3), 58–62. Baker, D.M., Jordan-Dahlgren, E., Maldonado, M.A., Harvell, C.D., 2010. Sea fan corals provide a stable isotope baseline for assessing sewage pollution in the Mexican Caribbean. Limnol. Oceanogr. 55 (5), 2139–2149. Baker, D.M., MacAvoy, S.A., Kim, K., 2007. Relationship between water quality, δ15N, and aspergillosis of Caribbean Sea fan corals. Mar. Ecol. Prog. Ser. 343, 123–130. Baker, D.M., Rodriguez-Martinez, R.E., Fogel, M.L., 2013. Tourism's nitrogen footprint on a Mesoamerican coral reef. Coral Reefs 32 (3), 691–699. Casciotti, K.L., Sigman, D.M., Hastings, M.G., Bohkle, J.K., Hilkert, A., 2002. Measurement of the oxygen isotopic composition of nitrate in seawater and freshwater using the denitrifier method. Anal. Chem. 74, 4905–4912. Choi, K.W., Lee, J.H.W., Kwok, K.W.H., Leung, K.M.Y., 2009. Integrated stochastic environmental risk assessment of the Harbour Area Treatment Scheme (HATS) in Hong Kong. Environ. Sci. Technol. 43 (10), 3705–3711. Connell, D.W., Wu, R.S.S., Richardson, B.J., Leung, K., Lam, P.S.K., Connell, P.A., 1998. Occurrence of persistent organic contaminants and related substances in Hong Kong marine areas: an overview. Mar. Pollut. Bull. 36 (5), 376–384. Cope, M.A., Morton, B., 1988. The scleractinian coral community at Hoi Ha Wan, Hong Kong. Asian Mar. Biol. 5, 41–52. Costanzo, S.D., Udy, J., Longstaff, B., Jones, A., 2005. Using nitrogen stable isotope ratios (δ15N) of macroalgae to determine the effectiveness of sewage upgrades: changes in the extent of sewage plumes over four years in Moreton Bay, Australia. Mar. Pollut. Bull. 51, 212–217. Fabricus, K.E., McCorry, D., 2006. Changes in octocoral communities and benthic cover along a water quality gradient in the reefs of Hong Kong. Mar. Pollut. Bull. 52, 22–33. Flanagan, L.B., Ehleringer, J.R., Pataki, D.E., 2004. Stable isotopes and biosphere – atmosphere reactions: Processes and controls. Factors that Control the Isotopic Composition of N2O from Soil Emissions. Academic Press, pp. 72–75. Fry, B., 2006. Landscape Ecology and Isotope Maps. Stable Isotope Ecology. Springer Science + Business Media LLC, New York, USA, pp. 50–51. Gaston, T.F., Kostoglidis, A., Suthers, I.M., 2004. The 13C, 15N and 34S signatures of a rocky reef planktivorous fish indicate different coastal discharges of sewage. Mar. Freshw. Res. 55, 689–699. Gonfiantini, R., Stichler, W., Rozanski, K., 1995. Standards and intercomparison materials distributed by the International Atomic Energy Agency for stable isotope measurements. Proceedings of a Consultants Meeting Vienna, Austria 1–3 Dec 1995. International Atomic Energy Agency, pp. 13–29. Hashimoto, K., Matsuda, M., Inoue, D., Ike, M., 2014. Bacterial community dynamics in a full-scale municipal wastewater treatment plant employing conventional activated sludge process. J. Biosci. Bioeng. 118 (1), 64–71. Heaton, T.H.E., 1986. Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere - a review. Chem. Geol. 59, 87–102.

9

HK EPD, 2014. Annual reports. (Accessed on 10.06.2014. (Retrieved from) http://www. epd.gov.hk/epd/english/resources_pub/publications/pub_reports_ap.html. HK EPD, 1986–2014. Marine Water Quality Data. Accessed on 10.06.2014. Retrieved from) http://epic.epd.gov.hk/EPICRIVER/marine/?lang=en. HKSAR DSD, 2014a. Sewage treatment facilities. (Accessed on 09.06.2014. Retrieved from) http://www.dsd.gov.hk/EN/Sewerage/Sewage_TreatmentFacilities/index.html. HKSAR DSD, 2014b. Harbour area treatment scheme (HATS). (Accessed on 08.05.2015.Retrieved from) http://www.dsd.gov.hk/others/HATS2A/en/Stage_2A_ KEY_ELEMENTS/index.html. Howarth, R., Anderson, D., Cloern, J., Elfring, C., Hopkinson, C., Lapointe, B., Malone, T., Marcus, N., McGlathery, K., Sharpley, A., Walker, D., 2000. Nutrient pollution of coastal rivers, bays and seas. Issues Ecol. 7, 1–15. Kellman, L., Hillaire-Marcel, C., 1998. Nitrate cycling in streams: using natural abundances of NO3 −–δ15N to measure in situ denitrification. Biogeochemistry 43, 273–292. Kendall, C., Elliott, E.M., Wankel, S.D., 2007. Tracing anthropogenic inputs of nitrogen to ecosystems. In: Michener, R.H., Lajtha, K. (Eds.), Stable Isotopes in Ecology and Environmental Science, second ed. Blackwell Publishing, pp. 375–449 (Chapter 12). Kendall, C., Young, M.B., Silva, S.R., 2010. Applications of stable isotopes for regional to national-scale water quality and environmental monitoring programs. Isoscapes: Understanding Movement, Pattern and Process on Earth Through Isotope Mapping http://dx.doi.org/10.1007/978-90-481-3354-3_5. Khayat, S., Geyer, S., Hotzl, H., Ghanem, M., Ali, W., 2006. Identification of nitrate sources in groundwater by delta N-15(nitrate) and delta O-18(nitrate) isotopes: a study of the shallow Pleistocene aquifer in the Jericho area, Palestine. Acta Hydrochim. Hydrobiol. 34, 27–33. Lagenbach, K., Kuschk, P., Horn, H., Kastner, M., 2009. Slow sand filtration of secondary clarifier effluent for wastewater reuse. Environ. Sci. Technol. 43, 5896–5901. Leavitt, P.R., Brock, C.S., Ebel, C., Patoine, A., 2006. Landscape-scale effects of urban nitrogen on a chain of freshwater lakes in central North America. Limnol. Oceanogr. 51 (5), 2262–2277. Luo, X., Jiao, J.J., Moore, W.S., Lee, C.M., 2014. Submarine groundwater discharge estimation in an urbanized embayment in Hong Kong via short-lived radium isotopes and its implication of nutrient loadings and primary production. Mar. Pollut. Bull. 82, 144–154. Magni, P., Rajagopal, S., Como, S., Jansen, J.M., van der Velde, G., Hummel, H., 2013. δ13C and δ15N variations in organic matter pools, Mytilus spp. and Macoma balthica along the European Atlantic coast. Mar. Biol. 160, 541–552. Mariotti, A., Lancelot, C., Billen, G., 1984. Natural isotopic composition of nitrogen as a tracer of origin or suspended organic matter in the Scheldt estuary. Geochim. Cosmochim. Acta 48, 529–555. Mayer, B., Boyer, E.W., Goodale, C., Jaworski, N.A., Breemen, N.V., Howarth, R.W., Seitzinger, S., Billen, G., Lajtha, K., Nadelhoffer, K., Van Dam, D., Hetling, L.J., Nosal, M., Paustian, K., 2002. Sources of nitrate in rivers draining sixteen watersheds in the northeastern US: isotopic constraints. Biogeochemistry 57 (58), 171–197. McClelland, J.W., Valiela, I., 1998. Linking nitrogen in estuarine producers to land-derived sources. Am. Soc. Limnol. Oceanogr. 43 (4), 577–585. McCorry, D., Blackmore, G., 1998. Tolo revisited: a resurvey of the corals and their metal burdens in Tolo Harbour and Channel twelve years and one million people later. Pages 6–26 the marine flora and fauna of Hong Kong and Southern China V. Proceedings of the Tenth International Marine Biology Workshop: The Marine Flora and Fauna of Hong Kong and Southern China, Hong Kong. Hong Kong: University Press, Hong Kong. Morton, B., 1988. Editorial: Hong Kong's first marine disaster. Mar. Pollut. Bull. 19, 299–300. Ng, W.C., Morton, B., 2003. Genetic structure of the scleractinian coral Platygyra sinensis in Hong Kong. Mar. Biol. 143, 963–968. Ochoa-Izaguirre, M.J., Soto-Jimenez, M.F., 2015. Variability in nitrogen stable isotope ratios of macroalgae: consequences for the identification of nitrogen sources. J. Phycol. 51, 46–65. Onodera, T., Kanaya, G., Syutsubo, K., Miyaoka, Y., Hatamoto, M., Yamaguchi, T., 2015. Spatial changes in carbon and nitrogen sable isotope ratios of sludge and associated organisms in a biological sewage system. Water Res. 68, 387–393. Phillips, D.L., Gregg, J.W., 2001. Uncertainty in source partitioning using stable isotopes. Oecologia 127 (2), 171–179. Piola, R.F., Moore, S.K., Suthers, I.M., 2006. Carbon and nitrogen stable isotope analysis of three types of oyster tissue in an impacted estuary. Estuar. Coast. Shelf Sci. 66, 255–266. Risk, M.J., 2014. Assessing the effects of sediments and nutrients on coral reefs. Curr. Opin. Environ. Sustain. 7, 108–117. Risk, M.J., Lapointe, B.E., Sherwood, O.A., Bedford, B.J., 2009. The use of δ15N in assessing sewage stress on coral reefs. Mar. Pollut. Bull. 58 (6), 793–802. Rodriguez, A., 1981. Marine and coastal environmental stress in the wider Caribbean region. Ambio 10, 283–294. Rogers, K.M., 1999. Effects of sewage contamination on macro algae and shellfish at Moa Point, New Zealand using stable carbon and nitrogen isotopes. New Zealand Journal of Marine and Freshwater Research]–>N. Z. J. Mar. Freshw. Res. 33, 181–188. Scott, P.J.B., 1990. Chronic pollution recorded in coral skeletons in Hong Kong. J. Exp. Mar. Biol. Ecol. 139, 51–64. Scott, P.J.B., Cope, M., 1982. The distribution of scleractinian corals at six sites within Tolo Harbor and Channel. Pages 575–586. Proceedings of the First Marine Biological Workshop: The Marine Flora and Fauna of Hong Kong and Southern China, Hong Kong, 1980. Hong Kong University Press, Hong Kong.


10


Sebilo, M., Billen, G., Grably, M., Mariotti, A., 2006. Assessing nitrification and denitrification in the Seine river and estuary using chemical and isotopic techniques. Ecosystems 9 (4), 564–577. Sigman, D.M., Casciotti, K.L., Andreani, M., Barford, C., Galanter, M., Bohlke, J.K., 2001. A bacterial method for the nitrogen isotopic analysis of nitrate in seawater and freshwater. Anal. Chem. 73, 4145–4153. Sigman, D.M., Karsh, K.L., Casciotti, K.L., 2009. Ocean process tracers: nitrogen isotopes in the ocean. (Accessed on 18.11.2015. Retrieved from) http://www.princeton.edu/ sigman/publications/pdf/Sigman-Karsh-Casciotti-09.pdf. Smith, S.V., Kirnmerer, W.J., Laws, E.A., Brock, R.E., Walsh, T.W., 1981. Kaneohe Bay sewage diversion experiment: perspectives on ecosystem responses to nutritional perturbation. Pac. Sci. 35, 279–402. Somes, C.J., Schmittner, A., Galbraith, E.D., Lehmann, M.F., Altabet, M.A., Montoya, J.P., Letelier, R.M., Mix, A.C., Bourbonnais, A., Eby, M., 2010. Simulating the global distribution of nitrogen isotopes in the ocean. Glob. Biogeochem. Cycles 24. http://dx.doi.org/ 10.1029/2009GB003767. Thibodeau, B., Helie, J.F., Lehmann, M.F., 2013. Variations of the nitrate isotopic composition in the St. Lawrence River caused by seasonal changes in atmospheric nitrogen inputs. Biogeochemistry http://dx.doi.org/10.1007/s10533-013-9834-4. Tucker, J., Sheats, N., Giblin, A.E., Hopkinsin, C.S., Montoya, J.P., 1999. Using stable isotopes to trace sewage-derived material through Boston Harbor and Massachusetts Bay. Mar. Environ. Res. 48 (4–5), 353–375.

van Vliet, M.T.H., Zwolsman, J.J.G., 2008. Impact of summer droughts on the water quality of the Meuse river. J. Hydrol. 353 (1–2), 1–17. Walker, D.I., Ormond, R.F.G., 1982. Coral death from sewage and phosphate pollution at Aqaba, Red Sea. Mar. Pollut. Bull. 13, 21–25. Wear, S.L., Vega Thurber, R.L., 2015. Sewage pollution: mitigation is key for coral reef stewardship. Ann. N. Y. Acad. Sci. http://dx.doi.org/10.1111/nyas.12785. Xu, E.G.B., Leung, K.M.Y., Morton, B., Lee, J.H.W., 2014. An integrated environmental risk assessment and management framework for enhancing the sustainability of marine protected areas: the Cape d'Aguilar Marine Reserve case study in Hong Kong. Sci. Total Environ. 505, 269–281. Xu, J., Lee, J.H.W., Yin, K., Liu, H., Harrison, P.J., 2011. Environmental response to sewage treatment strategies: Hong Kong's experience in long-term water quality monitoring. Mar. Pollut. Bull. 62 (11), 2275–2287. Yeung, K.P., 2000. Changes in zooxanthellae density and chlorophyll-a concentration in corals Porites lobata after short-term sediment burial. Directed Research Report, Environmental Science Programme. The Chinese University of Hong Kong. Yim, W.W.S., Nau, P.S., Rosen, B.R., 1982. Permian corals in the Tolo Harbour formation, Ma Shi Chau, Hong Kong. J. Paleontol. 55, 1298–1300. Yin, K., Harrison, P.J., 2007. Influence of the Pearl River estuary and vertical mixing in Victoria Harbour on water quality in relation to eutrophication impacts in Hong Kong waters. Mar. Pollut. Bull. 54, 646–656.
