See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/320381708

Large-scale multi-stage constructed wetlands

for secondary effluents treatment in northern
China: Carbon dynamics

Article in Environmental Pollution · October 2017

DOI: 10.1016/j.envpol.2017.09.048

CITATIONS READS

0 100

5 authors, including:

Haiming Wu Jinlin Fan

Northwest A & F University Shandong University
41 PUBLICATIONS 632 CITATIONS 27 PUBLICATIONS 575 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Study on the enhanced biological nitrogen removal mechanisms in constructed wetlands depending
on oxygen and organic substances release from plants View project

All content following this page was uploaded by Haiming Wu on 19 October 2017.

The user has requested enhancement of the downloaded file.

 Environmental Pollution xxx (2017) 1e10

Contents lists available at ScienceDirect

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

Large-scale multi-stage constructed wetlands for secondary effluents

treatment in northern China: Carbon dynamics*
Haiming Wu a, b, Jinlin Fan c, Jian Zhang b, *, Huu Hao Ngo d, Wenshan Guo d
a
College of Natural Resources and Environment, Northwest A&F University, Yangling, Shaanxi 712100, PR China
b
Shandong Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science & Engineering, Shandong University,
Jinan 250100, PR China
c
National Engineering Laboratory of Coal-Fired Pollutants Emission Reduction, Shandong University, Jinan 250061, PR China
d
School of Civil and Environmental Engineering, University of Technology Sydney, Broadway, NSW 2007, Australia

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

Article history: Multi-stage constructed wetlands (CWs) have been proved to be a cost-effective alternative in the
Received 3 May 2017 treatment of various wastewaters for improving the treatment performance as compared with the
Received in revised form conventional single-stage CWs. However, few long-term full-scale multi-stage CWs have been performed
12 September 2017
and evaluated for polishing effluents from domestic wastewater treatment plants (WWTP). This study
Accepted 16 September 2017
Available online xxx
investigated the seasonal and spatial dynamics of carbon and the effects of the key factors (input loading
and temperature) in the large-scale seven-stage Wu River CW polishing domestic WWTP effluents in
northern China. The results indicated a significant improvement in water quality. Significant seasonal and
Keywords:
Constructed wetlands
spatial variations of organics removal were observed in the Wu River CW with a higher COD removal
Secondary effluents efficiency of 64e66% in summer and fall. Obvious seasonal and spatial variations of CH4 and CO2
Organics removal emissions were also found with the average CH4 and CO2 emission rates of 3.78e35.54 mg m
2 d
1 and
Methane 610.78e8992.71 mg m
2 d
1, respectively, while the higher CH4 and CO2 emission flux was obtained in
Carbon dioxide spring and summer. Seasonal air temperatures and inflow COD loading rates significantly affected or-
ganics removal and CH4 emission, but they appeared to have a weak influence on CO2 emission. Overall,
this study suggested that large-scale Wu River CW might be a potential source of GHG, but considering
the sustainability of the multi-stage CW, the inflow COD loading rate of 1.8e2.0 g m
2 d
1 and tem-
perature of 15e20  C may be the suitable condition for achieving the higher organics removal efficiency
and lower greenhouse gases (GHG) emission in polishing the domestic WWTP effluent. The obtained
knowledge of the carbon dynamics in large-scale Wu River CW will be helpful for understanding the
carbon cycles, but also can provide useful field experience for the design, operation and management of
multi-stage CW treatments.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction the majority of untreated wastewater directly into urban rivers is a

common practice in many cities and small towns, which leads to
The issue of water scarcity and deterioration has become a the serious pollution of the river basins (Wu et al., 2015a). In
serious concern in the world, and this situation is becoming worse addition, in recent years great concern has been growing regarding
with the rapid urbanization, inadequate water/wastewater purifi- to the treated effluents, typically effluents from wastewater treat-
cation and management especially in developing countries in past ment plants (WWTP) (Vivant et al., 2016; Wu et al., 2017). Such
decades (Greenway, 2005). Due to lack of convenient and effective wastewaters contain excessive organic pollutants and nutrients,
wastewater treatments with lower cost (such as construction in- which are not properly/partially treated, are simply discharged into
vestment, operation costs and energy consumption), discharging rivers and estuaries, which may impact aquatic ecosystem health
and then pose potential risks to human health (Matamoros et al.,
2008; Wu et al., 2016a,b). In order to tackle this environmental
*
This paper has been recommended for acceptance by Dr. Jorg Rinklebe. concern, additional effective treatments could be employed to
* Corresponding author. purify WWTP effluents and improve the urban river water quality.
E-mail address: [email protected] (J. Zhang).

https://doi.org/10.1016/j.envpol.2017.09.048
0269-7491/© 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
2 H. Wu et al. / Environmental Pollution xxx (2017) 1e10

In fact, constructed wetlands (CWs), as an efficient and sus- such as dairy, tannery or domestic wastewater, however, further
tainable ecological treatment technology, have been proven to be studies reporting their treatment performance on the treatment of
an effective alternative to traditional wastewater treatment sys- effluent from WWTPs are still lacking. Additionally, few long-term
tems. The typical CW is generally comprised of vegetation, sub- full-scale multi-stage CW systems have been constructed and
strates, soils, microorganisms and water, and it can remove various evaluated for organic pollutant purification and processes of
pollutants (such as organics and nutrients) from wastewater by WWTP effluents. What is more, considering the sustainability and
means of physical, chemical, and biological mechanisms (microbial the potential of greenhouse gases (GHG) mitigation, not enough
degradation, plant uptake, sorption, sedimentation, filtration, pre- data are currently available on the dynamics and influencing factors
cipitation and volatilization etc.) (Saeed and Sun, 2012; Wu et al., of GHG emissions in the large-scale multi-stage CWs for polishing
2015a). Particularly, microbial removal mechanism plays a key WWTPs effluent.
role in the removal of pollutant in CWs (Mitsch et al., 2013; Meng Therefore, the aim of this study was to evaluate the carbon
et al., 2014). Those processes can be influenced by various envi- dynamics of a large-scale seven-stage FWS CW polishing secondary
ronmental conditions and operating parameters such as dissolved effluents from a conventional domestic WWTP in northern China.
oxygen (DO), hydraulic retention time (HRT), water depth, inflow Specific objectives of this study are: (i) to assess the removal per-
loading, water temperature, season and vegetation (Saeed and Sun, formance and variability of organics in the seven-stage FWS CW for
2012). According to the hydrological condition, CWs can typically polishing WWTPs effluent; (ii) to evaluate spatial and seasonal
be divided into free water surface (FWS) and subsurface flow (SSF) variation of methane (CH4) and carbon dioxide (CO2) emission in
wetlands with different technical characteristics. SSF CWs can be this multi-stage CW system; and (iii) to investigate the influence of
further classified into vertical flow (VF) and horizontal flow (HF) the key factors (input loading and temperature) on organics
CWs. There is another classification for CWs based on the type of removal and CH4 and CO2 emission. The results from this study
wetland plants used. In the past several years, CWs have been would be helpful in designing multi-stage CW systems in polishing
widely applied for removing various pollutants in domestic sewage, WWTP effluent for urban river water quality improvement as a
agricultural wastewater, industrial effluent, mine drainage, landfill sustainable wastewater treatment technology.
leachate, urban runoff, and polluted river water (Wu et al., 2014). In
addition, given their low cost and ease operation, CWs have been 2. Material and methods
shown to be able to efficiently remove organic pollutants, nutrients
and harmful bacteria from WWTP effluents with aim to improve 2.1. Study site description
the water quality and conserve ecological environment of the urban
river (Chen et al., 2014; Morvannou et al., 2015). But most of the Fig. S1 presents the site map of the studied CW (the Wu River
previous studies were performed in microcosm-scale or pilot-scale large-scale CW), which located on the west side of the Yi River in
CW systems with a small surface area (Vivant et al., 2016; Wu et al., Linyi, Shandong province, northern China at a latitude of 34 510 -
2015b). Matamoros et al. (2008) investigated the removal of a va- 35 06 N and a longitude of 118 060 -118 280 E. Wu River large-scale
riety of organic pollutants in a full-scale surface flow CW fed with CW as a treatment system for improving river water quality was
secondary effluent from a conventional WWTP, and reported that directly constructed along the riverbed of the Wu River, and Wu
CW was efficient for removing 12 organic micropollutants from River as the downstream of Yi River and Xianni River is currently
WWTP effluent discharged into the River Besos of northeastern the main urban drainage channel to receive the treated effluents
Spain. Chen et al. (2014) studied the removal efficiencies and the from a municipal WWTP in Linyi. The climate condition of the study
kinetics of disinfection byproducts in SSF CWs treating secondary site is warm temperate continental monsoon climate with a mean
effluent, and a high removal efficiently removed of >90% was annual temperature 14.1  C and an average precipitation of
achieved in laboratory-scale SSF CWs. In a recent study using FWS 818.8 mm, respectively.
CWs receiving secondary effluent from a France WWTP, Vivant et al.
(2016) found that FWS CWs could be an efficient treatment for 2.2. Large-scale system design and operation
extended-spectrum beta-lactamase-producing E. coli disinfection
of wastewater. However, it should be indicated that the treatment The Wu River large-scale CW is designed as a multi-stage CW
capacity of a single basic CW was still limited for polishing urban and comprises a rubber dam, a distribution ditch, a sequence of
rivers receiving large inputs of WWTP effluent. Concurrently, the seven stages of FWS CWs which are interconnected by using 500-
more application of multi-stage CWs by combining different types mm concrete pipes and an outlet (Fig. 1). In order to enhancing
of CWs has increased particularly for achieving the greatest po- reoxygenation or reaeration in CW system, several concrete over-
tential for wastewater polishing and more stable pollutant removal flow weirs were designed between the fourth wetland unit and the
(Wu et al., 2015b; Kato et al., 2013). This type multi-stage CW has fifth wetland unit of the CW system. The secondary effluent from
been proved to substantially improve the treatment performance the municipal WWTP as the influent was directly held by the
as compared with the conventional single-stage CWs (Vymazal, rubber dam, and flowed into the following FWS CW systems
2013a). Kato et al. (2013) designed the six multi-stage CWs for through the distribution ditch. Finally, the effluent from the CW
treating high-content wastewater in cold climates of northern system flowed into the Yi River through the outlet. The total land
Japan, and the satisfactory purification was obtained for organics, area of the system was about 8660000 m2 (over 15000 m long and
nitrogen and phosphorus. Jia et al. (2014) constructed a four-stage 120e320 m wide) with the treatment capability of 380000 m3 d
1.
wetland system for treating a heavily polluted river, and significant A natural soil layer of the Wu River was used as substrates for the
improvement in the water quality was observed. Moreover, several CW system, and the CW system had a 30e50 cm depth of water
important studies on multi-stage CWs have been conducted to treat above the substrate. The theoretical hydraulic retention time (HRT)
emerging organic contaminants (such as personal care products was around 7 d, and average hydraulic load was 3.5 cm d
1. The CW
and pharmaceuticals), and demonstrated the great capacity of a system was planted with a diversity of macrophyte types which
multi-stage wetland system as a cost-effective alternative or sup- was mainly the dominant native wetland plants included Phrag-
plementary to conventional WWTP (Avila et al., 2014a, 2014b). mites australis (P. australis), Typha orientalis (T. orientalis), Zizania
Above previous studies have convincingly proved the benefits of latifolia (Z. latifolia), Nelumbo nucifera (N. nucifera), Nymphaea tet-
multi-stage CWs in the treatment of various types of wastewater, ragona (N. tetragona), Potamogeton crispus (P. crispus), Lemna minor

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
 H. Wu et al. / Environmental Pollution xxx (2017) 1e10 3

Fig. 1. Schematic diagram of the Wu River multi-stage constructed wetland.

(L. minor) and water hyacinth. The Wu River CW was constructed 2.3. Field sampling and laboratory analysis
completely in February 2010 and then operated from May 2010
after a commissioning period (about three months). During the 6- Water samples were collected and analyzed weekly from
year operation, in order to realize the optimal treatment capabil- January 2014 to December 2014 to evaluate the efficiency and
ities of the CW system, a manual harvesting was performed variability of the Wu River large-scale CW on organics removal. In
annually in the Wu River large-scale CW in this study, but there was this study, water samples were collected from nine sampling sites
no mechanized harvesting. During the cold period (from November of the CW including 1# (Inflow), 2#, 3#, 4#, 5#, 6#, 7#, 8# and 9#
to December) of every year, when the plants (mainly the emergent (Outflow) at a depth of 0.2 m using a sampling bottle (Fig. 1), and
aquatic plants such as P. australis and T. orientalis) in the CW the detailed coordinates and description of each sampling site of
become yellow and ripe in late autumn, they were ready for har- the CW were described in Table 2. All water samples were trans-
vest. Specially, the aboveground parts of plants were harvested ferred immediately to the lab and stored at 4  C before analysis.
manually by cutting their upper parts at about 20 cm above the Collected water samples were analyzed for chemical oxygen
water level. The harvested biomass from the CW was used as a raw demand (COD), ammonium (NHþ 4 -N), total nitrogen (TN), nitrate
material for paper production in the local paper industry. According (NO


3 -N) and nitrite (NO2 -N) and total phosphorus (TP). Tempera-
to the long-term monitor from November 2010 to November 2014 ture, pH and dissolved oxygen (DO) were measured in situ by a DO
on the quality of the influent of the Wu River CWs, the character- meter (HQ 30d 53LED™ HACH USA) and a glass pH meter (SG2-T
istics of the influent during the operational period is showed in SevenGo pro™ MTD, Switzerland). All of the above analyses were
Table 1. The average influent concentrations of COD, NHþ

4 -N, NO3 - performed according to standard methods (APHA, 2005).
N, NO
2 -N, TN and TP were 41.94 mg L
1
, 4.61 mg L
1
, 5.72 mg L
1
, Gas samples were collected from above nine sampling sites of
0.36 mg L
1, 10.69 mg L
1 and 0.51 mg L
1, respectively. the CW weekly from January 2014 to December 2014, and were

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
4 H. Wu et al. / Environmental Pollution xxx (2017) 1e10

Table 1
Characteristics of influent and effluent of the Wu River constructed wetland and design criteria for water quality.
a b
Water quality parameters Influent Effluent Effluent requirement (Grade III)

COD (mg L
1) 41.94 ± 8.17 18.44 ± 2.41 20

NHþ
1
4 -N (mg L ) 4.61 ± 1.15 1.06 ± 0.92 1.0
NO

1
3 -N (mg L ) 5.72 ± 1.01 4.01 ± 2.15 e
NO

1
2 -N (mg L ) 0.36 ± 0.19 0.13 ± 0.08 e
TN (mg L
1) 10.69 ± 1.06 5.19 ± 2.18 1.0
TP (mg L
1) 0.51 ± 0.11 0.01 ± 0.01 0.2
pH 7.63 ± 0.11 8.71 ± 0.12 6e9
DO (mg L
1) 7.13 ± 2.17 19.41 ± 2.72 5.0
Water temperature ( C) 18.63 ± 9.51 19.05 ± 11.74 e
a
The final effluent after the 7 FWS CW treatments.
b
According to the environmental functions and protective objectives of surface waters, all surface water in Mainland China are divided into five grades, and Grade III means
the surface water for drinking purpose, water for aquaculture (common species).

Table 2
The coordinates and description of sampling sites of the Wu River constructed wetland in the present study.

Sampling sites Coordinates Environmental conditions

Latitude Longitude

1# 34 520 12.2900 N 118 210 01.2300 E With a maximum depth of 0.5 m, and covered with sparse L. minor in spring and summer, but with
no vegetation cover in autumn and winter.
 0 00  0 00
2# 34 52 01.43 N 118 20 44.01 E With a depth of 0.3e0.5 m, and covered with dense L. minor in spring and summer, but with
no vegetation cover in autumn and winter.
3# 34 510 45.1600 N 118 200 32.2400 E Open water with a fast water flow rate; Covered with water hyacinth in summer, and withering
and dying off in autumn and winter.
4# 34 510 11.9400 N 118 200 12.7800 E Covered with P. australis (2e3 m in height) during the growing season, but decaying and
dying off in winter.
5# 34 500 40.9200 N 118 190 59.8400 E Covered with dense T. orientalis (about 2 m in height) during the growing season and with
a slow water flow rate.
6# 34 500 16.3000 N 18 190 46.7900 E The mixed area of N. nucifera, N. tetragona and P. crispus, and, but decaying and dying off in winter.
7# 34 490 50.9700 N 118 190 14.3400 E Open water with the vegetation of submerged macrophytes, and N. tetragona were found in a small area.
8# 34 490 28.0200 N 118 180 34.5600 E Covered with P. australis and T. orientalis with the vegetation height 2e3 m during the growing
season, but almost decaying and dying off in winter.
9# 34 480 56.2300 N 118 180 04.5300 E Open water with a slow water flow rate and low water depth, but submerged macrophytes were
found in some area.

analyzed to evaluate the variation of CH4 and CO2 emission from 3. Results and discussion
the CW. Gas sampling in this study was done using a commercially-
available surface emission isolation flux chamber during the 3.1. Variations of environmental parameters and water quality
experimental period. The surface emission isolation flux chamber
was made of a cylindrical stainless steel flux hood with the diam- Fig. 2 shows the seasonal variation of the average air tempera-
eter, volume and sampling area of 0.41 m, 30 L and 0.13 m2, ture and relative humidity from January 2014 to December 2014.
respectively. Gas samples were collected in 100 ml polypropylene There was an obvious seasonal variation for air temperature which
syringes at 10 min intervals for 30 min, and the details of collecting was ranged from
10.1  C to 36.5  C in 2014. In addition, relative
steps of gas samples can be found in the previous studies (Ren et al.,
2013). The concentrations of CH4 and CO2 were determined using
the gas chromatography (SP-6890, China) equipped with a flame
ionization detector (GC-FID) and stainless steel packed columns
(GDX502). The operating conditions for the GC were: 375  C
reformer temperature, 40  C oven temperature and 200  C detector
temperature. The carrier gas was ultra-high purity N2
(30 mL min
1). CH4 and CO2 fluxes (mg m
2 d
1) were determined
from the increase in concentration in the chambers over time with
linear regression analysis according to the described method in the
previous studies (Wang et al., 2011).
Dissolved gas sampling was also performed to investigate the
variation of the dissolved CH4 and CO2 concentration in the water of
the CW. The dissolved gas samples were collected and analyzed
using the headspace air method as described by Ren et al. (2013).
Water (30 ml) and argon gas (70 ml) were injected into a 100 ml
preevacuated vial, and 1 ml HgCl2 (20 mM) was added to inhibit
microbial activity. The water was collected at the same time as the
gas sampling. After a shaking (1 h), the vial was left at room tem-
perature for 10 min. The headspace gas phase in the vial was
collected for determining dissolved CH4 and CO2 concentration Fig. 2. Variation of air temperature and relative humidity during the experimental
based on Henry’ s Law and mass balance (Ren et al., 2013). period.

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
 H. Wu et al. / Environmental Pollution xxx (2017) 1e10 5

humidity during the experimental period varied significantly, and and macrophyte root transfer into the plant rhizosphere, and
the average relative humidity was 54.8%, with the higher value in anaerobic organics removal can proceed inside the sediment,
July (74.2%) and the lower value in January (35.6%). The variation of lacking oxygen (Saeed and Sun, 2012). In this study, seasonal and
average water temperature in the CW, which ranged from 3.9  C in spatial variations of COD removal in the Wu River CW over a period
Winter to 31.1  C in Summer, was similar to air temperature during of one year (from January 2014 to December 2014) were evaluated,
the experimental period, and the mean seasonal water temperature and the average concentrations of COD in the nine sites of the Wu
was slightly lower than the mean seasonal air temperature except River CW in different seasons are illustrated in Fig. 3. The data
for Winter (Table S1). Table 1 gives the effluent characteristics (the demonstrate a sustained and stable removal of organics from the
final effluent after the 7 FWS CW treatments) of the Wu River CW secondary effluent of the municipal WWTP. However, average
that were measured during the experimental period, and it showed effluent concentrations of COD of the Wu River CW changed
that the final average effluent concentrations of COD, NHþ

4 -N, NO3 - significantly with the different COD removal efficiency in different
N, NO
2 -N, TN and TP were 18.44 mg L
1
, 1.06 mg L
1
, 4.01 mg L
1
, seasons of the one year. Furthermore, concentrations of COD varied
0.13 mg L
1, 5.17 mg L
1 and 0.01 mg L
1, respectively. The high considerably within each sampling site. In spring (March to May),
oxygen level (DO > 10 mg L
1) in the effluent indicates the pre- the concentration of COD gradually decreased along the Wu River
dominance of aerobic conditions. On the whole, the main water CW from 50.61 mg L
1 at the inlet to 20.69 mg L
1 in the outlet,
quality parameters (i.e. COD, NHþ 4 -N, pH and DO) can comply with indicating a removal efficiency of 59.11% (Fig. 3a). Concentrations of
the Grade III of Environmental Quality Standards for Surface Water COD also varied considerably within each sampling location. A
(GB3838-2002) in China, which indicated a significant improve- significant decrease was observed in the third sampling site (i.e. the
ment in water quality of secondary effluents from a conventional first wetland unit), and then there was a slightly fluctuation along
WWTP by Wu River CW treatment. However, other parameters the CW. In summer (June to September), the similar dynamics of
such as TN and TP could not meet the Grade III. In addition, there COD removal was obtained in the Wu River CW as those observed
was seasonal fluctuation and in average removal performance of during the spring season. COD concentrations of the effluent was
pollutants, which may be due to the seasonal variation of some 16.53 mg L
1 which resulting a higher removal efficiency of 65.98%
important influencing factors such as macrophyte species and (Fig. 3b). However, a significant decrease in the concentration of
density, diversity of plants, loading rates and climatic condition COD, reaching 15.57 mg L
1, was observed in the fourth sampling
(especially air temperature) during the whole experimental period site (i.e. the second wetland unit), and then the value of COD
(Brisson and Chazarenc, 2009; Vymazal, 2013b; Wu et al., 2011). concentration increased weakly with the stable variation in the
following sampling sites of the CW. But these differences were not
3.2. Seasonal and spatial variations of organics removal statistically significant. It was found that CW system also showed
better COD removal effects in autumn (October to November), and
For organic pollutants removal, degradation aerobically and the effluent concentration was 15.54 mg L
1 which indicating a
anaerobically is the classic removal routes in CW treatments. Aer- removal efficiency of 63.86%. Even if the air temperature in this
obic degradation can be accomplished based on the available ox- season began decrease, COD concentrations in the different sam-
ygen via atmospheric oxygen diffusion, convection (wind effect), pling sites except inflow site were all stabilized at the range of

Fig. 3. Seasonal and spatial variation of COD concentrations in the Wu River multi-stage constructed wetland.

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
6 H. Wu et al. / Environmental Pollution xxx (2017) 1e10

8.72e22.12 mg L
1 (Fig. 3c), especially in the seventh sampling site

(nearly the fifth wetland unit) with lowest COD concentration.
During the coldest season (i.e. December, January and February),
COD concentration in different sampling sites had a distinct fluc-
tuating trend of early gradual decrease and later sudden increase
along the CW (Fig. 3d). Unlike in the warm months, a significant
increase in the concentration of COD, reaching 29.22 mg L
1, was
observed in the sixth wetland unit. Then, the effluent COD con-
centration stabilized at 21.22 mg L
1, which basically meets the
Grade III of Environmental Quality Standards for Surface Water
(GB3838-2002) in China.
On the whole, results in this study showed that significant
seasonal influences on organics removal were presented in the
wetland system, and the mean removal efficiencies were obvious
pronounced in spring, summer and autumn compared to winter.
Moreover, there were obvious spatial variations in seasonal or-
ganics removal effects. Organics in CWs can mainly be degraded by
both aerobic and anaerobic microbial processes as well as by
sedimentation and filtration of particulate organic compounds
(Saeed and Sun, 2012). Organics removal processes in FWS CWs
may be generally influenced by the different environmental pa-
rameters and operational conditions such as: temperature, loading
rate, retention time, vegetation and harvesting (Wu et al., 2011,
2014; Saeed and Sun, 2012). Particularly in order to investigate
whether the efficiency of organics removal of the multi-stage CW
was impacted by seasonal variations in temperature and loading
rate, COD removal by the CW observed for each season was
compared to average seasonal air temperatures and inflow COD
loading rates. Fig. 4 presents the effect of seasonal variations in air
temperature and inflow COD loading rate on the rate of COD
removal in the CW. Linear relationship between air temperature
and COD removal rate during different seasons was found to be
significant (R2 ¼ 0.98, p < 0.01). As is shown in Fig. 4a, COD removal
rate increased significantly with the air temperature rising from
2  C (Winter) to 18  C (Spring), but there was less significant Fig. 4. Effect of seasonal variations in temperature (a) and inflow loading rate (b) on
the rate of organics removal in the Wu River multi-stage constructed wetland. R2
variation in COD removal rate when air temperature was between represents the coefficient of determination of the linear model between seasonal
18  C (Spring) and 33  C (Summer). This suggested that a proper air temperature (inflow loading rate) and organics removal.
temperature range for COD removal was around 18e33  C under
certain conditions. The possible explanation might be that organics
removal mechanisms in CWs was mainly microbial processes by COD removal rate in CWs rose clearly with the increase of loading
microbial community, and microbial activity could be directly rate of influent wastewater. These results are consistent with pre-
affected by water temperature which was caused by air tempera- vious research which reported that increase of organic loading
ture variation (Saeed and Sun, 2012; Lv et al., 2017). On the other often coincided with greater organics removal rate (Sun and Saeed,
hand, air temperature could significantly affect the growth and 2009; Saeed and Sun, 2011). However, excessive loading rates
establishment of wetlands plants which in turn influenced water would cause accumulation of organic matter and lower removal
purification (Rai et al., 2013; Meng et al., 2014). In the CW, wetland efficiencies in CWs (Landry et al., 2007). For polishing secondary
plants and microbes grow well in warm season, but during the effluent of the WWTP in this study, the inflow COD loading rate of
winter season, plant litter was decomposed resulting in the release 1.8e2.0 g m
2 d
1 could be appropriate for achieving the higher
of organic matter into water. Moreover, the diversity of different organics removal efficiency in the multi-stage CW. But it should be
macrophyte types in different treatment units would cause signif- noted that it was difficult to identify an optimal loading rate that
icant spatial dynamics of microbial community in CWs (Button corresponded to the highest organics removal efficiency in CWs
et al., 2016; Lv et al., 2017). The influence of evapotranspiration under complicated conditions.
should also be considered. In this study, it is estimated that
evapotranspiration from different CW cells was high in summer 3.3. Seasonal and spatial variations of CH4 emission
(approximately 8 mm d
1), and evaporation values in CW cells with
more vegetation were higher than that in CW cells with more open In order to understand the carbon dynamics in a large-scale CW
water area. Evapotranspiration might have a negative effect on COD system for the maximum environmental benefit, a measurement of
removal rates because it reduces the amount of water and increases seasonal and spatial CH4 emissions from the Wu River CW over a
the concentration of pollutants (Białowiec et al., 2014). As is shown period of one year were conducted. Fig. 5 presents the average
in Fig. 4b, significant line relationship between COD loading rate emissions of CH4 from the CW in different seasons throughout the
and COD removal rate was also observed (R2 ¼ 0.99, p < 0.01), long-term field study. There was significant seasonal variation in
which indicated a positive correlation among the influent waste- CH4 emission, with the average CH4 flux varying from 3.78 mg m
2
water quality and organics removal. It can be observed that the d
1 to 35.54 mg m
2 d
1. The higher flux of CH4 was observed in
wastewater discharged into the river was significantly more spring compared to summer (Fig. 5a and b), and a lower CH4 flux
contaminated during spring, summer and fall than in winter, and occurred in fall and winter (Fig. 5c and d). On the whole, the

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
 H. Wu et al. / Environmental Pollution xxx (2017) 1e10 7

Fig. 5. Seasonal and spatial CH4 fluxes and dissolved CH4 concentrations in the Wu River multi-stage constructed wetland.

average CH4 fluxes in spring and summer were 5e10 times higher study also indicated that the CH4 emission from the CW could be
than that in fall and winter. Average CH4 emission fluxes also had affected by air temperature which represented seasonal water
obvious spatial variations in different sampling sites of the CW. temperature directly. Fig. 6a shows the polynomial regression
Specially in spring, summer and fall, and there was strong evidence relationship between CH4 fluxes and air temperature (R2 ¼ 0.82,
of CH4 fluxes in the middle sites of the CW being significantly p < 0.01). It seemed that the rate of CH4 flux from CW increased
increased compared to inflow and outflow sites. However, the with the rise of air temperature, and this is in agreement with
insignificant spatial variation of CH4 fluxes was found in winter findings from earlier studies which indicated that temperature is
when the CW became a weak source of CH4. It should be noted that positively correlated with greenhouse gas emissions in CWs
when CH4 emissions were high, the variation in the emission rates (Maucieri et al., 2017). The possible reason may be that microbial
was also greatest. Both the minimum and maximum CH4 emission activity in wetlands generally increases at a proper temperature
fluxes were measured in warm seasons. Particularly, an obvious (Mander et al., 2014). Temperature can also affect vegetation pro-
uptake of CH4 was found in the outflow site in summer and winter, ductivity, cover, and biomass, which will then influence CH4
which indicating a net CH4 sink. The above results suggested that emission by transporting CH4 to the atmosphere through plant
seasons might have a remarkable effect on spatial variation of CH4 tissues directly (Mitsch et al., 2013; Murray et al., 2017). On the
emissions in the Wu River CW. The corresponding seasonal and other hand, it is well recognized that temperature affects directly
spatial change of dissolved CH4 concentrations was also observed plant photosynthesis and evapotranspiration, and thus increases
with the variation of the CH4 fluxes. In spring, the dissolved CH4 gas transportation, which would give a positive effect on CH4
concentration measured in the Wu River CW was gradually emission (Zhao et al., 2016). Pangala et al. (2010) investigated the
increasing along the CW from 0.02 mg L
1 in the inlet to response of CH4 emissions to temperature in CWs, and showed that
0.04 mg L
1 in the outlet (Fig. 5a). But compared with the variation CH4 emissions from CWs increased exponentially and significantly
in the spring season, the different dynamics of dissolved CH4 con- with water temperature from 10  C to 30  C. CH4 emission may be
centration in the Wu River CW was detected in summer, fall and influenced by other factors such as influent pollutants concentra-
winter, which showed a variation trend of a significant decrease in tions (Wang et al., 2008). Our results also indicated that influent
the initial sampling sites with a significant increase later, and then a COD loading rate significantly influence CH4 emission (Fig. 6b). As is
suddenly drop at the outflow sampling location. shown in Fig. 6b, a positive correlation between inflow loading rate
GHG emission from CWs can generally be influenced by various and CH4 emission rate was observed (R2 ¼ 0.97, p < 0.01). These
environmental conditions and operating parameters such as DO, results are in agreement with findings from earlier studies which
HRT, water depth, available organic carbon, temperature and indicated that a significant correlation between the inflow organics
vegetation (Wu et al., 2017; Mander et al., 2014). The results in this loading and CH4 emission values in CWs based on the multiple

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
8 H. Wu et al. / Environmental Pollution xxx (2017) 1e10

that the variation in the flux rates was greatest when CO2 emissions
were high. Both the minimum and maximum CO2 fluxes were
measured in summer with the CO2 flux varying
from
1019.72 mg m
2 d
1 to 42549.78 mg m
2 d
1. In addition,
strong CO2 emission occurred in the front and rear sites of the CW
while negative CO2 emission was found in the middle sites. The
evident spatial variation of CO2 fluxes was also found in spring, fall
and winter, however, the negative CO2 emission was only observed
at the outflow site. These results can be explained by the previous
research that plant autotrophic respiration and plant photosyn-
thesis increased with high biomass production, which may result a
higher net exchange of CO2 between CWs and the atmosphere
(Mander and Teiter, 2005; Mander et al., 2014). In addition, it is
clearly reported that the significant role of wetland plants, which
bring a large quantity of atmospheric CO2 into the CW, thereby
showing the small net GHG sink (Mitsch et al., 2013). The corre-
sponding seasonal and spatial change of dissolved CO2 concentra-
tions was also observed with the variation of the CO2 fluxes (Fig. 7).
In spring, the dissolved CO2 concentration measured in the Wu
River CW was significantly increasing along the CW from 0.12 g L
1
in the inlet to 0.74 g L
1 in the middle sites, and then decreased to
0.11 g L
1 in the outlet (Fig. 7a). However, the different dynamics of
dissolved CO2 concentration in the Wu River CW was detected in
summer, fall and winter, which showed a variation trend of a
gradual decrease from the initial sampling sites to the outflow
sampling location.
To investigate whether the CO2 emission of the CW was
impacted by seasonal variations in temperature and inflow loading
rates, the average CO2 emission in the CW observed for different
seasons was compared to average seasonal air temperatures and
inflow COD loading rates. As indicated in Fig. 6a, the polynomial
regression relationship between air temperature and CO2 emission
rates revealed that CO2 emission rate was associated with seasonal
air temperature (R2 ¼ 0.81, p < 0.01). A slight increase in rate of CO2
Fig. 6. Effect of seasonal variations in temperature (a) and inflow loading rate (b) on emission from the CW was presented under high temperature
emission fluxes of CH4 and CO2 from the Wu River multi-stage constructed wetland. R2
represents the coefficient of determination of the polynomial model between seasonal
conditions. The possible reason may be that microbial activity for
temperature (inflow loading rate) and emission of CH4 and CO2. organic degradation in CWs generally increases at a proper climatic
condition (such as temperature) (Mander et al., 2014). In addition,
temperature affects directly plant photosynthesis and respiration
regression analysis (Mander et al., 2014). However, it should be as well as evaporation, which influences directly CO2 emission
noted that there was some limit level for CH4 emission at a certain (Mitsch et al., 2013). This is in agreement with the previous results
loading rates. The possible reason is that the rate of methano- which reported that the CO2 flux from the wetlands might be
genesis might be inhibited by increased concentration of ammonia affected by the seasonal temperature variation (Mander and Teiter,
or accumulation of volatile fatty acids (Mata-Alvarez et al., 2000; 2005), and Trumbore et al. (1996) indicated that temperature had a
Wang et al., 2008). On the whole, for the multi-stage CW polish- strong influence on soil carbon dynamics regulating CO2 uptake or
ing secondary effluent of the WWTP in this study, the inflow COD emission to the atmosphere. A positive correlation between
loading rate of 1.8e2.0 g m
2 d
1 could be appropriate for mini- average air temperature and CO2 emission in the CWs was also
mizing CH4 emission with a higher organic removal efficiency. confirmed by Maucieri et al. (2014). But there was an evident dif-
ference in the importance of temperature in CO2 emissions from
3.4. Seasonal and spatial variations of CO2 emission different designed or operated CWs in diverse climate sites (Wu
et al., 2016a,b). Hence, the relationship between CO2 emissions
Although CO2 is a major GHG, there are only few studies and influent COD loading rates in the large-scale CW system in
reporting on CO2 emission in CWs (Mander et al., 2014; Wu et al., different seasons was assessed, but as shown in Fig. 6b, an insig-
2017). Therefore, the CO2 emissions from different sites of the Wu nificant correlation was observed between inflow loading rate and
River CW were also measured in this study for understanding the CO2 emission rate (R2 ¼ 0.52, p < 0.01). This appears to be due to an
variation of CO2 emission in the large-scale CW system for the obscure interaction between CO2 fluxes and organic loading rates
treatment of secondary effluent. As shown in Fig. 7, the emission when it is at low loading conditions, although the characteristics
rates of CO2 from the CW were significant different in different (i.e. biodegradability and loading) of the organic matter was the key
seasons during the one year. The average CO2 emission rates factor in influencing the intensity of organics removal processes in
were higher in spring (2889.41 mg m
2 d
1) and summer CWs (Saeed and Sun, 2012; Mander et al., 2014). Generally, for the
(8992.71 mg m
2 d
1), but lower in autumn (610.78 mg m
2 d
1) multi-stage CW polishing secondary effluent of the WWTP, the
and winter (1908.35 mg m
2 d
1). The average CO2 fluxes in spring proper inflow COD loading rate for a low CO2 emission and high
and summer were 2e15 times higher than that in fall and winter. organics removal efficiency is uncertain and worthy of further long-
The significant spatial variation in the average CO2 emission rates term investigation. On the other hand, it should also be noted that a
was also found in different sampling sites of the CW. It is indicated comprehensive analysis of the environmental and ecological

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
 H. Wu et al. / Environmental Pollution xxx (2017) 1e10 9

Fig. 7. Seasonal and spatial CO2 fluxes and dissolved CO2 concentrations in the Wu River multi-stage constructed wetland.

benefit would be required for the CW as a water purification Acknowledgements

technology when considering the aspect of the CW sustainability.
Specifically, the potential contribution of GHG (CH4 and CO2) We gratefully acknowledge financial support by the National
emission from the CW may reduce the environmental and Science Foundation of China (NSFC, Project No. 51508466, 51578321),
ecological benefits although an excellent organic removal perfor- and the Project Supported by Natural Science Basic Research Plan in
mance was achieved in this study. Therefore, a feasible method or Shaanxi Province of China (Program No. 2017JQ4024).
management for mitigating GHG emissions in CWs should be paid
special attention. Appendix A. Supplementary data

4. Conclusions Supplementary data related to this article can be found at

https://doi.org/10.1016/j.envpol.2017.09.048.
The long-term investigation of the large-scale multi-stage CW
for polishing domestic WWTP effluents was performed in northern References
China, and the results indicated that water quality of Wu River CW
could comply with the Grade III of Environmental Quality Stan- APHA, 2005. Standard Methods for the Examinations of Water and Wastewater,
twenty-first ed. APHA and AWWA and WEF, DC, Washington.
dards for Surface Water in China. However, Wu River CW might be a ~ a, B., Casado, M., Mita, L., Rivetti, C.,
Avila, C., Matamoros, V., Reyes-Contreras, C., Pin
potential emission source of CH4 (3.78e35.54 mg m
2 d
1) and CO2 Barata, C., García, J., Bayona, J.M., 2014a. Attenuation of emerging organic
(610.78e8992.71 mg m
2 d
1). Significant seasonal variations of contaminants in a hybrid constructed wetland system under different hydraulic
loading rates and their associated toxicological effects in wastewater. Sci. Total
organics removal and emission of CH4 and CO2 were observed in Environ. 470e471, 1272e1280.
the Wu River CW with a higher COD removal efficiency in summer Avila, C., Bayona, J.M., Martín, I., Juan, S.J., García, J., 2014b. Emerging organic
and fall and higher CH4 and CO2 emission rate in spring and sum- contaminant removal in a full-scale hybrid constructed wetland system for
wastewater treatment and reuse. Ecol. Eng. 80, 108e116.
mer. The rates of COD removal and CH4 emission increased signif- Białowiec, A., Albuquerque, A., Randerson, P.F., 2014. The influence of evapotrans-
icantly with the rise of air temperature and inflow COD loading piration on vertical flow subsurface constructed wetland performance. Ecol.
rates, but CO2 emission rate was observed to be insignificantly Eng. 67, 89e94.
Brisson, J., Chazarenc, F., 2009. Maximizing pollutant removal in constructed wet-
associated with air temperature and inflow COD loading rates. Our
lands: should we pay more attention to macrophyte species selection? Sci. Total
results will be helpful for understanding the dynamics and cycles of Environ. 47, 3923e3930.
carbon in large-scale multi-stage CWs, and further field experi- Button, M., Rodriguez, M., Brisson, J., Weber, K.P., 2016. Use of two spatially sepa-
ments will be required to optimize the design, operation and rated plant species alters microbial community function in horizontal subsur-
face flow constructed wetlands. Ecol. Eng. 92, 18e27.
management of multi-stage CW treatments and its environmental Chen, Y., Wen, Y., Tang, Z., Li, L., Cai, Y., Zhou, Q., 2014. Removal processes of
sustainability. disinfection byproducts in subsurface-flow constructed wetlands treating

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
10 H. Wu et al. / Environmental Pollution xxx (2017) 1e10

secondary effluent. Water Res. 51, 163e171. Ren, Y.G., Wang, J.H., Li, H.F., Zhang, J., Qi, P.Y., Hu, Z., 2013. Nitrous oxide and
Greenway, M., 2005. The role of constructed wetlands in secondary effluent treat- methane emissions from different treatment processes in full-scale municipal
ment and water reuse in subtropical and arid Australia. Ecol. Eng. 25, 501e509. wastewater treatment plants. Environ. Technol. 34, 2917e2927.
Jia, H., Sun, Z., Li, G., 2014. A four-stage constructed wetland system for treating Saeed, T., Sun, G., 2011. The removal of nitrogen and organics in vertical flow
polluted waterfrom an urban river. Ecol. Eng. 71, 48e55. wetland reactors: predictive models. Bioresour. Technol. 102, 1205e1213.
Kato, K., Inoue, T., Ietsugu, H., Koba, T., Sasaki, H., Miyaji, N., Kitagawa, K., Saeed, T., Sun, G., 2012. A review on nitrogen and organics removal mechanisms in
K.Sharma, P., Nagasawa, T., 2013. Performance of six multi-stage hybrid wetland subsurface flow constructed wetlands: dependency on environmental param-
systems for treating high-content wastewater in the cold climate of Hokkaido. eters, operating conditions and supporting media. J. Environ. Manage 112,
Jpn. Ecol. Eng. 51, 256e263. 429e448.
Landry, G.M., Chazarenc, F., Comeau, Y., Troesch, S., Brisson, J., 2007. Effects of Sun, G., Saeed, T., 2009. Kinetic modelling of organic matter removal in 80 hori-
artificial aeration, macrophyte species, and loading rate on removal efficiency in zontal flow reed beds for domestic sewage treatment. Process Biochem. 42,
constructed wetland mesocosms treating fish farm wastewater. J. Environ. Eng. 717e722.
Sci. 6, 409e414. Trumbore, S.E., Chadwick, O.A., Amundson, R., 1996. Rapid exchange between soil
Lv, T., Carvalho, P.N., Zhang, L., Zhang, Y., Button, M., Arias, C.A., Weber, K.P., Brix, H., carbon and atmospheric carbon dioxide driven by temperature change. Science
2017. Functionality of microbial communities in constructed wetlands used for 272, 393e396.
pesticide remediation: influence of system design and sampling strategy. Water Vivant, A.L., Boutin, C., Prost-Boucle, S., Papias, S., Hartmann, A., Depret, G.,
Res. 110, 241e251. Ziebal, C., Le Roux, S., Pourcher, A.M., 2016. Free water surface constructed
Mander, Ü., Dotro, G., Ebie, Y., Towprayoon, S., Chiemchaisri, C., Nogueira, S.F., wetlands limit the dissemination of extended-spectrum beta-lactamase pro-
Jamsranjav, B., Kasak, K., Truu, J., Tournebize, J., Mitsch, W.J., 2014. Greenhouse ducing Escherichia coli in the natural environment. Water Res. 104, 178e188.
gas emission in constructed wetlands for wastewater treatment: a review. Ecol. Vymazal, J., 2013a. The use of hybrid constructed wetlands for wastewater treat-
Eng. 66, 19e35. ment with special attention to nitrogen removal: a review of a recent devel-
Mander, U., Teiter, S., 2005. Emission of N2O, N2, CH4, and CO2 from constructed opment. Water Res. 47, 4795e4811.
wetlands for wastewater treatment and from riparian buffer zones. Ecol. Eng. Vymazal, J., 2013b. Emergent plants used in free water surface constructed wet-
25, 528e541. lands: a review. Ecol. Eng. 61, 582e592.
Mata-Alvarez, J., Mace , S., Llabre
s, P., 2000. Anaerobic digestion of organic solid Wang, J., Zhang, J., Xie, H., Qi, P., Ren, Y., Hu, Z., 2011. Methane emissions from a full-
wastes: an overview of research achievements and perspectives. Bioresour. scale A/A/O wastewater treatment plant. Bioresour. Technol. 102, 5479e5485.
Technol. 74, 3e16. Wang, Y., Inamori, R., Kong, H., Xu, K., Inamori, Y., Kondo, T., Zhang, J., 2008. In-
Matamoros, V., García, J., Bayona, J.M., 2008. Organic micropollutant removal in a fluence of plant species and wastewater strength on constructed wetland
full-scale surface flow constructed wetland fed with secondary effluent. Water methane emissions and associated microbial populations. Ecol. Eng. 2, 22e29.
Res. 42, 653e660. Wu, H., Zhang, J., Li, P., Zhang, J., Xie, H., Zhang, B., 2011. Nutrient removal in con-
Maucieri, C., Barberab, A.C., Vymazal, J., Borin, M., 2017. A review on the main structed microcosm wetlands for treating polluted river water in northern
affecting factors of greenhouse gases emission in constructed wetlands. Agr. China. Ecol. Eng. 37, 560e568.
For. Meteorol. 236, 175e193. Wu, H., Zhang, J., Ngo, H.H., Guo, W., Hu, Z., Liang, S., Fan, J., Liu, H., 2015a. A review
Maucieri, C., Borin, M., Barbera, A.C., 2014. Role of C3 plant species on carbon di- on the sustainability of constructed wetlands for wastewater treatment: design
oxide and methane emissions in Mediterranean constructed wetland. Ital. J. and operation. Bioresour. Technol. 175, 594e601.
Agron. 9, 120e126. Wu, H., Fan, J., Zhang, J., Ngo, H., Guo, W., Liang, S., Hu, Z., Liu, H., 2015b. Strategies
Meng, P., Pei, H., Hu, W., Shao, Y., Li, Z., 2014. How to increase microbial degradation and techniques to enhance constructed wetland performance for sustainable
in constructed wetlands: influencing factors and improvement measures. Bio- wastewater treatment. Environ. Sci. Pollut. Res. 22, 14637e14650.
resour. Technol. 157, 316e326. Wu, H., Fan, J., Zhang, J., Ngo, H., Guo, W., Liang, S., Lv, J., Hu, Z., Lu, S., Wu, W., Wu, S.,
Mitsch, W.J., Bernal, B., Nahlik, A.M., Mander, Ü., Zhang, L., Anderson, L., 2016a. Intensified organics and nitrogen removal in the intermittent aerated
Jørgensen, S.E., Brix, H., 2013. Wetlands, carbon, and climate change. Landsc. constructed wetland using a novel sludge-ceramsite as substrate. Bioresour.
Ecol. 28, 583e597. Technol. 210, 101e107.
Morvannou, A., Forquet, N., Michel, S., Troesch, S., Molle, P., 2015. Treatment per- Wu, H., Lin, L., Zhang, J., Guo, W., Liang, S., Liu, H., 2016b. Purification ability and
formances of French constructed wetlands: results from a database collected carbon dioxide flux from surface flow constructed wetlands treating sewage
over the last 30 years. Water Sci. Technol. 71, 1333e1339. treatment plant effluent. Bioresour. Technol. 219, 768e772.
Murray, K.R., Barlow, N., Strack, M., 2017. Methane emissions dynamics from a Wu, H., Zhang, J., Ngo, H.H., Guo, W., Liang, S., 2017. Evaluating the sustainability of
constructed fen and reference sites in the Athabasca Oil Sands Region, Alberta. surface flow constructed wetlands: methane and nitrous oxide emissions.
Sci. Total Environ. 583, 369e381. J. Clean. Prod. 147, 152e156.
Pangala, S.R., Reay, D.S., Heal, K.V., 2010. Mitigation of methane emissions from Wu, S., Kuschk, P., Brix, H., Vymazal, J., Dong, R., 2014. Development of constructed
constructed farm wetlands. Chemosphere 78, 493e499. wetlands in performance intensifications for wastewater treatment: a nitrogen
Rai, U.N., Tripathi, R.D., Singh, N.K., Upadhyay, A.K., Dwivedi, S., Shukla, M.K., and organic matter targeted review. Water Res. 57C, 40e55.
Mallick, S., Singh, S.N., Nautiyal, C.S., 2013. Constructed wetland as an eco- Zhao, Z., Chang, J., Han, W., Wang, M., Ma, D., Du, Y., Qu, Z., Chang, S., Ge, Y., 2016.
technological tool for pollution treatment for conservation of Ganga river. Effects of plant diversity and sand particle size on methane emission and ni-
Bioresour. Technol. 148, 535e541. trogen removal in microcosms of constructed wetlands. Ecol. Eng. 95, 390e398.

Please cite this article in press as: Wu, H., et al., Large-scale multi-stage constructed wetlands for secondary effluents treatment in northern
China: Carbon dynamics, Environmental Pollution (2017), https://doi.org/10.1016/j.envpol.2017.09.048
View publication stats