Environmental Pollution 233 (2018) 387e394

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Tracking major endocrine disruptors in coastal waters using an

integrative approach coupling field-based study and hydrodynamic
modeling*
Elvis Genbo Xu a, b, *, S.N. Chan c, K.W. Choi c, Joseph H.W. Lee c, Kenneth M.Y. Leung a, d, **
a
The Swire Institute of Marine Science and School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China
b
Department of Environment Sciences, University of California, Riverside, CA 92521, USA
c
Department of Civil and Environmental Engineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China
d
State Key Laboratory in Marine Pollution, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China

a r t i c l e i n f o a b s t r a c t

Article history: Many of the world's large coastal cities discharge partially treated wastewater effluents containing
Received 4 April 2017 various endocrine disrupting chemicals (EDCs) to coastal environments. Nonylphenols (NP) and
Received in revised form bisphenol A (BPA) were found to be the most abundant EDCs in sewage effluents in Hong Kong. The
14 September 2017
environmental fate and ecological risk of these two EDCs remains largely unknown, particular for coastal
Accepted 23 October 2017
systems with complex hydrodynamic flows. Based on a validated three-dimensional (3D) multiple-scale
hydrodynamic model, a field-based study was conducted to track the two EDCs from potential sources to
the only marine reserve in Hong Kong. The two compounds were detected in all seawater, suspended
Keywords:
Endocrine disrupting chemicals
particle, and sediment samples, with higher aqueous concentrations in wet season than in dry season.
Marine reserve High concentrations in sediments suggest sediment is a sink, posing an ecological risk to the benthos.
Submarine outfall The fate and transport of the two EDCs was predicted using a 3D near-field Lagrangian jet model
Hydrodynamic model seamlessly coupled with a 3D shallow water circulation model. The results suggested the NP and BPA in
Ecological risk the marine reserve cannot be solely attributed to the nearby submarine sewage outfall, but likely
concurrently contributed by other sources. This study calls for more effective measures of reducing the
use and release of these EDCs, and research to investigate their impacts on the marine benthos.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction degradation products of alkylphenol polyethoxylates, which have

Endocrine disrupting chemicals (EDCs) are ubiquitous in coastal
marine environments of urbanized coastal cities like Hong Kong
(Kueh and Lam, 2008). We previously found that two common
EDCs, namely nonylphenols (NP) and bisphenol A (BPA), exhibited
elevated concentrations in seawater samples collected across the
marine environment of Hong Kong (Xu et al., 2014, 2016). The two
compounds have been identified as major anthropogenic contrib-
utors to endocrine-disrupting activities in aquatic environments
worldwide (Auriol et al., 2006). Nonylphenols are the main
* major source of the EDCs being released into aquatic environments,
This paper has been recommended for acceptance by Prof. von Hippel Frank A.
* Corresponding author. Department of Environment Sciences, University of
California, Riverside, CA 92521, USA.
** Corresponding author. School of Biological Sciences, The University of Hong
Kong, Pokfulam, Hong Kong, China.
E-mail addresses: genboxu@ucr.edu (E.G. Xu), kmyleung@hku.hk (K.M.Y. Leung).
been used widely as surfactants in household detergents, agricul-
ture and textile industry (White et al., 1994). NP can cause an in-
crease of vitellogenin and a decrease in the growth rate of testes in
male rainbow trout, and disturb the estrogen receptor (ER)
pathway in juvenile Atlantic salmon at low environmental con-
centrations (White et al., 1994; Bonefeld-Jørgensen et al., 2007).
BPA is an industrial raw material mainly used in plastic, rubber,
adhesive, and cable industries, and is known to cause a delay in the
hatching of eggs and a suppression of growth in juvenile rainbow
trout (Aluru et al., 2010), and decrease their sperm production at
20 ng/g (Von Saal et al., 1998). It has been widely recognized that
effluent discharges from sewage treatment plants (STPs) are the
including rivers, estuaries and coastal marine waters (Niven et al.,
2001; Zhang and Zhou, 2008; Grover et al., 2011). Submarine out-
falls are commonly used for disposal of partially treated effluents
from STPs in urbanized coastal cities, like Hong Kong. These

https://doi.org/10.1016/j.envpol.2017.10.086
0269-7491/© 2017 Elsevier Ltd. All rights reserved.
388 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394

partially treated sewage discharges are typically located in rela-
tively shallow waters (5e20 m depth), and often close to sensitive
receivers such as bathing beaches, fishing grounds and marine
protected areas (MPAs) (Choi et al., 2009; Xu et al., 2015a, 2016).
Twelve phenolic EDCs have been detected in treated effluents
discharged from the STPs in Hong Kong, which may pose risk to
marine organisms inhabiting in the receiving water body, including
those living in the Cape D' Aguilar Marine Reserve (CAMR), which is
situated on the southeastern tip of Hong Kong Island (Fig. 1) (Xu
et al., 2014, 2015b).
In 1995, the Hong Kong government, recognizing the impor-
tance and urgency of protecting the environment of Cape D' Aguilar,
designated this area as the first and only marine reserve in Hong
Kong. The CAMR has a rich marine biodiversity, with numerous
inhabitants such as fishes, corals, and marine invertebrates
(Morton, 1995). This reserve is considered to be of immense sci-
entific and environmental significance, and is managed as a single
ecological unit under special protection by Hong Kong legislation
(i.e., Cap 476: Marine Parks Ordinance). However, due to its small
size (0.2 km2) and close proximity to human activities in coastal
areas, the conservation of this reserve may be jeopardized by
anthropogenic impacts, such as partially-treated sewage discharges
and contaminated surface runoff (Xu et al., 2015a).
Due to concerns about the adverse impacts of NP and BPA on
aquatic organisms worldwide, the levels of these chemicals have
been quantified in different riverine and estuarine environments
(Naylor et al., 1992; Blackburnet et al., 1999; Isobe et al., 2001;
Ferguson et al., 2001; Rice et al., 2003; Jonkers et al., 2003; Li

Fig. 1. Sampling locations along transects from Big Wave Bay, Shek O Sewage Treatment Plant and Shek O swimming beach to Cape D'Aguilar Marine Reserve (CAMR) in Hong Kong
(a). Concentrations (mean ± SD, n ¼ 3) of nonylphenols (NP) and bisphenol A (BPA) in seawater (SW), suspended particles (SP) and sediments (SD), and E. coli in seawater from Big
Wave Bay to Shek O to the Cape D' Aguilar Marine Reserve (CAMR) in March 2013 (b and c) and August 2013 (d and e).
 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394
et al., 2004); however as far as we are aware only a few related
studies have been conducted in open coastal marine waters (Bester
et al., 2001; Chen et al., 2006). Oceanic currents can affect the
transport of different scales of matters in seawater, from trace
chemicals such as inorganic nutrients to visible suspended particles
(Christiansen et al., 2000; Leonard and Reed, 2002). The interaction
of the discharge momentum and buoyancy with the oceanic flow
can result in rapid initial mixing and dilution of the sewage effluent
close to the outfall discharge point. The pollutant concentrations in
the treated sewage effluent can be significantly reduced by the time
the mixed effluent rises to the surface or reaches a trapped level in
the presence of ambient stratification (Choi et al., 2009). In a pre-
vious study, we found that effluents from different STPs in Hong
Kong contained significant concentrations of varying EDCs, with NP
and BPA being the most abundant (Xu et al., 2014). Little is known,
however, about the transport and environmental fate of these two
compounds in the coastal waters after discharge from the subma-
rine sewage outfall. This is particularly true for coastal systems with
complex ecological and hydrodynamic characteristics.
In this study, we hypothesized that the monsoon-induced
onshore oceanic flows interacting with the tidal currents could
bring the aqueous NP and BPA from the Shek O STP effluent
discharge southward to the CAMR and its vicinity (Fig. 1). Given
that the distribution of phenolic EDCs is often closely associated
with sedimental organic carbon through hydrophobic interaction
(Ferguson et al., 2001; Rice et al., 2003; Jonkers et al., 2003; Chen
et al., 2006; Johnson et al., 1998), we predict that the concentra-
tions of both compounds are significantly positively correlated with
the level of organic matter content in the suspended particles and
sediment, while concentrations of the two EDCs increase vertically
downward from surface water to bottom water and to sediment.
To test these hypotheses, we investigated the horizontal and
vertical distributions of NP and BPA in different environmental
compartments (i.e., water, suspended particles and sediment
samples) along transects in close proximity to three pollution
sources (i.e., Shek O STP, Shek O Beach and Big Wave Bay Bathing
Beach) to the CAMR during both dry and wet seasons. Given the
limited treatment capacity of the preliminary Shek O STP, the
sewage outfall is expected to be an important source of NP and BPA.
To track the transport and fate of the two EDCs, we applied a
validated multi-scale 3D hydrodynamic model, which dynamically
couples a near-field 3D Lagrangian jet model with a 3D shallow
water circulation model at grid scale to model and forecast the
transport and distribution of NP and BPA in aqueous phase from the
Shek O STP outfall to the CAMR, while other possible EDC pollution
sources were also discussed. The results of this study provide
essential information about the current status of EDC pollution in
subtropical Hong Kong waters, and shed light on the water quality
management strategy for regulating these pollutants in coastal
cities worldwide.
2. Materials and methods
2.1. Sampling
Seawater and sediment samplings were carried out on March
28, 2013 (dry season) and August 14, 2013 (wet season). To trace NP
and BPA from the pollution sources to the receiving water at the
CAMR, twelve sampling sites were selected along three transects,
from Big Wave Bay, Shek O STP outfall and Shek O Beach to the
CAMR (see Fig. 1). The distance between any two sampling stations
was approximate 500 m. Sites B1, S4 and O were the locations of the
three pollution sources (i.e., Big Wave Bay, Shek O Beach and Shek O
STP outfall), and the CAMR was the target sensitive receiving water
of interest. The outfall of Shek O STP is configured with three 10 m
389
deep pipes (information from Drainage Services Department of the
Hong Kong Special Administrative Region (HKSAR) Government).
Water samples collected from surface and bottom of the water
column as well as their combinations were analyzed for NP, BPA
and E.coli. Accordingly, 500 mL water samples were taken from
both 0.5 m and 10 m below the water surface along each of the
three transects. For the site where the maximum water depth is less
than 10 m deep, the bottom sample was taken at 0.5 m above the
seabed. Both composite samples and individual samples from the
surface and bottom water layers were taken at each sampling sta-
tion in triplicates. At the location of the STP outfall, vertical
seawater samples of triplicate were collected at 1, 4, 7, 10 m below
the water surface and at 0.5 m above the sea bottom. About 50 mL
of methanol was immediately added to each 1 L water sample to
suppress biotic activities and the pH was adjusted to 3e4 using 4 M
H2SO4 on the research boat. Water samples were transported on ice
in a cooler to the laboratory and stored at 4  C in the dark; subse-
quent chemical analysis was performed within 24 h. Surface sedi-
ment samples (up to top 20 cm) were collected using a stainless
steel grab sampler, scooped into glass jars and stored at 20  C
until further analysis.
2.2. Analytical methods
2.2.1. Characteristics of water quality and surface oceanic flows
Seawater temperature, salinity, and pH were measured with a
portable salinity meter (AZ-8371, AZ Instrument Corp., Taiwan),
and a portable pH Eco Sense Meter (YSI, US). The properties of the
seawater samples are summarized in Table S1. Flow velocity and
direction were measured at each sampling site in situ. All the
measurements were taken 1.5 m below seawater surface using an
Acoustic Doppler Velocimeter (ADV) with a 2D probe (SonTek),
attached to a wading rod with a compass. The operation of the ADV
followed the Flow-Tracker Handheld ADV Principles of Operation
(May 1, 2001) (Meral et al., 2008). The measurement range of ADV is
between 0.001 and 4.0 m/s with an accuracy of ±1%.
2.2.2. Measurement of total organic carbon and E. coli
The analytical methods for total organic carbon in the sediment
samples were documented in Ran et al. (2002). In brief, about 5 g of
the freeze-dried sediment sample was treated with 10% HCl for 24 h
to remove inorganic carbon, and then dried overnight at 105  C. The
organic carbon contents were determined on an elemental analyzer
(Vario EL III Elementar, Germany) with acetanilide used as external
standard. Sample duplicates were processed and their mean values
are reported in Table S1. E. coli and total coliform groups in seawater
samples were measured using a membrane filtration method
adopted by the Environmental Protection Department of Hong
Kong Special Administrative Region (HKSAR) Government (http://
www.beachwq.gov.hk/en/programme.aspx).
2.2.3. NP and BPA extraction and analysis
NP and BPA were extracted using solid phase extraction and
ultrasonic-assisted solvent extraction, and determined with gas
chromatography-mass spectrometry (GC-MS) based on the
methods described in Zhao et al. (2011). Detailed extraction, clean-
up, and analytical procedures are described in the Supplementary
Information. The average recoveries of NP and BPA ranged between
72 and 116%, and 83 and 108%, respectively. The method detection
limits (MDLs) were evaluated by a signal-to-noise ratio of 3, and
MDLs of NP and BPA were 12.3 and 2.1 ng/L for seawater respec-
tively, and 7.0 and 2.0 ng/g dry weight (dw) for sediment,
respectively.
390 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394
2.2.4. Three-dimensional (3D) near field Lagrangian jet model
The sewage effluent from Shek O STP is discharged through an
outfall on the southeastern shore of Hong Kong Island to the
receiving waters in the form of three horizontal jets in a water
depth of about 10 m to reduce its impact from the nearby Shek O
bathing beach. The outfall design is aimed to direct the effluent at
high velocity from a rock cliff away from the beach area as much as
possible. The jets have a nozzle diameter of 3.6 cm. The salinity of
sewage in Hong Kong is about 10 ppt (due to the use of seawater for
toilet flushing), with a density of 1010 kg/m3. The salinity of
ambient seawater is about 33 ppt in dry season, with a density of
1030 kg/m3. The temperatures of both sewage and ambient are
about 20  C in the dry season and have little effect on density dif-
ference. Based on an average sewage flow rate of Q ¼ 864 m3/
d (0.01 m3/s), jet velocity U0 ¼ 3.27 m/s (information from Drainage
Services Department of HKSAR Government) and the density dif-
ference between sewage and seawater Dr/r ¼ 0.02, p the jet densi-
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi was from 1 March to 30 March 2013, coincidence with the dry
metric Froude number was calculated using Fr ¼ U0 = ðDr=rÞgD as
37 (Fig. 2). A schematic representation of the outfall and a 3D
depiction of the sewage jets were predicted using VISJET 2.5 (http://
www.aoe-water.hku.hk/visjet/visjet.htm). VISJET is an interactive
computer modeling system that predicts the impact of an effluent
discharge into the water environment. The model provides 3D flow
visualization of the predicted path and mixing of an arbitrarily
inclined buoyant plume in moving receiving waters which may be
density-stratified (Lee and Chu, 2003). The ambient conditions for
the jet model are provided by the 3D hydrodynamic model pre-
diction, driven by real tidal condition and measured wind forcing
for the dry season period of 1e30 March 2013 (see below).
2.2.5. 3D hydrodynamic tidal model
The hydrodynamics of Hong Kong coastal waters is governed by
astronomical tides, monsoon induced oceanic currents, and Pearl
River freshwater discharge. The 3D hydrodynamic model was
developed based on the Environmental Fluid Dynamics Code
(EFDC) which solves the shallow-water equation numerically using
a curvilinear structural grid and sigma-coordinates in the vertical
direction (Hamrick, 1992). The two-equation Mellor-Yamada tur-
bulence model was used for predicting the turbulent diffusivity.
The model consists of about 21,500 computational grid cells in the
horizontal direction and 10 equal sigma layers. The horizontal grid
size ranges from about 100 m to 10 km, within the study region the
grid size is about 200e500 m. A computational time step size of 3 s
is used. The Distributed Entrainment Sink Approach (DESA) was
used to dynamically couple the near field buoyant jet model
Fig. 2. Schematic representation of the Shek O outfall (a), the 3D trajectory of the jets (b), and vertical profiles of nonylphenols (NP), bisphenol A (BPA) and E. coli at the Shek O
outfall (c). SW, seawater; SP, suspended particles.
predictions with the far field shallow water circulation model
computations at grid cell level (Choi and Lee, 2007). Essentially the
physical action of the turbulent entrainment of the sewage jets on
the surrounding flow was modeled as a series of sinks of known
strength along the predicted jet trajectories. It is a sub-grid scale
model; the effects of the sewage outfall configuration, including the
size and orientation of the discharge ports, were included as the
input to the jet model, and refined model grid sizes around the
discharge would not be needed. This model has been extensively
validated and applied for daily beach water quality forecast, envi-
ronmental impact assessment of sewage discharge, setting of
effluent standards, optimization of chlorine disinfection for sewage
and emergency response in Hong Kong (Chan et al., 2013a, 2013b;
Lee et al., 2014; Lee et al., 2015).
The model simulation was carried out with a 30-day spin-up
period from a cold start. The modeled period for the EDC study
season sampling period. The real astronomical tide with 8 tidal
constituents was used as ocean boundary forcing, with measured
daily wind corrected to 10 m above sea level as surface forcing. The
uniform salinity of 33 ppt was prescribed at the open boundary. The
hydrodynamic model has been extensively validated against
measured tide level, tidal current and salinity data in Hong Kong
recorded during 1998e2007. Details of the model set-up and vali-
dation can be found in Choi et al. (2012) and Chan et al. (2013a).
The present modeling aimed to evaluate the contribution of the
Shek O sewage outfall to the levels of EDCs in the marine reserve by
tidal mixing and transport. The concentration of NP in the Shek O
outfall effluent was assumed to be 1500 ng/L based on the reported
data of Xu et al. (2014) for dry season. The transport of NP by the
tidal flow was simulated by solving the advective-diffusion equa-
tion with the pollutant as a conservative tracer.
2.2.6. Data analysis
A non-parametric Wilcoxon test was used to examine for sea-
sonal differences in the concentrations of NP or BPA in the seawater,
suspended particles and sediment samples collected along the
three transects. For each season, the difference of concentrations of
each EDC between bottom seawater and surface seawater was also
tested using a Wilcoxon test. Cluster analysis was conducted based
on the interaction of NP, BPA, E. coli and TOC concentrations at each
location using PRIMER 6 (Clarke and Gorley, 2006). A dendrogram
was constructed based on Euclidean distance using PRIMER 6. For
all statistical tests, p < 0.05 was considered significant for rejecting
the null hypothesis.
 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394
3. Results and discussion
3.1. Horizontal, vertical and seasonal distribution of NP and BPA
Both NP and BPA were detected in all samples, ranging from 76.5
to 1353.6 and 4.4e114.1 ng/L in seawaters, from 97.7 to 1492.5 and
5.3e141.8 ng/L in suspended particles, and from 114.6 to 2665.0 and
21.3e236.8 ng/g dw in sediments, respectively. Consistently, NP
had higher concentrations at all sites compared to those of BPA
(Table S2). Their concentrations were lower than those observed
from the Pearl River Estuary of China to the west of Hong Kong
(Gong et al., 2012), but higher than those previously reported in
Hong Kong (Kueh and Lam, 2008; Li et al., 2007). The highest
concentrations of NP and BPA in seawaters, suspended particles and
sediments were all found at site O, where the sewage outfall of Shek
O STP is located (Fig. 1). Bennett and Metcalfe (2000) also found the
distribution of NP was localized to areas close to (e.g., 1 km) the
discharge points of sewage effluents.
During the sampling in March 2013, the oceanic current was
measured in situ. There was a southward oceanic current flow from
the outfall to site MR, with an average velocity of 0.17 m/s
(Table S1). The model prediction of tidal current in the dry season is
shown in Fig. S1. Net oceanic current was observed from the
northeast direction due to the prevailing northeast monsoon in the
dry season. During flood tide, the tidal current came from the
eastern part of Hong Kong waters. The flooding tidal current was
diverted northwestwards to the Victoria Harbor at the east of Shek
O and CAMR, and also to the south of CAMR, with a maximum
velocity of 0.3 m/s. During ebb tide, there is an eastward flow from
Victoria Harbor and south of Hong Kong Island that converges
around the east of Shek O and CAMR. Velocity magnitude during
ebb tide was relatively weaker than flood tide due to the interaction
with the oceanic monsoon current, about 0.1e0.2 m/s. In general,
the water concentrations of NP and BPA in “downstream” seawater
samples (i.e., S1 to S8, see Fig. 1) were 10 times lower than that at
site O (i.e., the outfall), although site CAMR showed higher con-
centrations than other “downstream” sites despite the fact it is at
the far end of the sampling transect (Fig. 1). This decreasing trend
from the sewage outfall may be attributed to the jet dilution and
tidal mixing effect at the “downstream” sites. A similar distribution
trend was also observed for NP and BPA in suspended particles and
E. coli in seawaters. It was interesting to find that the second highest
level of E. coli was found in site B1 (Big Wave Bay) in August 2013;
the large number of tourists on this popular beach in the summer
may furnish another potential source of high E. coli concentration.
As with sediment samples, the highest concentrations of NP and
BPA were also found at site O (i.e. the outfall), which were five and
three times higher than those at the other sampling sites, respec-
tively (Fig. 1). Untreated municipal wastewaters from Shek O area
were discharged into the sea before the commissioning of the
screening plant in 1990. Large amounts of NP and BPA could thus
accumulate in the sediments. The Shek O wastewater had not been
treated until 1990 when the Shek O preliminary STP was built;
nevertheless, this preliminary treatment plant showed poor
removal of NP and BPA (Xu et al., 2014). Therefore, significant
concentrations of NP and BPA were detected in all sediment sam-
ples nearby the Shek O STP in the present study.
In seawater, NP showed significantly higher concentrations in
August 2013 (wet season) than those in March 2013 (dry season)
(Wilcoxon test; p < 0.01). This result is consistent with previous
reports of higher concentrations of NP in seawater found at higher
summer temperatures (Xu et al., 2014; Isobe et al., 2001). The
observed seasonal variation could be due to the higher degradation
rate of primary compounds of NP, and an increased storm water and
surface runoff input containing these EDCs during summer. Both NP
391
and BPA in suspended particles sampled in August 2013 were
significantly higher than those sampled in March 2013 (Wilcoxon
test; p < 0.05), while the temporal differences of their concentra-
tions were not significant in sediment (Wilcoxon test; p > 0.05).
Many parameters could affect the distribution dynamics of EDCs,
such as temperature, the degradation rate of mother compounds,
the activity of microbiology, characteristics of the sediment, drift of
sediment, sedimentation rate and content of oxygen (Li et al.,
2004). Zhao et al. (2011) also found that the seasonal differences
of EDCs in sediment were not as obvious as those in surface water
suggesting that sediments are a sink for NP and BPA.
The concentrations of NP and BPA in surface and bottom water
layers found in this study were summarized in Fig. S2. The con-
centrations of NP in the bottom water layer were significantly
higher than those in the surface water (Wilcoxon test; p < 0.05),
with a maximum magnification (bottom/surface > 5) at site O (i.e.,
the outfall). Similar to NP, concentrations of BPA were also higher in
bottom water than in surface water at most stations. A similar trend
has been observed in other investigations, for example, a higher
level BPA was observed in the bottom water layer of Natara River in
Japan (Funakoshi and Kasuya, 2009), and higher levels of NP was
found in the bottom water layer of Changjiang River in China (Shao
et al., 2002). The different concentrations between surface and
bottom water layers could be explained by the adsorption of NP and
BPA to sediments, resuspension into the bottom water, and pro-
tection from aerobic degradation and photolysis of these chemicals
in the bottom water (Mao et al., 2012). In the clustering method, all
locations within a cluster were considered to have similar features
in NP, BPA, E. coli and TOC concentrations. Based on the dendrogram
(Fig. S3), the sampling sites were classified into three main groups
(highly-polluted station, moderately-polluted stations and the
rest). Station O was identified as the most polluted and was clearly
separated from other sites probably due to the direct discharge of
EDC-contaminated sewage at this station (Fig. S3). Shek O Swim-
ming Beach (station S4) and the CAMR belong to the second cluster,
moderately-polluted group. These two stations are either close to
the Shek O or the CAMR and the Swire Institute of Marine Science,
and are expected to receive more domestic sewage compared to
other stations.
In comparison with other investigations, the concentrations of
NP and BPA measured in the present study fell in the lower ranges
of reported observations in similar settings worldwide (Table S3).
The concentration of BPA was around 1e3 orders of magnitude
lower than the reported acute toxicity levels (microgram per liter)
(Levy et al., 2004), but sedimental concentrations of NP detected in
this study was comparable with the concentration that may induce
adverse endocrine disruptive effects in marine organisms (Naylor,
1995; Nice et al., 2000). Furthermore, the environmental risk of
NP and BPA was discussed on the basis of environmental quality
standards from the EU. Risk quotients (RQs) were employed for the
initial evaluation of the ecological risks. The RQs were calculated as
follows: RQi ¼ MECi/PNECi, where MECi is the measured concen-
tration and PNEC the predicted no-effect concentration for the i
compound. There is a risk when RQ > 1. The PNECs for NP and BPA
are 300 and 150 ng/L in marine water, and 39 and 6.3 ng/g dw in
sediment, respectively (EC 2008, 2010; EU, 2002, 2010). Risks were
found for both NP (average RQ ¼ 14) and BPA (average RQ ¼ 8) in
sediments to the benthic organisms in the southeast Hong Kong
Island including the CAMR.
3.2. Partitioning of NP and BPA between dissolved and particulate
phases
There were significant amounts of NP and BPA present in par-
ticulate phase in seawaters collected along the transects (51% for NP
392 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394
and 41% for BPA on average) (Table S2). In general, NP was more
adsorbed to suspended particles than BPA, which may be due to
their differences in hydrophobicity (Isobe et al., 2001). Similar
partition patterns for the two chemicals were reported by other
investigators (28e67% for NP and 5e40% for BPA) (Li et al., 2004;
Arditsoglou and Voutsa, 2010). Spatial variations of the concen-
trations of NP and BPA in seawaters and suspended particles were
similar. Such similarities are reflected in the significant positive
relationship between the chemical in seawater and in suspended
solids, with a highly significant coefficient of determination (r2)
(p < 0.001; Fig. S4). The value of r2 for NP and BPA is 0.98 and 0.64,
respectively. A similar linear correlation for NP between water and
suspended particles was also observed by Xu et al. (2006).
The total organic matter in the sediments appeared to be an
important factor in controlling the distribution and fate of NP and
BPA. The present study demonstrated that sediments with high
total organic carbon contents (TOC) showed high concentrations of
NP and BPA in the sediments. Fig. S4 presents the significant cor-
relation (p < 0.01) between TOC and sediment concentrations of NP
(r2 ¼ 0.72) and BPA (r2 ¼ 0.26). The correlation between BPA with
TOC is weaker than that between NP with TOC. This difference can
be explained by many factors. One key factor is associated with
their hydrophobicity. The log Kow value of BPA (3.32) is lower than
that of NP (4.48) suggesting a lower affinity of BPA to TOC than that
of NP (Ferguson et al., 2001; Gong et al., 2012).
The in situ organic carbon-normalized particle-water partition
coefficient (Koc) can be calculated using the concentration in dis-
solved phase and particulate phase and organic carbon contents in
the particles with the following formula: Koc¼Cs/Caq/foc, where Cs is
the solid-phase concentration on a unit weight basis (ng/g), Caq is
the aqueous phase concentration on a unit volume basis (ng/mL),
and foc is the mass fraction of organic carbon on the particles (Isobe
et al., 2001). The mean log Koc values for NP and BPA were calcu-
lated as 5.82 ± 0.54 and 4.65 ± 0.39 (±SD) (n ¼ 72), respectively.
These values are within the ranges reported from other field studies
(Table S4), suggesting that the water-particle partition behaviors of
NP and BPA are generally similar regardless of the origin and nature
of the sources, water quality, and chemical analysis methods (Gong
Fig. 3. Predicted depth-averaged NP level (ng/L) at high tide (a) and low tide (b) from Shek O Sewage Treatment Plant outfall to the Cape D'Aguilar Marine Reserve (CAMR) and
nearby fields.
et al., 2012). A strong positive correlation between the Koc and Kow
(i.e., organic carbon-normalized particle-water partition coeffi-
cient) for EDCs was found by previous investigators (Arditsoglou
and Voutsa, 2010; Schwarzenbach et al., 1993). In summary, the
partition of NP and BPA in marine environments is dependent not
only on their hydrophobicity but also on factors such as contents of
organic matter and particulate matter, and hydrodynamic condi-
tions in the marine system (Yang et al., 2011).
3.3. The sewage jets and vertical distribution profiles at outfall
A schematic representation of the outfall and a 3D depiction of
the sewage jets predicted using VISJET is shown in Fig. 2. At the
outfall site, NP and BPA in seawater and suspended particles, and
E. coli in seawater samples were low at the surface, but they slightly
increased with depth to 7 m, and sharply increased from 10 m to
the seabed (Fig. 2). The concentrations of NP and BPA reached the
maxima at the bottom (12 m) but not at 10 m, which could be
explained by the inputs of settling particles or the resuspension of
bottom sediments or their combined effects. Yoshida et al. (2009)
also demonstrated the highest concentration of NP and the high-
est concentration of BPA were found in the same water layer.
Concentrations of NP showed significant correlations with several
water quality parameters (see Table S5). The negative correlations
between concentrations of NP and salinity (r ¼ 0.989, p < 0.001)
and dissolved oxygen (r ¼ 0.993, p < 0.001) suggested that the
origin of NP may come from freshwater surface runoff or partially
treated wastewater discharge outfall. However, such correlation
was not significant for BPA e a point which deserves further
investigation.
3.4. Transport simulation of Shek O sewage-derived NP and BPA
The results of the cluster analysis and the vertical distribution
profiles of NP and BPA suggested that the submarine outfall of the
nearby sewage treatment plant is a possible source of NP and BPA in
the marine environment of southeast Hong Kong Island. In order to
confirm the influence of the sewage-induced NP and BPA on the
 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394
CAMR, the transport of NP and BPA by the tidal flow was simulated
by solving the advective-diffusion equation with NP and BPA as the
conservative tracers. The near field dilution of the Shek O outfall
was about 500e2000, depending on the tidal current. The
maximum measured current was 0.3 m/s during flood tide and
0.2 m/s during ebb tide, respectively (Fig. S1). Larger dilution was
observed during peak tidal current and the sewage plumes were
predicted to reach the surface due to the weak stratification in dry
season. Based on the near field dilution prediction, the concentra-
tions of NP at the outfall were 0.5e3 ng/L above the background
level. The measurement of NP around the coastal region of Shek O
and the CAMR region was 100e200 ng/L. Such elevated levels could
be due to the combined effects of persistent background level of NP
in Hong Kong waters (Kueh and Lam, 2008), relatively slow
seawater exchange rate with current velocity of 0.033 m/s in the
CAMR, and treated sewage effluent discharged from the Swire
Institute of Marine Science and surface runoff at this location, but
not solely induced by the Shek O outfall. Fig. 3 shows the 3D-model
predicted contours of NP induced by Shek O outfall alone. During
flood tide and high water (Fig. 3a), the sewage plume extended
along the eastern coast of Hong Kong Island, with an average
concentration of NP about 0.1 ng/L. During ebb tide (Fig. 3b), the
sewage plume was transported to the south to the CAMR by the
tidal current; the NP concentration induced by the Shek O outfall
was almost negligible due to weaker tidal current during ebb. It is
reasonable that a similar conclusion could be made for explaining
the concentration of BPA in the marine reserve and nearby coastal
water, where the concentration of BPA in effluents of Shek O outfall
was about 700 ng/L while that in the nearby coastal waters are
measured in the order of 5e60 ng/L. Again, such concentrations of
BPA could not be induced by the Shek O STP outfall alone. Possible
sources of background EDCs in Hong Kong waters include surface
runoff from the hinterland (Kueh and Lam, 2008), re-suspension of
polluted sediments and Pearl River discharge (Fu et al., 2003),
shipping activities, accidental spill, and atmospheric deposition. In
particular, the seabed sediment might be a source of the EDCs to the
water column and the dynamics of sediment particles are governed
by oceanic waves, tidal currents and turbulent mixing. Further
studies on particle resuspension model, sediment transport model,
non-point source pollution model and atmospheric deposition
model may shed more light on the interpretation of the other
possible EDC pollution sources.
4. Conclusions
Overall, the environmental transport and fate of NP and BPA was
investigated in the marine environment of southeast Hong Kong
Island. The two compounds were detected in all samples with
higher concentrations in seawater and suspended particles during
the hot, wet season than in the cold, dry season. The concentrations
of NP and BPA in the sediment samples were associated with TOC,
and exceeded their sediment quality benchmarks, indicating
contamination of NP and BPA in the studied areas might have
caused potential estrogenic effects on marine benthos including
those living in the marine reserve. The results from the hydrody-
namic and transport model simulations suggested the elevated
concentrations of NP and BPA detected in the marine reserve and its
vicinity could not be solely attributed to the Shek O outfall effluent
discharge; other possible sources include surface runoff from the
hinterland, re-suspension of polluted sediments, and Pearl River
discharge. This study highlights the need for a multidisciplinary
approach in ecological risk assessment, involving environmental
chemistry and hydrodynamics. The ubiquitous occurrence of NP
and BPA in the coastal environment calls for more effective mea-
sures of controlling these two widespread and abundant EDCs, and
393
studies of their impacts on benthic marine organisms in this region.
Acknowledgements
The work described in this paper is supported by Research
Grants Council of the Hong Kong SAR Government via Collaborative
Research Fund (project no. C7044-14G) and Theme-based Research
Scheme (project no. T21-711/16-R). Elvis Xu would like to thank
John Swire & Sons Limited for awarding the James Henry Scott PhD
Scholarship. The authors also thank Mr. Ming Cheung and Mr. M.H.
Cheung for their assistance in field sampling, and Dr. Ken T.M.
Wong for his help in oceanic current measurement. We would also
like to thank Ms. Jessie Lai, Ms. Helen Leung, and Ms. Holly Wong
for their technical assistance in chemical analysis. The authors are
grateful to Mr. Scott Coffin of University of California, Riverside, USA
for proofreading and improving the use of English in an early draft
of this manuscript.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.envpol.2017.10.086.
References
Aluru, N., Leatherland, J.F., Vijayan, M.M., 2010. Bisphenol A in oocytes leads to
growth suppression and altered stress performance in juvenile rainbow trout.
PLoS One 5 (5), e10741.
Arditsoglou, A., Voutsa, D., 2010. Partitioning of endocrine disrupting compounds in
inland waters and wastewaters discharged into the coastal area of Thessaloniki,
Northern Greece. Environ. Sci. Pollut. Res. 17, 529e538.
Auriol, M., Filali-Meknassi, Y., Tyagi, R.D., Adams, C.D., Surampalli, R.Y., 2006.
Endocrine disrupting compounds removal from wastewater, a new challenge.
Process Biochem. 41 (3), 525e539.
Bennett, E.R., Metcalfe, C.D., 2000. Distribution of degradation products of alkyl-
phenol ethoxylates near sewage treatment plants in the lower Great Lakes.
N. Am. Environ. Toxicol. Chem. 19 (4), 784e792.
Bester, K., Theobald, N., Schroder, H. Fr, 2001. Nonylphenols, nonylphenol-
ethoxylates, linear alkylbenzenesulfonates (LAS) and bis(4-chlorophenyl)-sul-
fone in the German Bight of the North Sea. Chemosphere 45, 817e826.
Blackburnet, M.A., Kirby, S.J., Waldock, M.J., 1999. Concentrations of alkylphenol
polyethoxylates entering UK estuaries. Mar. Pollut. Bull. 38, 109e118.
Bonefeld-Jørgensen, E.C., Long, M., Hofmeister, M.V., Vinggaard, A.M., 2007. Endo-
crine-disrupting potential of bisphenol A, bisphenol A dimethacrylate, 4-n-
nonylphenol, and 4-n-octylphenol in vitro: new data and a brief review. En-
viron. Health Perspect. 115, 69.
Chan, S.N., Thoe, W., Lee, J.H.W., 2013a. Real-time forecasting of Hong Kong beach
water quality by 3D deterministic model. Water Res. 2013 (47), 1637e1641.
Chan, S.N., Thoe, W., Choi, K.W., Lee, J.H.W., 2013b. Application of 3D deterministic
model on marine beach water quality management in Hong Kong, China. In:
Proc. Of 35th IAHR World Congress, Chengdu, China, September 8-13, 2013.
Chen, B., Duan, J.C., Mai, B.X., Luo, X.J., Yang, Q.S., Sheng, G.Y., Fu, J.M., 2006. Dis-
tribution of alkylphenols in the Pearl River delta and adjacent northern south
China sea, China. Chemosphere 63, 652e661.
Choi, K.W., Chan, S.N., Wong, K.T.M., Lee, J.H.W., 2012. WATERMAN Territorial and
Regional Model for Hong Kong Waters. Project WATERMAN Technical Note TN-
2012-01. Croucher Laboratory of Environmental Hydraulics, the University of
Hong Kong, p. 2012.
Choi, K.W., Lee, J.H.W., 2007. Distributed entrainment sink approach for modelling
mixing and transport in intermediate field. J. Hydraulic Eng. 133, 804e815.
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), 3705e3711.
Christiansen, T., Wiberg, P.L., Milligan, T.G., 2000. Flow and sediment transport on a
tidal salt marsh surface. Estuarine, Coast. Shelf Sci. 50, 313e331.
Clarke, K.R., Gorley, R.N., 2006. PRIMER V6: User Manual/Tutorial. PRIMER-E, Ply-
mouth, 192pp.
EC (European Council), 2008. Directive 2008/56/EC Establishing a Framework for
Community Action in the Field of Marine Environmental Policy (Marine Strat-
egy Framework Directive). OJ L164, 25.6.2008, pp. 19e39.
EC, 2010. Commission Decision on Criteria and Methodological Standards on Good
Environmental Status of Marine Waters (Notified under Document C (2010)
5956), 2010/477/EU, L232/14, 2.9.2010.
EU, 2002. Risk Assessment Report. 4-Nonylphenol (Branched) and Nonylphenol.
European Commission. EUR 20387 EN.
EU, 2010. Updated European Union Risk Assessment Report. 4,40e Iso-
propylidenediphenol (BisphenoleA). European Commission. EUR 24588 EN.
394 E.G. Xu et al. / Environmental Pollution 233 (2018) 387e394
Ferguson, P.L., Iden, C.R., Brownawell, B.J., 2001. Distribution and fate of neutral
alkylphenol ethoxylate metabolites in a sewage impacted urban estuary. En-
viron. Sci. Technol. 35, 2428e2435.
Fu, J.M., Mai, B.X., Sheng, G.Y., Zhang, G., Wang, X.M., Peng, P.A., Xiao, X.M., Ran, Y.,
Cheng, F.Z., Peng, X.Z., Tang, U.W., 2003. Persistent organic pollutants in envi-
ronment of the Pearl River Delta, China, an overview. Chemosphere 52,
1411e1422.
Funakoshi, G., Kasuya, S., 2009. Influence of an estuary dam on the dynamics of
bisphenol A and alkylphenols. Chemosphere 75 (4), 491e497.
Gong, J., Ran, Y., Chen, D., Yang, Y., Zeng, E.Y., 2012. Association of endocrine-
disrupting chemicals with total organic carbon in riverine water and sus-
pended particulate matter from the Pearl River, China. Environ. Toxicol. Chem.
31 (11), 2456e2464.
Grover, D.P., Zhou, J.L., Frickers, P.E., Readman, J.W., 2011. Improved removal of es-
trogenic and pharmaceutical compounds in sewage effluent by full scale
granular activated carbon, Impact on receiving river water. J. Hazard. Mater. 185,
1005e1011.
Hamrick, J., 1992. A Three-dimensional Environmental Fluid Dynamics Computer
Code: Theoretical and Computational Aspects. The College of William and Mary,
Virginia Institute of Marine Science. Special Report 317.
Isobe, T., Nishiyama, H., Nakashima, A., Takada, H., 2001. Distribution and behavior
of nonylphenol, octylphenol, and nonylphenol monoethoxylate in Tokyo
metropolitan area, their association with aquatic particles and sedimentary
distributions. Environ. Sci. Technol. 35, 1041e1049.
Johnson, A.C., White, C., Besien, T.J., Jurgens, M.D., 1998. The sorption of octylphenol,
a xenobiotic oestrogen, to suspended and bed-sediments collected from in-
dustrial and rural reaches of three English rivers. Sci. Total Environ. 210/211,
271e282.
Jonkers, N., Laane, R.P.M., De Voogt, P., 2003. Fate of nonylphenol ethoxylates and
their metabolites in two Dutch estuaries, evidence of biodegradation in the
field. Environ. Sci. Technol. 37, 321e327.
Kueh, C.S.W., Lam, J.Y.C., 2008. Monitoring of toxic substances in the Hong Kong
marine environment. Mar. Pollut. Bull. 57, 744e757.
Lee, J.H.W., Chan, S.N., Choi, K.W., Chan, P.K., 2014. Application of the WATERMAN
real-time coastal water quality management system to environmental engi-
neering and control. In: DSD International Conference 2014, November 12-14,
2014, Hong Kong.
Lee, J.H.W., Choi, K.W., Chan, S.N., 2015. Field experiment of disinfection dosage
optimization for the Hong Kong harbour area treatment Scheme. In: Pro-
ceedings of 5th International Conference on Estuaries and Coasts (ICEC 2015)
November 2-4, 2015, Muscat, Oman, pp. 47e53.
Lee, J.H.W., Chu, V., 2003. Turbulent Jets and Plumess A Lagrangian Approach.
Kluwer Academic Publishers, Norwell, MA, p. 2003.
Leonard, L.A., Reed, D.J., 2002. Hydrodynamics and sediment transport through
tidal marsh canopies. J. Coast. Res. SI 36, 459e469.
Levy, G., Lutz, I., Krüger, A., Kloas, W., 2004. Bisphenol A induces feminization in
Xenopus laevis tadpoles. Environ. Res. 94, 102e111.
Li, X., Luan, T., Liang, Y., Wong, M., Lan, C., 2007. Distribution patterns of octylphenol
and nonylphenol in the aquatic system at Mai Po Marshes Nature Reserve, a
subtropical estuarine wetland in Hong Kong. J. Environ. Sci. 19, 657e662.
Li, Z., Li, D., Oh, J.R., Je, J.G., 2004. Seasonal and spatial distribution of nonylphenol in
Shihwa lake, Korea. Chemosphere 56 (6), 611e618.
Mao, Z., Zheng, X.F., Zhang, Y.Q., Tao, X.X., Li, Y., Wang, W., 2012. Occurrence and
biodegradation of nonylphenol in the environment. Int. J. Mol. Sci. 13 (1),
491e505.
Meral, R., Degirmenci, H., Gencoglan, C., Akyuz, A., Sesveren, S., 2008. International
Meeting on Soil Fertility Land Management and Agroclimatology. Turkey, vol.
2008, pp. 271e277.
Morton, B., 1995. An Introduction to the Cape D’ Aguilar Marine Reserve, Hong
Kong. Hong Kong University Press, Hong Kong.
Naylor, C.G., 1995. Environmental fate and safety of nonylphenol ethoxylates. Text.
Chem. Colorists 27, 29e33.
Naylor, C.G., Mieure, J.P., Adams, W.J., Weeks, J.A., Castaldi, F.J., Ogle, L.D.,
Romano, R.R., 1992. Alkylphenol ethoxylates in the environment. J. Am. Oil
Chem. Soc. 69, 695e703.
Nice, H.E., Thorndyke, M.C., Morritt, D., Steele, S., Crane, M., 2000. Development of
Crassostrea gigas larvae is affected by 4-nonylphenol. Mar. Pollut. Bull. 40,
491e496.
Niven, S.J., Snape, J., Hetheridge, M., Evans, M., Mcevoy, J., Sutton, P.G., Rowland, S.J.,
2001. Investigations of the origins of estrogenic A-ring aromatic steroids in UK
sewage treatment works effluents. Analyst 126, 285e287.
Ran, Y., Huang, W., Rao, P.S.C., Liu, D., Sheng, G., Fu, J., 2002. The role of condensed
organic matter in the nonlinear sorption of hydrophobic organic contaminants
by a peat and sediments. J. Environ. Qual. 31 (6), 1953e1962.
Rice, C.P., SchmitzeAfonso, J.S., LoyoeRosales, J.E., et al., 2003. Alkylphenol and
alkylphenol-ethoxylates in carp, water and sediment from the Cuyahoga River,
Ohio. Environ. Sci. Technol. 37, 3447e3754.
Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M., 1993. Environmental Organic
Chemistry, vol. 1993. John Wiley & Sons, New York, 275pp.
Shao, B., Hu, J.Y., Yang, M., 2002. Determination of nonylphenol ethoxylates in the
aquatic environment by normal phase liquid chromatographyeelectrospray
mass spectrometry. J. Chromatogr. A 950 (1), 167e174.
Von Saal, F., Cooke, P.S., Buchanan, D.L., Palanza, P., Thayer, K.A., Nagel, S.C.,
Parmigiani, S., Welshons, W.V., 1998. A physiologically based approach to the
study of bisphenol A and other estrogenic chemicals on the size of reproductive
organs, daily sperm production, and behavior. Toxicol. Industrial Health 14,
239e260.
White, R., Jobling, S., Hoare, S.A., Sumpter, J.P., Parker, M.G., 1994. Environmentally
persistent alkylphenolic compounds are estrogenic. Endocrinology 135,
175e182.
Xu, E.G., Ho, P.W.L., Tse, Z., Ho, S.L., Leung, K.M.Y., 2016. Revealing ecological risks of
priority endocrine disrupting chemicals in four marine protected areas in Hong
Kong through an integrative approach. Environ. Pollut. 215, 103e112.
Xu, E.G., Leung, K.M.Y., Morton, B., Lee, J.H.W., 2015a. 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, 269e281.
Xu, E.G., Liu, S., Ying, G.G., Zheng, G.J., Lee, J.H.W., Leung, K.M.Y., 2014. The occur-
rence and ecological risks of endocrine disrupting chemicals in sewage efflu-
ents from three different sewage treatment plants, and in natural seawater
from a marine reserve of Hong Kong. Mar. Pollut. Bull. 85 (2), 352e362.
Xu, E.G., Morton, B., Lee, J.H.W., Leung, K.M.Y., 2015b. Environmental fate and
ecological risks of nonylphenols and bisphenol A in the Cape D'Aguilar marine
reserve, Hong Kong. Mar. Pollut. Bull. 91 (1), 128e138.
Xu, J., Wang, P., Guo, W., Dong, J., Wang, L., Dai, S., 2006. Seasonal and spatial dis-
tribution of nonylphenol in Lanzhou reach of yellow River in China. Chemo-
sphere 65, 1445e1451.
Yang, G.P., Ding, H.Y., Cao, X.Y., Ding, Q.Y., 2011. Sorption behaviour of nonylphenol
on marine sediments, effect of temperature, medium, sediment organic carbon
and surfactant. Mar. Pollut. Bull. 62, 2362e2369.
Yoshida, H., Kudari, S., Hori, T., Sugiyama, M., 2009. Distribution of particulate
nonylphenol in lake Biwa, Japan. Water Air Soil Pollut. 200, 267e276.
Zhang, Y., Zhou, J.L., 2008. Occurrence and removal of endocrine disrupting
chemicals in wastewater. Chemosphere 73, 848e853.
Zhao, J.L., Ying, G.G., Chen, F., Liu, Y.S., Wang, L., Yang, B., Liu, S., Tao, R., 2011. Es-
trogenic activity profiles and risks in surface waters and sediments of the Pearl
River system in South China assessed by chemical analysis and in vitro bioassay.
J. Environ. Monit. 13 (4), 813e821.