Spatial-temporal characteristics of nitrogen degradation in typical Rivers of Taihu Lake Basin, China

Spatial-temporal characteristics of nitrogen degradation in typical Rivers of Taihu Lake Basin, China

Science of the Total Environment 713 (2020) 136456 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www...

3MB Sizes 0 Downloads 3 Views

Science of the Total Environment 713 (2020) 136456

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Spatial-temporal characteristics of nitrogen degradation in typical Rivers of Taihu Lake Basin, China Jiaxun Guo a, Lachun Wang a,⁎, Long Yang a, Jiancai Deng b,⁎, Gengmao Zhao c, Xiya Guo d a

School of Geography and Ocean Science, Nanjing University, Nanjing, Jiangsu Province, China State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing, Jiangsu Province, China College of Resources and Environmental Sciences, Nanjing Agricultural University, Nanjing, Jiangsu Province, China d Jiangsu Provincial Academy of Environmental Science, Nanjing, Jiangsu Province, China b c

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• In-situ degradation coefficient of TN, NH+ 4 -N and NO3-N in rivers are measured by an innovative experimental device. • Concentration of nitrogen and water temperature significantly determine spatial-temporal variability of nitrogen degradation. • Impervious ratio within a small buffering zone of the sampling sites shows strong correlations with nitrogen degradation.

a r t i c l e

i n f o

Article history: Received 24 October 2019 Received in revised form 3 December 2019 Accepted 31 December 2019 Available online 07 January 2020 Editor: Jurgen Mahlknecht Keywords: Dissolved inorganic nitrogen degradation Spatial-temporal variation Urbanization Taihu Lake Basin

a b s t r a c t In this study, we focus on the measurement of different nitrogen (N) forms and investigate the spatial-temporal variability of degradation coefficient in river channels. We aim to provide a new approach of deriving in-situ degradation coefficients of different N forms, and highlight factors that determine the spatial-temporal variability of degradation coefficients. Our results are based on a two-year field survey in 34 channels around the Taihu Lake Basin, eastern China. The derived degradation coefficients of different N forms based our newly-developed exper−1 , 0.022–1.175 imental device are: degradation coefficients of TN, NH+ 4 -N and NO3-N range from 0.006–0.449 d d−1 and -0.096–2.402 d−1, respectively. The degradation coefficients of N show strong dependence on N concentration and water temperature. The seasonal difference of water temperature and N concentration leads to spatial-temporal variability of degradation coefficients. The derived degradation coefficients of N are further verified through one-dimensional water quality model simulations. The degradation coefficient obtained in this study and the influencing factors of its spatial-temporal variability provide invaluable reference for studies in aquatic environment. © 2020 Elsevier B.V. All rights reserved.

1. Introduction ⁎ Corresponding authors. E-mail addresses: [email protected] (L. Wang), [email protected] (J. Deng).

https://doi.org/10.1016/j.scitotenv.2019.136456 0048-9697/© 2020 Elsevier B.V. All rights reserved.

The global water environment has been witnessed with a great pressure to the increase of available nitrogen in the past century (Galloway

2

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

et al., 2008). Excess nitrogen affects ecology, quality and value of aquatic ecosystems (Dodds et al., 2009; Galloway et al., 2004; Mulholland et al., 2008; Seitzinger, 1988). For example, two-thirds of American coastal rivers and bays suffer from moderate to severe degradation of water environment due to nitrogen pollution (U.S. EPA, 2001). N85% lakes and 82% river branches over China suffer from serious nitrogen pollution (Novotny et al., 2010). Eutrophication induced by nitrogen pollution in major rivers and lakes can significantly jeopardize the availability of surface water resources. In addition to placing controls on the input of nitrogen into water bodies, increasing the degradation capacity of surface water to nitrogen can be equally or even more important (Singh et al., 2005; Yang and Liu, 2010). Self-purification is defined as the sum of all physical, chemical, and biological processes reducing the load and the concentration of pollutant in a water body, respectively (Obst, 2003; Vagnetti et al., 2003). It is an important concept for decreasing the pollutants in water ecosystems. Self-purification capacity is closely associated with physical and chemical processes, such as dilution, adsorption, oxidative and reductive reactions and others, and biological process such as microbial metabolism, phytoplankton absorption (Obst, 2003; Vagnetti et al., 2003). The load of nitrogen in surface water bodies can be reduced through a variety of processes, including burial of organic matter into sediments, denitrification, sediment adsorption, and absorption by plants and microorganisms (Seitzinger et al., 2002). The reduced rates of nitrogen in water bodies are subject to several factors. Previous studies show that nitrogen concentration and water temperature are the controlling factors in the biological processes such as nitrification and denitrification (Guan et al., 2013; Iriarte et al., 1997; Wu et al., 2013; Zhao et al., 2015; Zhong et al., 2010). High temperature accelerates the rate of nitrification and denitrification of microorganisms (Zhou et al., 2018). On the other hand, large nitrogen content provides more substrates for microorganisms that can accelerate the microbial process (Yao et al., 2016). The nitrification rate is also associated with other indicators of water environment, e.g., Dissolved oxygen (DO), pH, total dissolved solids (TDS) and oxidation-reduction potential (ORP). Previous studies show that DO is related to nitrification rates (Wang and Yang, 2004; Zhao et al., 2015), while difference in pH and ORP can affect the biochemical reaction conditions and further affect the degradation process (Zhou et al., 2018). Degradation coefficient is routinely used to describe self-purification capacity of pollutant, and reflects reduced rates of pollutants in natural water ecosystems. It is one of the most critical parameter in water quality models, and mainly serves the basis for predicting pollutant concentration, estimating river environmental capacity and formulating pollutant discharge strategy (Li et al., 2014a; Shenk and Linker, 2013). At present, degradation coefficient is mainly calculated by empirical methods (Li and Liao, 2002), indoor simulation experiment method (Huang et al., 2017) and water mass tracking method (Kotnala et al., 2016; Tian et al., 2011). An important adequacy of these approaches is lack of consideration in the influence of in-situ water environment on the values of degradation coefficients. For instance, the empirical estimation method is unable to reflect the internal pattern of the pollutants degradation (Huang et al., 2017). The hydraulic and hydrological characteristics and the behaviors of the river course cannot be well represented in indoor experiments (Huang et al., 2017). This is a particular issue, as previous studies found that river hydrological conditions show significant influences on degradation processes. The influence demonstrate itself mainly through changing the probability of contact between nitrogen and microorganisms or particles, as well as affecting the absorption efficiency of microorganisms and suspended substances (Huang et al., 2017; Panda et al., 2015; Wang et al., 2015). The water mass tracking method is unable to exclude the effects of other contamination inputs, and may lead to uncertainty in estimating the reduced rates of specific pollutants. In this study, we examine the spatial-temporal characteristics of nitrogen degradation coefficient and their influencing factors. The

degradation coefficients are obtained based on an innovative experimental device that can consider in-situ conditions of the river systems. The device is placed in river channels to ensure the maintenance of pristine water ecosystems, such as dissolved oxygen, pH, microbial communities, flow rates, and water temperature. Our study region is the Taihu Lake basin. It is the core of the Yangtze River delta region, and homes to several megacities (e.g., Shanghai, Hangzhou, etc.) in China. Cyanobacteria blooms erupt frequently north of the Taihu Lake (Chen et al., 2003), and lead to serious drinking water crisis in the city of Wuxi in 2007 (Li et al., 2014b; Liu et al., 2012; Zhang et al., 2008). Fast urbanization leads to deterioration of water quality in the rivers that feed into the Taihu Lake. 61.1% of nitrogen and phosphorus in the Lake is conveyed by the surrounding rivers (Wang et al., 2017). Non-point source pollution from urban rivers turns into a more serious problem than agricultural non-point sources on the deterioration of water quality in Taihu Lake (Li et al., 2000; Wei et al., 2011; Zhao et al., 2013a). Our study region is covered by diverse land use types, and contrasting hydrological and hydraulic features in the river systems (Guo et al., 2018; Lian et al., 2018). Therefore, it is necessary to determine the degradation coefficient that represents naturally contrasting ambient environments. Our results pertaining to the degradation coefficient can provide valuable references for future modeling studies. We also expect to shed light on effective water quality management through examining the spatial and temporal characteristics of the degradation coefficients in this region. 2. Materials and methods 2.1. Study area Taihu Lake is the third largest freshwater lake in China, with a lake surface area of 2338 km2 and a mean water depth of 1.9 m. Its drainage basin, the Taihu Lake Basin (TLB), is located in the center of the Yangtze River Delta, which is one of the most economically developed regions across China (Fig. 1). TLB covers 45 municipal counties in Jiangsu and Zhejiang Provinces, and the Shanghai metropolitan region. TLB is located in a humid subtropical climate zone. The annual mean temperature is 16 °C, while the annual precipitation is 1115 mm (Xia et al., 2016). More than 60% of the annual precipitation occurs during the period from June to September, controlled by east Asia Summer Monsoon. N200 rivers connect to the Taihu Lake (Lu et al., 2012), and the total length of channels in TLB is 12,000 km (Qin et al., 2007). A distinct feature of the river networks in TLB is low gradients of river channels due to the flat landscape in this region. In addition, the flow regime is severely regulated by hydraulic structures, e.g., sluices, dikes. We collect water samples from 18 channels far from the lake in July 2015, February 2016, October 2016 and December 2016. The 18 channels are distributed in northwestern and southeastern part of TLB, including the Taige river (TG), the grand canal-Weidun (JA), the grand canal-Pingwang (PW), the grand canal-Jiaxing (JX), Danjinlicao river (DJ), Nanxi river (NX), East Tiaoxi (TX), Taipu river (TP). There are another 16 channels that are directly connected to the lake, with their samples collected in December 2014, October 2015, March 2016 and July 2016 (Fig. 1). The 16 channels are distributed from the north towards south, including the Shedu river (SD), Guandu river (GD), Xinzhuang river (XZ), Chendong river (CD), Dapu river (DP), Zhudu river (ZD), Huangdu river (HD), Wuxi river (WX), Dacha river (DC), Hexixingang river (HX), Xintang (XT), Yangjiapu (YJ), Changdou river (CG), Daqian river (DQ), Huanlou river (HL) and Taipu river (TP0). 2.2. Sample collection and analysis We directly measure nitrogen degradation coefficient in each channel based on a self-developed experimental apparatus (Patent No. ZL201520816420.7, Fig. S1). The experimental apparatus enables us to derive nitrogen degradation coefficient with in-situ river conditions.

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

3

Fig. 1. Map that shows the sampling sites, river network and urbanization area in TLB.

The procedures for deriving nitrogen degradation coefficients are described as follows. First, three culture bottles were filled with river water without any treatments (e.g., without nutrients or microorganisms) collected near both river shores and from the open channel at a depth of 50 cm. The culture bottles sealed with a rubber stopper throughout the duration of the experiment to ensure no exchange with river water. The inside of the culture bottles is connected to the atmosphere through a latex tube to ensure aerobic conditions in the bottles. The culture bottles were suspended from the steel ring with a diameter of 80 cm and height of 2.5 cm and placed in rivers (Fig. S1). To avoid disturbing suspended solids settled at the bottom of the bottles, the experimental apparatus is taken out of the water slowly after reaching the preset incubation time (about 1 day). Samples of culture water from the upper layer of each bottle were collected from a side sampling port. Water samples are then placed in ice and transported to the laboratory, filtered (Whatman GF/C filters) and frozen until anal− ysis. Filtered water and original water was analyzed for NH+ 4 -N, NO3 -N and TN using a flow injection analyzer with detection limits of − 0.001 mg/L (Skalar Analytical, Breda, Netherlands). NH+ 4 -N, NO3 -N and TN in three culture bottles at each sampling point were analyzed separately. Their average values were used to represent the degradation coefficient at each sampling sites. Previous studies show that the degradation process of water pollutants conforms to the first-order reaction kinetics model when the turbulent and the flow rate is small (Wang et al., 2006; Zhang et al., 2005).. This method is generally used in indoor simulation calculation of degradation coefficient. In this area, river channels in this urbanized region are heavily controlled by sluices. The flow rates are around 0.03–0.42 m/s. We assume that the one-dimensional reaction kinetics equation also applicable in this study. The degradation coefficient (k, in d−1) is calculated using the first-order reaction kinetics model (Feng et al., 2016; Tien et al., 2011), and takes the form as: k¼

1 C0  ln t Ct

2.3. Measurements of other variables In order to analyze the dependence of nitrogen degradation coefficient on river properties, a paralleling suite of measurements are made at each sampling site. A multi-parameter water quality probe (YSI, USA) was used to measure water temperature, DO,

Table 1 Physical and chemical parameters measured in different temperature ranges. Temperature TN(mg/L) [7–10) [10–13) [13–16) [16–19) [19–22) [22–25) [25–28) [28–31) [31–34)

where t is the culture time (d), ct is the pollutant concentration at time t (mg/L), while c0 is the initial concentration (mg/L).

1.5 1.0 0.6 1.6 1.2 0.9 0.5

− NH+ 4 -N(mg/L) NO3 -N(mg/L) DO(mg/L) pH

1.3 1.4 0.2 0.8 0.9 0.4 0.8 1.1 0.7

± 1.2 ± 0.8 ± 0.2 ± 1.4 ± 0.9 ± 0.8 ± 0.3

1.9 2.7 4.6 2.4 1.3 1.7 1.9 1.5 1.0

± 0.8 ± 0.6 ± 0.8 ± 0.8 ± 0.6 ± 0.3 ± 0.3

7.5 6.9 8.9 7.3 5.5 3.3 4.1 5.0 5.1

± 2.1 ± 1.3 ± 1.8 ± 2.1 ± 1.3 ± 2.3 ± 1.5

7.5 7.7 7.1 7.5 8.1 8.0 7.8 7.8 7.5

± 0.3 ± 0.3 ± 0.3 ± 0.6 ± 0.2 ± 0.4 ± 0.4

Temperature Velocity (m/s)

Water flow (m3/s)

SS (mg/L)

TDS (g/L)

ORP (mV)

[7–10)

94.8 ± 116.3

16.2 ± 14.3 32.8 ± 42.5 7.1 50.2 ± 15.0 63.0 ± 30.4 65.5 27.9 ± 18.8 27.6 ± 16.5 24.1 ± 9.4

0.24 ± 0.04 0.26 ± 0.08 0.05 0.16 ± 0.04 0.19 ± 0.05 0.20 /

188.9 ± 25.1 190.4 ± 37.7 324.0 265.5 ± 35.6 135.9 ± 89.6 108.0 /

0.12 ± 0.01 0.12 ± 0.02

203.5 ± 21.9 215.5 ± 25.9

[10–13) [13–16) [16–19) [19–22) [22–25) [25–28) [28–31)

ð1Þ

4.6 ± 5.7 ± 5.5b 4.4 ± 4.0 ± 3.84 4.3 ± 4.6 ± 3.6 ±

a

[31–34) a b

0.20 ± 0.12 0.16 ± 0.15 0.20 0.36 ± 0.23 0.19 ± 0.12 0.17 0.29 ± 0.25 0.10 ± 0.08 0.14 ± 0.18

19.9 ± 18.4 5.4 59.6 ± 69.2 38.8 ± 56.8 58.0 117.7 ± 145.9 21.3 ± 22.3 68.7 ± 186.3

Data presented as the mean ± standard deviation. Only one value in the water temperature range.

4

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

pH, TDS, ORP. The River Surveyor M9 (SonTek, Xylem Inc.) was used to measure river velocity and discharge across each crosssection of sampled channels. These parameters are used to analyze the influencing factors of river nitrogen degradation coefficients. Due to equipment failure and other factors, some data of physical-chemical parameters were not obtained, including TDS, ORP in the rivers disconnected to Taihu Lake (the 18 sample sites) in summer; velocity, water flow, TDS and ORP in the rivers connected with Taihu Lake in autumn (the 16 sample sites), DO, pH velocity, water flow, ORP and water temperature in the rivers around the lake in winter (the 16 sampling sites).

2.4. Downstream concentration prediction We set up a one-dimensional steady-state water quality model for each sample site. We sample eleven reaches for the points in both upstream and downstream (Table S1). We predict nitrogen concentration of the downstream river based on the derived degradation coefficients (Fisher, 1975; Jiang and Wang, 1997). We assume that the concentration of pollutant is evenly distributed horizontally and vertically. Changes in pollutant concentration are mainly caused by degradation of different pollutants in the longitudinal dispersion and advection. This, however, is not often the case, and may leads uncertainty in the

− Fig. 2. TN NH+ 4 -N and NO3 -N concentrations and degradation coefficients in different groups.

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

estimation (see discussion in Section 4.3). The mathematical equation of the one-dimensional steady-state water quality model (without considering source and sink terms) takes the form as:   ∂c ∂ ∂c ∂c ∂x þu ¼D −kc ∂t ∂x ∂x u2 B2 D ¼ 0:011 pffiffiffiffiffiffiffiffi H gHI

ð2Þ

ð3Þ

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi ux 4kD 1− 1 þ 2D 86400u2 c ¼ c0 e

dimensionless; x is the distance from the upstream section, m; k is degradation coefficient (d−1). We adopt the calculated degradation coefficient as the parameter k of the one-dimensional steady-state water quality model. The concen− tration of TN, NH+ 4 -N and NO3 -N in six downstream monitoring sections were predicted using the formula (2)–(4). Table S1 summarizes the parameters used in the calculation. The accuracy of the concentration was evaluated based on relative error (RE, formula (5)). The calculation formula is as follows: RE ¼

ð4Þ

where c and c0 are pollutant concentration in the downstream and upstream, respectively, mg/L; t is time with the unit of day (d); u is the average flow velocity (m/s); D is the longitudinal dispersion coefficient (m2/s); B is the river width, m; H is the average water depth, m; g is the gravitational acceleration, m/s2; I is the slope of the river channel,

5

  C predict −C observe  C observe

 100

ð5Þ

where Cpredict is predicted value and Cobserve is observed value. The smaller the RE, the predicted value is closer to the observed value, which also verifies the applicability of the degradation coefficient in this study. Spatial and temporal differences in mean degradation coefficient among different sites and seasons were tested using one-way analysis of variance (ANOVA) followed by Fisher's Least Significant Difference

− Fig. 3. Spatial distribution of the average concentration (a) (b) (c) and degradation coefficient (DeC) (d) (e) (f) of TN NH+ 4 -N and NO3 -N.

6

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

(LSD) test. The hierarchical clustering was adopted by Ward method and square Euclidean distance method based on the nitrogen concentration of four seasons in all sites (El-Hames et al., 2013). Principal component analysis (PCA) was used to analyze the impacts of environmental conditions on degradation coefficient of nitrogen. All statistical analyses were performed in SPSS version 19.0 (SPSS Inc., Chicago, Illinois).

3. Results 3.1. Spatial-temporal characteristics of physical and chemical parameters We provide a summary of the physical (including SS, TDS, Velocity − and Discharge) and chemical (inducing TN, NH+ 4 -N, NO3 -N, DO, pH and ORP) properties of each of the sampled channels in Table S2. As can be seen from Table S2, the sampled channels represent diverse en− vironmental background. The concentration of TN, NH+ 4 -N and NO3 -N is 1.2–9.1 mg/L, 0.1–6.3 mg/L and 0.2–4.6 mg/L, with the coefficients of variation of 33.7%, 104.3% and 50.7%, respectively.

We sort all of the measured samples by temperature, so as to represent the influence of seasonality. The average value of each temperature − range was calculated. The values of TN, NH+ 4 -N and NO3 -N are the highest in the temperature range of 10–16 °C and are the lowest in the range of 31–34 °C. However, DO and TDS are higher in low temperature ranges, while pH and SS are the highest in the medium temperature ranges. The values of ORP are the highest in the temperature range of 13–19 °C (Table 1). In order to analyze the spatial characteristics of nitrogen concentration, we categorized the 34 sample sites into five groups based on the cluster analysis method focusing on the seasonal concentration − of TN, NH+ 4 -N and NO3 -N (Fig. 2 and Fig. S2). The spatial distribution of the five groups is shown in Fig. 2. As can be seen from Fig. 2, there is a concentration order of TN and NH+ 4 -N as Group 4 N Group 5 N Group 1, 2 N Group 3, while the order of NO− 3 -N is as Group 1 N Group 5 N Group 4 N Group 2 N Group 3. The spatial variability − of the average concentration of TN, NH+ 4 -N and NO3 -N can be summarized as: the concentration is first decreasing from northwest to the southwest, and then increasing from southwest to the south of the lake (Fig. 3).

− Fig. 4. Average degradation coefficients of each temperature segment of TN (a), NH+ 4 -N (b) and NO3 N (c) vary with temperature. And average degradation coefficients of each group vary − with concentrations of TN (d), NH+ 4 -N (e) and NO3 N (f).

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

7

3.2. Spatial-temporal characteristics of degradation coefficients

3.3. Physicochemical factors of degradation coefficient

− We calculate the degradation coefficients of TN, NH+ 4 -N, NO3 -N based on in-situ experiments in 34 sampled channels around the lake. The degradation coefficient of TN is 0.006–0.449 d−1 (with the median and mean values being 0.156 d−1 and 0.177 d−1, respectively); the deg− −1 radation coefficients for NH+ (0.232 4 -N and NO3 -N are 0.022–1.175 d −1 −1 −1 d of median, 0.302 d of mean) and − 0.096–2.402 d (0.130 d−1 of median, 0.196 d−1 of mean), respectively. The degradation coefficients show strong linear correlation with water temperature. The degradation coefficients of TN show positive correlation (R2 = 0.87) with temperature at the significance level of − 0.01 (Fig. 4a). The degradation coefficients of NH+ 4 -N and NO3 -N show cubic polynomial correlations with water temperature. The corre− lation coefficients are 0.78 and 0.58 for NH+ 4 -N and NO3 -N, respectively + (Fig. 4b and c). The degradation coefficient of NH4 -N attains its first peak when the temperature is 16–19 °C, and further increases with temperature after crossing a local minimum when the temperature is 22–25 °C. Changes of NO− 3 -N with temperature are similar with that + of NH+ 4 -N, but the temperature is slightly higher than NH4 -N when the peak values are attained. The maximum degradation coefficients − for TN, NH+ 4 -N, and NO3 -N are obtained when the water temperature is 31–34 °C. Based on the clustering analysis of nitrogen concentration, the degradation coefficients of TN for each group is ranked as Group 3 N Group 2 N Group 5 N Group 1 N Group 4, while for NH+ 4 -N, the order is Group 3 N Group 1 N Group 2 N Group 5 N Group 4. The rank of average degradation coefficient of NO− 3 -N is Group 3 N Group 2 N Group 4 N Group 1 (Fig. 2). There is significantly negative correlation between degradation coefficients and concentrations of TN (R2 = 0.90), 2 2 − NH+ 4 -N (R = 0.98) and NO3 -N (R = 0.89) (Fig. 4). The spatial pattern − of degradation coefficients of TN, NH+ 4 -N and NO3 -N are opposite to that of concentration, with a decreasing trend from southwest to south, and an increasing trend from north to southwest of the lake (Fig. 3).

We show correlations between key physiochemical factors of degradation coefficients in Fig. 5. As can be seen from Fig. 5, there is a significantly negative correlation between nitrogen concentration and degradation coefficient. Water temperature is significantly correlated with degradation coefficient of TN (P b 0.01), NH+ 4 -N (P b 0.01) and NO− 3 -N (P b 0.01). TDS shows a negative correlation with the degrada− tion coefficient of TN (P b 0.01), NH+ 4 -N (P b 0.01) and NO3 -N (P b 0.01) (Fig. 5). pH is significantly correlated with the degradation coefficient of TN (P b 0.01) and NO− 3 -N (P b 0.01) (Fig. 5a and c). Velocity, discharge and ORP had no significant correlation with degradation coefficient (P N 0.05). Based on the Principal Component Analysis, the physiochemical factors of degradation coefficients can be divided into three main components. The three components consist of different factors, and are termed as concentration factors (PC1), chemical factors (PC2), and hydrodynamic factors (PC3) (Table 2). The first two components contribute to 54% of the total variance. The hydrodynamic factors alone only contribute to 14% of the total variance. The degradation coefficients show dependence on impervious ratios over the buffering zone of the sample sites. Large impervious ratios in the buffering zone lead to higher degradation coefficients (Fig. 6). However, the correlation only exists when the radius of the buffering zone is smaller than 2.5 km. The correlation becomes statistically insignificant when the radius of the buffering zone exceeds 2.5 km (Fig. 6). 3.4. Downstream concentration prediction We further apply measured degradation coefficients in the onedimensional water quality model to predict concentration of TN downstream of each sampling site. The predicted concentration of TN shows consistency with measurements, with an average relative error (RE) being 13% (Fig. 7 and Table 3). The minimum RE is in spring, while the maximum error is in winter. The average RE of NH+ 4 -N is 34.1%. The

− Fig. 5. Correlations of factors with degradation coefficient of TN (a), NH+ 4 -N (b) and NO3 -N (c). The numbers in the figure represent Pearson's correlation coefficients. The red and blue scatters represent negative and positive correlation, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

8

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

Table 2 PCA loadings, percentages of variance explained for environmental parameters measured at all sampling sites. Parameters

TN TDS NO3-N Water temperature NH4-N DO ORP pH Velocity Water flow % of variance

Components PC1

PC2

PC3

0.900 0.785 0.720 −0.714 0.687 0.331 −0.058 0.198 −0.183 −0.414 32.836

−0.206 0.038 0.343 −0.334 −0.335 0.751 0.732 −0.662 0.366 0.260 21.250

−0.048 0.156 0.103 −0.297 −0.092 −0.037 −0.424 0.411 0.723 0.637 14.128

predicted and observed values of NH+ 4 -N are more consistent in spring and winter, with RE being 17.2% and 17.3%, respectively, while it is less consistent during summer, with RE being 66.3%. The RE between the predicted and observed concentration of NO− 3 -N is comparatively larger than TN and NH+ 4 -N, with lowest (highest) RE in spring (winter). 4. Discussion 4.1. Analysis of spatial variability of degradation coefficient Nitrogen degradation in aquatic environment mainly involves three types of processes, i.e., physical processes, biological processes and chemical processes (Obst, 2003; Vagnetti et al., 2003). The negative correlations between concentration and degradation coefficients of TN, NH+ 4 -N, NO3-N determine its spatial variation, with the highest

degradation coefficients lie in the region with the lowest concentration of nitrogen (Fig. 8). PCA-based analysis further confirm that concentration plays a major role in determining the spatial variability (Table 2). Our results are consistent with previous studies regarding to the spatial variation in this region (Wu et al., 2018). Suspended solids in rivers can adsorb nitrogen to their surfaces, and then settle away from the water (Guo and Yu, 2006). In addition, nitrifying and denitrifying bacteria can take different forms of nitrogen as substrates for biological transformation (Liu et al., 2013; Xia et al., 2004). Those bacteria are more active in aquatic environment with high concentration of nitrogen (Wu et al., 2013; Zhao et al., 2015). For this study, the concentration of TN, NH+ 4 -N and NO3-N is high in this region. High concentration of nitrogen prohibits adsorption deposition and biological reactions. Previous studies on the adsorption of ammonia nitrogen showed that when concentration increases from 0 mg/L to 2 mg/L, the maximum quantity of NH+ 4 N absorbed on suspended solid ranged from 0 mg N /g to 0.4 mg N /g (Xia et al., 2004). The average SS concentration is 26.4 mg/L in our study region. We thus estimate the value of NH+ 4 -N absorbed ranged from 0 mg/L to 0.01 mg/L which is much less than the concentration of NH+ 4 -N (0.2–2.4 mg/L). High nitrogen, on the other hand, is prone to lead to eutrophication, which provides negative feedbacks through reducing the degradation capacity of rivers (Zhu et al., 2013). The concentration of nitrogen in receiving waters is also closely tied to land use patterns, urbanization rate, population density in surrounding environment (Janardan and Chang, 2018; Lian et al., 2018; Mouri et al., 2011; Tu, 2013). Lian et al. (2018) analyzed the influence of human activities on long-term change of nitrogen import and export in the TLB. The Group 1, Group 4 and Group 5 in our study are located in high urbanization areas. A large number of arable land caused high nitrogen export (3000–4000 kg/km2·yr). The rivers in Group 2 are covered by relatively low urban land uses, with the nitrogen export of 2000–3000 kg/km2·yr. The urbanization rate in Group 3 is low, with

Fig. 6. Correlation between impervious rate and degradation coefficient of TN and NH+ 4 -N.

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

9

− Fig. 7. Predicted and observed concentrations of TN, NH+ 4 -N and NO3 -N in spring, summer, autumn and winter, respectively.

the upper reaches of the river covered by forest with the nitrogen export of 2000–4000 kg/km2·yr (Lian et al., 2018). Some rivers in Group 3 are outflow from the lake, where the water quality is relatively good (Zhang et al., 2014). Our results echo with theirs, and confirm that impervious coverage is an important player in characterizing degradation coefficient (Fig. 6). Low nitrogen degradation coefficient generally occurs in rivers located in high urbanization areas with high nitrogen concentration, and vice versa. Spatial heterogeneity in urban land use lead to spatial contrast in nitrogen concentration and degradation coefficients in this region. 4.2. Analysis of temporal variability of degradation coefficient Previous studies show that the concentration of nitrogen is higher in low water temperature (spring and winter), while the degradation coefficient is higher in medium and higher water temperature (summer and autumn) (Wang et al., 2012; Zhang et al., 2015). This is also consistent with the negative correlation between concentration and degradation coefficient of TN, NH+ 4 -N and NO3-N. The nitrogen concentration in TLB is lower in summer and autumn than that in spring and winter (Xu et al., 2009). In the TLB, 70% rainfall occur during April to September. The generated runoff carries pollution into rivers, meanwhile a larger amount of runoff also dilute the pollutants (Lang et al., 2013). In our study, TDS and degradation coefficient of nitrogen showed significantly negative correlation (Fig. 5). TDS is smaller in summer/autumn than spring/winter when the temperature is relatively lower, highlighting the dilution effects of rainfall-runoff on pollutants (Table 2).

Water temperature is another important factor affecting the spatial variation of degradation coefficient. High water temperature promotes biochemical reactions and the degradation coefficient of nitrogen (Zhao et al., 2013b; Zhao et al., 2015). Water temperature in summer and autumn (30.3 and 20.2 °C) is higher than spring and winter (14.3 and 10.2 °C). There is a positive correlation between temperature and degradation coefficient (Fig. 4 and 5), which can explain the temporal variation of degradation coefficient. Previous studies show that fast disturbance of water can accelerate the degradation of pollutants (Huang et al., 2017). In this study, there is no significant correlation between velocity and nitrogen degradation coefficient (Fig. 5), which is possibly tied to the unique river regimes in this region. Flow velocities among the four seasons show small variations due to flat channel beds in this region. (Table 2). Previous studies also show that DO is related to nitrification rates (Wang and Yang, 2004; Zhao et al., 2015), while differences in pH and ORP can affect biochemical reaction conditions and further affect the degradation process (Zhou et al., 2018). However, there are no significant correlations between DO, ORP and degradation coefficient. DO and ORP in natural rivers are not the limiting factors for the degradation process (DON5.0 mg/L). 4.3. Reliability analysis of degradation coefficient The TN degradation coefficient derived in this study (i.e., 0.006–0.449 d−1), which is consistent with the results of TN comprehensive degradation coefficient of 0.006–0.450 d−1 in the

Table 3 − Relative error (in %) in predicted sties of TN, NH+ 4 -N and NO3 -N. Predicted sites

DJ12 DJ32 NX3 CD TG13 TG23 TP1 TP2 JX21 JX31 CG

NO− 3 -N

NH+ 4 -N

TN Spring

Summer

Autumn

Winter

Spring

Summer

Autumn

Winter

Spring

Summer

Autumn

Winter

0.4 6.8 7.6 15.5 1.6 1.4 36.0 10.8 3.0 8.2 11.4

8.3 26.3 2.2 46.9 18.5 24.2 3.3 17.6 3.1 3.2 8.3

1.2 1.5 7.9 24.7 6.9 1.4 37.7 23.7 2.8 7.9 9.8

6.1 8.9 1.7 11.7 2.4 6.1 69.5 3.8 2.2 8.5 59.7

1.6 0.5 24.8 6.9 8.7 3.3 48.2 24.5 3.7 20.7 46.7

140.8 1.6 38.3 62.0 40.9 59.7 226.5 13.9 27.7 53.7 64.5

45.3 35.7 22.8 50.1 30.4 36.3 40.9 46.7 12.6 46.2 22.3

11.9 3.0 8.9 16.3 2.8 1.1 90.8 9.5 23.8 13.9 8.2

9.0 14.5 19.9 7.1 3.6 2.9 24.2 13.3 10.0 8.1 0.5

20.7 32.6 0.6 14.7 5.8 20.7 60.5 20.4 126.5 104.1 109.4

21.1 45.3 11.1 33.9 83.9 102.9 89.4 47.6 9.6 39.9 80.3

23.3 6.6 3.3 43.9 10.8 5.3 80.7 15.2 1.2 15.9 881.2

10

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

Fig. 8. Sampling sites. The orange and blue arrows represent the decreasing trends of degradation coefficients and concentrations, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

upper Mississippi River (Table S3) (Alexander et al., 2000). The −1 degradation coefficient of NH+ , and the max4 -N is 0.022–1.175 d imum value is much higher than the comprehensive degradation coefficient of ammonia nitrogen in typical rivers across China (0.058–0.27 d−1) (Table S3), while the mean value is relatively consistent (0.302 d−1 in our study and 0.240 d −1 in others). 59.1% water samples have degradation coefficient of NH+ 4 -N in the range of Chinese rivers. Previous studies focus on degradation coefficient of NH+ 4 -N in limited channels with small variations in environmental conditions (Table S3). There is not degradation coefficient of NO− 3 -N in previous researches. Our present study comprehensively analyzed degradation coefficients of TN, NH+ 4 -N and NO− 3 -N in different environmental conditions in TLB. The negative value of nitrate degradation coefficient may be attributed to the fact that the amount of nitrate produced by nitrification is greater than that of denitrification and deposition degradation. One-dimensional hydrodynamic simulation of downstream nitrogen concentration further validates the reliability of degradation coefficient obtained through the in-situ experiments. In general, the calculated value is relatively close to the observed value with lower relative error in spring, but higher RE in other seasons. There are several possible reasons for the large relative errors: (1) average flow rates, river widths and depths in upstream and downstream are adopted, which may be unable to represent the hydrological characteristics of specific rivers, especially in summer and autumn with the flow relatively higher. The role of water disturbance and river width is under-represented (Cai et al., 2010; Guo and Yu, 2006; Huang et al., 2017); (2) Failure to exclude the inflow of pollutants from tributaries and the pollution input from sewage outlets along the river, which may lead to biases in model simulation; (3) There are water gates and other water conservancy facilities along some of the long channels (e.g., NX3-CD, TX1-DQ and TP-TP1). These properties are not well represented in the one-dimensional hydrodynamic model. However, the derived degradation coefficients are within a reliable range of values, although caution is needed when translating the results to other regions.

5. Summary and conclusions In this study, the degradation coefficients of TN, NH+ 4 -N and NO3N under different surrounding environments were monitored and calculated based on an innovative in-situ experimental device. We investigate the spatial and temporal characteristics of degradation coefficients. The values of degradation coefficients of TN, NH+ 4 -N and NO 3-N range from 0.006–0.449 d −1 , 0.022–1.175 d −1 and -0.096–2.402 d−1, respectively. The spatial variations of degradation coefficient show a decreasing trend from southwest to northwest and south of Taihu lake. Larger values of degradation coefficients are observed in summer and autumn. The spatial-temporal characteristics of degradation coefficients are directly influenced by nitrogen concentration and water temperature. Impervious coverages within the buffering zone of the receiving waters lead to high concentration of nitrogen and degradation coefficient. We further test the reliability of derived values of degradation coefficients based on one-dimensional hydrodynamic model simulations, with reasonable predicted values of nitrogen obtained. We note that caution is needed when transferring the values of degradation coefficients to other regions. Key elements that should consider include the surrounding environment of the river, nitrogen concentration, water temperature and other hydrological conditions (e.g., velocity, discharge, and sluices et al.). Acknowledgments The authors would like to acknowledge Haitao Zhang, Yun Chen for their assistance in the field experiments. Funding This research is financially supported by the Major Science and Technology Program for Water Pollution Control and Treatment (2017ZX07203002-02-01), National Natural Science Foundation of China (41271213) and the Nanjing University Innovation and Creative Program for PhD Candidates (CXCY18-23).

J. Guo et al. / Science of the Total Environment 713 (2020) 136456

Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2019.136456.

References Alexander, R.B., Smith, R.A., Schwarz, G.E., 2000. Effect of stream channel size on the delivery of nitrogen to the Gulf of Mexico. Nature 403, 758. Cai, J., Pan, W., Cao, Y., Wu, Y., Zhang, L., 2010. Impact of river morphology on urban river self-purification capacity in Guangzhou. Water Resources Protection 26, 16–19. Chen, Y., Fan, C., Teubner, K., Dokulil, M., 2003. Changes of nutrients and phytoplankton chlorophyll- a in a large shallow lake, Taihu, China: an 8-year investigation. Hydrobiologia 506-509, 273–279. Dodds, W.K., Bouska, W.W., Eitzmann, J.L., Pilger, T.J., Pitts, K.L., Riley, A.J., et al., 2009. Eutrophication of U.S. freshwaters: analysis of potential economic damages. Environmental Science & Technology 43, 12–19. El-Hames, A.S., Hannachi, A., Al-Ahmadi, M., Al-Amri, N., 2013. Groundwater quality zonation assessment using GIS, EOFs and hierarchical clustering. Water Resour. Manag. 27, 2465–2481. Feng, S., Li, X., Deng, J., 2016. Biodegradation coefficients of typical pollutants in the plain rivers network. Environmental Science 37, 1724–1733. Fisher, B.H., 1975. Simple method for predicting dispersion in streams. J. Environ. Eng. 101, 453–455 American Society of Civil Engineering. Galloway, J.N., Dentener, F.J., Capone, D.G., Boyer, E.W., Howarth, R.W., Seitzinger, S.P., et al., 2004. Nitrogen cycles: past, present, and future. Biogeochemistry 70, 153–226. Galloway, J.N., Townsend, A.R., Jan Willem, E., Mateete, B., Zucong, C., Freney, J.R., et al., 2008. Transformation of the nitrogen cycle: recent trends, questions, and potential solutions. Science 320, 889–892. Guan, D., Fan, J., Chen, J., Feng, S., Zhao, H., 2013. Seasonal variation of denitrification rate in estuary sediments of Liaohe, China. Third International Conference on Intelligent System Design & Engineering Applications. Guo, C., Yu, G., 2006. Adsorption properties of sediment to pollutants of contaminated river water. Ecology & Environment 15, 1151–1155. Guo, J., Wang, L., Guo, X., Zhao, G., Deng, J., Zeng, C., 2018. Spatio-temporal differences in nitrogen reduction rates under biotic and abiotic processes in river water of the Taihu Basin, China. Int. J. Environ. Res. Public Health 15, 2568. Huang, B., Hong, C., Du, H., Qiu, J., Liu, D., 2017. Quantitative study of degradation coefficient of pollutant against the flow velocity. Journal of Hydrodynamics Ser B 29, 118–123. Iriarte, A., Sota, A.D.L., Orive, E., 1997. Seasonal variation of nitrification along a salinity gradient in an urban estuary. Hydrobiologia 362, 115–126. Janardan, M., Chang, H., 2018. Landscape and anthropogenic factors affecting spatial patterns of water quality trends in a large river basin, South Korea. J. Hydrol. 564, 26–40. Jiang, Z., Wang, J., 1997. Study and modifacation on one dimension water quality model in naturalr ivers. Journal of Hunan University 24, 91–95. Kotnala, G., Dobhal, S., Chauhan, J.S., 2016. Monitoring the self-purification capacity of the river Alaknanda stretch at Srinagar, Uttarakhand, India. International Journal of River Basin Management 14, 491–498. Lang, M., Li, P., Yan, X., 2013. Runoff concentration and load of nitrogen and phosphorus from a residential area in an intensive agricultural watershed. Sci. Total Environ. 458–460, 238–245. Li, J., Liao, W., 2002. The effect of water flow on the biodegradation of organic pollutant. Res. Environ. Sci. 15, 45–48. Li, R., Xia, Y., Wu, A., Qian, Y., 2000. Pollutants sources and their discharging amount in Taihu Lake area of Jiangsu Province. Journal of Lakeence 12, 147–153. Li, K., Zhang, L., Li, Y., Zhang, L., Wang, X., 2014a. A three-dimensional water quality model to evaluate the environmental capacity of nitrogen and phosphorus in Jiaozhou Bay, China. Mar. Pollut. Bull. 91, 306–316. Li, Y., Tang, C., Yu, Z., Acharya, K., 2014b. Correlations between algae and water quality: factors driving eutrophication in Lake Taihu, China. International Journal of Environmental Science & Technology 11, 169–182. Lian, H., Lei, Q., Zhang, X., Yen, H., Wang, H., Zhai, L., et al., 2018. Effects of anthropogenic activities on long-term changes of nitrogen budget in a plain river network region: a case study in the Taihu Basin. Sci. Total Environ. 645, 1212–1220. Liu, Q.Q., Gao, Y., Xu, X.F., 2012. An ecological dynamics model of algae growth in Taihu Lake and the analysis of its influencing factors. Adv. Mater. Res. 518–523, 4961–4966. Liu, T., Xia, X., Liu, S., Mou, X., Qiu, Y., 2013. Acceleration of denitrification in turbid rivers due to denitrification occurring on suspended xediment in oxic waters. Environmental Science & Technology 47, 4053–4061. Lu, S.Y., Yuan, Y., Jin, X., Jiao, W., Yao-Jie, W.U., Ren, D.Y., et al., 2012. Speciation distribution of nitrogen in sediments of 7 rivers around Taihu Lake. Environmental Science 33, 1497–1502. Mouri, G., Takizawa, S., Oki, T., 2011. Spatial and temporal variation in nutrient parameters in stream water in a rural-urban catchment, Shikoku, Japan: effects of land cover and human impact. J. Environ. Manag. 92, 1837–1848. Mulholland, P.J., Helton, A.M., Poole, G.C., Hall, R.O., Hamilton, S.K., Peterson, B.J., et al., 2008. Stream denitrification across biomes and its response to anthropogenic nitrate loading. Nature 452, 202–205. Novotny, V., Wang, X., Englande, A.J., Bedoya, D., Promakasikorn, L., Tirado, R., 2010. Comparative assessment of pollution by the use of industrial agricultural fertilizers in four rapidly developing Asian countries. Environment Development & Sustainability 12, 491–509. Obst, U., 2003. Strategies of maintaining the natural purification potential of rivers and lakes. Environmental Science & Pollution Research International 10, 251–255.

11

Panda, U.S., Mahanty, M.M., Rao, V.R., Patra, S., Mishra, P., 2015. Hydrodynamics and water quality in Chilika lagoon-a modelling approach. Procedia Engineering 116, 639–646. Qin, B., Xu, P., Wu, Q., Luo, L., Zhang, Y., 2007. Environmental issues of Lake Taihu, China. Hydrobiologia 581, 3–14. Seitzinger, S.P., 1988. Denitrification in freshwater and coastal marine ecosystems: ecological and geochemical significance. Limnology & Oceanography 33, 702–724. Seitzinger, S.P., Styles, R.V., Boyer, E.W., Alexander, R.B., Billen, G., Howarth, R.W., et al., 2002. Nitrogen retention in rivers: model development and application to watersheds in the northeastern U.S.A. Biogeochemistry 57/58, 199–237. Shenk, G.W., Linker, L.C., 2013. Development and application of the 2010 Chesapeake bay watershed total maximum daily load model. Jawra Journal of the American Water Resources Association 49, 1042–1056. Singh, K.P., Malik, A., Sinha, S., 2005. Water quality assessment and apportionment of pollution sources of Gomti river (India) using multivariate statistical techniques—a case study. Anal. Chim. Acta 538, 355–374. Tian, S., Wang, Z., Shang, H., 2011. Study on the self-purification of Juma River. Procedia Environ. Sci. 11, 1328–1333. Tien, C.J., Chuang, T.L., Chen, C.S., 2011. The role of naturally occurring river biofilms on the degradation kinetics of Diazinon. Acta Hydrochim. Hydrobiol. 39, 931–938. Tu, J., 2013. Spatial variations in the relationships between land use and water quality across an urbanization gradient in the watersheds of northern Georgia, USA. Environ. Manag. 51, 1–17. Environmental Protection Agency, U.S., 2001. National Coastal Condition Report. EPA620/R-01/005, Washington DC. Vagnetti, R., Miana, P., Fabris, M., Pavoni, B., 2003. Self-purification ability of a resurgence stream. Chemosphere 52, 1781–1795. Wang, J., Yang, N., 2004. Partial nitrification under limited dissolved oxygen conditions. Process Biochem. 39, 1223–1229. Wang, Y., Zhou, Z., Wang, L., Zhang, J., Sun, Y., 2006. Test on degradation coefficient of ammonia-nitrogen in Yellow River adjacent to Lanzhou. Journal of Lanzhou University of Technology 32, 72–74. Wang, L., Zhang, H., Xie, J., Luo, J., Li, Y., 2012. Research on dynamic vharacteristics of river pollution degradation coefficient between Longmen and Sanmenxia on the Yellow River. Journal of Xian University of Technology 28, 293–297. Wang, C., Zheng, S., Wang, P., Hou, J., 2015. Interactions between vegetation, water flow and sediment transport: a review. Journal of Hydrodynamics, Ser. B 27, 24–37. Wang, J., Zhao, Q., Pang, Y., Hu, K., 2017. Research on nutrient pollution load in Lake Taihu, China. Environ. Sci. Pollut. Res. 24, 1–10. Wei, S., Wen, C., Cheng, C., Shuang, G., Yao, G., 2011. Constraint regionalization of water environment and the guidance for industrial layout: a case study of Jiangsu Province. J. Geogr. Sci. 21, 937–948. Wu, J., Chen, N., Hong, H., Lu, T., Wang, L., Chen, Z., 2013. Direct measurement of dissolved N2 and denitrification along a subtropical river-estuary gradient, China. Mar. Pollut. Bull. 66, 125–134. Wu, Z., Wang, X., Chen, Y., Cai, Y., Deng, J., 2018. Assessing river water quality using water quality index in Lake Taihu Basin, China. Sci. Total Environ. 612, 914–922. Xia, X., Yang, Z., Huang, G., Zhang, X., Yu, H., Rong, X., 2004. Nitrification in natural waters with high suspended-solid content––A study for the Yellow River. Chemosphere 57, 1017–1029. Xia, Y., Ti, C., She, D., Yan, X., 2016. Linking river nutrient concentrations to land use and rainfall in a paddy agriculture–urban area gradient watershed in Southeast China. Sci. Total Environ. 566–567, 1094–1105. Xu, H., Yang, L., Zhao, G., Jiao, J., Yin, S., Liu, Z., 2009. Anthropogenic impact on surface water quality in Taihu Lake region, China. Pedosphere 19, 765–778. Yang, S., Liu, P., 2010. Strategy of water pollution prevention in Taihu Lake and its effects analysis. J. Great Lakes Res. 36, 150–158. Yao, X., Zhang, L., Zhang, Y., Xu, H., Jiang, X., 2016. Denitrification occurring on suspended sediment in a large, shallow, subtropical Lake (Poyang lake, China). Environ. Pollut. 219, 501. Zhang, X., Yang, Z., Xia, X., 2005. An experimental study on nitrification in natural water of the yellow river, China. Environ. Chem. 24, 245–249. Zhang, L., Xia, M., Zhang, L., Wang, C., Lu, J., 2008. Eutrophication status and control strategy of Taihu Lake. Front. Environ. Sci. Eng. China 2, 280–290. Zhang, Q., Kong, M., Tang, W., Zhang, L., Yin, H., Cai, Y., et al., 2014. Water quality assessment in the surface waters from Taihu Lake and its main tributaries. Resources & Environment in the Yangtze Basin 23, 73–80. Zhang, Y., Shen, J., Shi, S., Han, L., Yao, Z., 2015. Dynamic change of the river pollutions' composite degradation coefficient in Huaihe River's tributary. Environmental Monitoring in China 31, 64–67. Zhao, H., You, B., Duan, X., Becky…, S., 2013a. Industrial and agricultural effects on water environment and its optimization in heavily polluted area in Taihu Lake Basin, China. Chin. Geogr. Sci. 23, 203–215. Zhao, Y., Xia, Y., Kana, T.M., Wu, Y., Li, X., Yan, X., 2013b. Seasonal variation and controlling factors of anaerobic ammonium oxidation in freshwater river sediments in the Taihu Lake region of China. Chemosphere 93, 2124–2131. Zhao, Y., Xia, Y., Ti, C., Shan, J., Li, B., Xia, L., et al., 2015. Nitrogen removal capacity of the river network in a high nitrogen loading region. Environmental Science & Technology 49, 1427–1435. Zhong, J., Fan, C., Liu, G., Zhang, L., Shang, J., Gu, X., 2010. Seasonal variation of potential denitrification rates of surface sediment from Meiliang Bay, Taihu Lake, China. J. Environ. Sci. 22, 961–967. Zhou, S., Xia, C., Huang, T., Zhang, C., Fang, K., 2018. Seasonal variation of potential denitrification rate and enhanced denitrification performance via water-lifting aeration technology in a stratified reservoir—A case study of Zhoucun reservoir. Chemosphere 211, 1123–1136. Zhu, X., Shen, W., Gao, K., Liu, J., 2013. Study on the integrated attenuation coefficient of NH_3-N in the soctors of main stream of Songhua River in Jilin Province. Science Technology & Engineering 13, 2758–2773.