Pedosphere 19(6): 765–778, 2009 ISSN 1002-0160/CN 32-1315/P c 2009 Soil Science Society of China Published by Elsevier Limited and Science Press
Anthropogenic Impact on Surface Water Quality in Taihu Lake Region, China∗1 XU Hai1,2 , YANG Lin-Zhang2 , ZHAO Geng-Mao1 , JIAO Jia-Guo1 , YIN Shi-Xue3 and LIU Zhao-Pu1,∗2 1 College
of Resources and Environmental Science, Nanjing Agricultural University, Nanjing 210095 (China) of Soil Science, Chinese Academy of Sciences, Nanjing 210008 (China) 3 Agricultural College, Yangzhou University, Yangzhou 225007 (China) 2 Institute
(Received December 3, 2008; revised September 22, 2009)
ABSTRACT Taihu Lake region is one of the most industrialized areas in China, and the surface water is progressively susceptible to anthropogenic pollution. The physicochemical parameters of surface water quality were determined at 20 sampling sites in Taihu Lake region, China in spring, summer, autumn, and winter seasons of 2005–2006 to assess the eﬀect of human activities on the surface water quality. Principal component analysis (PCA) and cluster analysis (CA) were used to identify characteristics of the water quality in the studied water bodies. PCA extracted the ﬁrst three principal components (PCs), explaining 80.84% of the total variance of the raw data. Especially, PC1 (38.91%) was associated with NH4 -N, total N, soluble reactive phosphorus, and total P. PC2 (22.70%) was characterized by NO3 -N and temperature. PC3 (19.23%) was mainly associated with pH and dissolved organic carbon. CA showed that streams were inﬂuenced by urban residential subsistence and livestock farming contributed signiﬁcantly to PC1 throughout the year. The streams inﬂuenced by farmland runoﬀ contributed most to PC2 in spring and winter compared with other streams. PC3 was aﬀected mainly by aquiculture in spring, rural residential subsistence in summer, and livestock farming in fall and winter seasons. Further analyses showed that farmlands contributed signiﬁcantly to nitrogen pollution of Taihu Lake, while urban residential subsistence and livestock farming also polluted water quality of Taihu Lake in rainy season. The results would be helpful for the authorities to take sound actions for an eﬀective management of water quality in Taihu Lake region. Key Words:
anthropogenic pollution, cluster analysis, non-point pollution, principal component analysis
Citation: Xu, H., Yang, L. Z., Zhao, G. M., Jiao, J. G., Yin, S. X. and Liu, Z. P. 2009. Anthropogenic impact on surface water quality in Taihu Lake region, China. Pedosphere. 19(6): 765–778.
INTRODUCTION Taihu Lake region is one of the most industrialized areas in China with high population density and fast economic development. The area covers 0.4% of the total area of China and provides more than 14% of China’s gross domestic production (Shen et al., 2000). During the last decade, because of rapid development of industries and agriculture, pollutants have been increasingly produced and discharged into rivers and the lake, resulting in severe degradation of water quality and restricting the sustainable development of the local economies (Li et al., 2000). As a result of nutrient overload, Taihu Lake becomes currently eutrophic, and algal bloom in the lake has extended coverage and persists throughout the summer season, which seriously aﬀects the lake as a service of drinking water supply (Qin et al., 2007). Despite the eﬀorts dedicated to solve the problem of eutrophication, water quality of the lake showed little improvement in recent years (Chen, 2001; Zhang et al., 2001). It is evident that anthropogenic activities are generally responsible for deterioration of surface water quality of aquatic systems in watersheds. Taihu Lake region receives several types of inputs, such as urban, industrial, and agricultural wastes (Li et al., 2000). Currently, industrial pollution has been ∗1 Project supported by the Knowledge Innovation Key Project of the Chinese Academy of Sciences (No. KZCX1-YW-14-5) and the National Natural Science Foundation of China (No. 30600086). ∗2 Corresponding author. E-mail: [email protected]
H. XU et al.
controlled to a certain extent, and thus rural and agricultural pollution sources are becoming the major concern (Zuo et al., 2003). Considerable eﬀorts have been made to study the loss and load of pollutants from diﬀerent pollution sources (Quan and Yan, 2002; Zhang et al., 2003; Gao et al., 2004; Guo et al., 2004; Qin et al., 2004; Zhang et al., 2007). However, there is limited information on the eﬀect of diﬀerent anthropogenic activities on the water quality of Taihu Lake. Some studies concentrated on the pollution level of Taihu Lake (Huang and Zhu, 1996; Cai et al., 1997; Ye et al., 2007), whereas others focused on the status of pollution in principal rivers around the lake (Wang et al., 2007; Xie et al., 2007). These studies provided good insights into the character of pollution in the lake or in other large water systems, which usually behave as a sink for pollutants. However, the intensiﬁcation of agriculture and the rapid urbanization of the basin are changing the natural ﬂows in the drainage network, creating cross-connections between urban and agricultural pathways, and thus making it diﬃcult to assess various anthropogenic impacts on the water quality of surface water. In this study, physicochemical parameters of surface water quality directly aﬀected by diﬀerent pollution sources were monitored over one-year period. Multivariate statistical techniques, such as principal component analysis (PCA) and cluster analysis (CA), were used to evaluate the eﬀects of human activities on the characteristics of surface water quality. These techniques identify the possible factors or sources that inﬂuence water systems and thus oﬀer a valuable tool to develop appropriate strategies for eﬀective management of the water resources (Adams et al., 2001; Lee et al., 2001; Wunderlin et al., 2001; Reghunath et al., 2002; Simeonov et al., 2004). MATERIALS AND METHODS Study area Taihu Lake is located between 119◦ 54 and 120◦ 36 N and between 30◦ 56 and 31◦ 33 E with a surface area of 2 338 km2 and a catchment area of 36 500 km2 . The lake has an average water depth of 1.9 m. Located on the northern border of the subtropical broad-leaved forest region, it belongs to the southeast monsoon climate area. The rainy season appears in summer. The annual average air temperature of this region is 14.9–16.2 ◦ C. The annual mean precipitation is 1 000–1 400 mm and the annual mean runoﬀ into the lake is 4 100 Mm3 . Over 200 rivers enter into the lake at present. With the Zhihu Port in Wuxi City (located at the north of Taihu Lake) and Wulou Port (located at the south of Taihu Lake) as borderline, the Taihu Lake basin is divided into two parts: upstream and downstream regions. Water tributaries in the upstream region, including Tiaoxi, Yili and Tiaohe rivers, and Wujin and Zhihu ports, ﬂow into Taihu Lake; and water tributaries in the downstream region, including the channels draining northwardly to the Yangtze River, Huangpu River, and the channels draining southwardly to Hangzhou Bay, ﬂow out from Taihu Lake (Wang and Dou, 1998; Qin et al., 2004). The site for this study is in the Yili River watershed, which is an important water tributary in the upstream region. Paddy rice is the dominant crop in this area, and the other major crops are wheat, rapeseed, and soybeans. The dominant cultivation crop rotations are rice-wheat or rice-rapeseed. To evaluate the eﬀects of anthropogenic activities on surface water quality, 20 sampling sites were chosen (Fig. 1). The sites were categorized according to their pollution sources derived from diﬀerent anthropogenic activities (Table I). Category 1: sites 1–6 are on the streams that receive pollutants from agricultural activities; Category 2: sites 7–10 are on the streams that receive pollutants from rural domestic sewage and litter; Category 3: sites 11–13 are on the intensive ﬁsh ponds in an aquaculture area, which are inﬂuenced by residual ﬁsh food and ﬁsh excreta; Category 4: sites 14–16 are on the streams that are polluted by animal waste and washing tool from nearby livestock farm; Category 5: sites 17 and 18 are on the streams that receive pollutant from urban domestic sewage of Yixing City; Category 6: sites 19 and 20 are the reservoirs in drinking water protection area free of nearby factories and other pollution sources. These both sites are located in relatively non-aﬀected area and allow a comparison between relatively pristine water bodies and anthropogenic inﬂuenced water bodies.
ANTHROPOGENIC IMPACT ON SURFACE WATER QUALITY
Study area and 20 sampling sites in the Yili River watershed of Taihu Lake, China.
TABLE I Categories of sampling sites based on pollution sources from diﬀerent land use types in the Yili River watershed of Taihu Lake, China Category Sampling site No. Land use type
Category Sampling site No. Land use type
1 2 3 4 5 6 7 8 9 10
Agricultural area Agricultural area Agricultural area Agricultural area Agricultural area Agricultural area Rural residential area Rural residential area Rural residential area Rural residential area
11 12 13 14 15 16 17 18 19 20
Aquaculture area Aquaculture area Aquaculture area Livestock farming area Livestock farming area Livestock farming area Urban residential area Urban residential area Drinking water protection area Drinking water protection area
Sampling and analyzing Surface water samples (0–0.5 m) were collected from 20 sites in spring (April), summer (August), autumn (November) in 2005, and winter (January) in 2006. On each sampling date, three replicates were collected at each sampling site. The water samples were preserved in 1-L polypropylene sampling bottles at 4 ◦ C in darkness and analyzed within 48 h. A saturated mercuric chloride solution was used at a ﬁnal concentration of 0.2 mL L−1 to stop all microbiological activities in the water samples. The measured parameters of water quality included temperature, transparency, pH, electrical conductivity (EC), dissolved oxygen (DO), dissolved organic carbon (DOC), total nitrogen (TN), ammonium nitrogen (NH4 -N), nitrate nitrogen (NO3 -N), total phosphorus (TP), and soluble reactive phosphorus (SRP). The temperature, transparency, pH, EC, and DO were determined in situ. The temperature and EC were measured with Yellow Springs Instrument (YSI) meter Model 33 (YSI, Yellow Springs, Ohio, USA), the pH with a pH meter Beckman Model F8 253 (Beckman, Fullerton, CA, USA), the transparency with a Secchi disc with 30-cm diameter, and the DO with a YSI oxygen meter Model 57 (YSI, Yellow Springs, Ohio, USA). The NO3 -N, NH4 -N, TN, SRP, and TP were determined in laboratory using a Shimadzu UV mini-1240 spectrophotometer (Shimadzu, Kyoto, Japan) with a minimum detection limit of 0.001 mg L−1 . The NO3 -N concentration was measured directly using colorimetry, with double wavelengths set at 220 and 275 nm. For the determination of NH4 -N, the samples were developed with indophenol blue and then were measured using colorimetry with the wavelength set at 635 nm (Lu, 2000). For TN measurement, samples were digested with potassium persulfate and analyzed with the same method used for NO3 -N. The SRP concentration was measured using the ammonium molybdate method with wavelength set at 660 nm (Lu, 2000). The samples for TP measurement were
H. XU et al.
digested with potassium persulfate in acidic conditions and analyzed with the ammonium molybdate method similar to SRP. The DOC was determined using high temperature combustion techniques on a Shimadzu model TOC-5000 (Shimadzu, Kyoto, Japan), equipped with an ASI-5000A autosampler. In April 2005, water samples collected from some sites and the center of the lake were passed through + 15 N anion and cation exchange resins to extract and concentrate NO− 3 and NH4 for measurement of − + 15 −1 natural abundance (δ N). The NO3 and NH4 were eluted from column with 1 mol L HCl, separately. + NO− 3 in the resultant solution was converted into NH4 by distillation in the presence of Cu-Al-Zn alloy. All samples were acidiﬁed with HCl, concentrated in a water bath at 80 ◦ C, and analyzed for their δ 15 N using MAT-251 isotope mass spectrometer, with an analytical error of 0.2 δ 15 N (Sun et al., 1991). The quality of analytical data was ensured through careful standardization, procedural blank measurements, and spiked and duplicate samples. All chemical analyses were carried out at least in duplicate, and results did not diﬀer by more than 5% of the mean. Distilled deionized water was substituted for test solution when running blanks for the chemical parameters measured. Multivariate analysis Multivariate analysis of the water quality data set was performed through PCA and CA using the SPSS 11.0 statistic software packages. The experimental data set was ﬁrst standardized through z-scale transformation to avoid misclassiﬁcation due to wide diﬀerences in data dimensions. PCA, based on the standardized sample matrix (20 sites × 4 seasons × 10 variables), was performed to identify signiﬁcant parameters for the characterization of the water quality. The scree plot was applied to identify the number of principal components (PCs) to understand the underlying data structure. Any factor with an eigenvalue greater than “1” was considered signiﬁcant according to the criteria of Cattell and Jaspers (Liang and Yu, 2000). A correlation matrix of these variables was computed, and factor loading was deﬁned to explore the nature of variation and principal patterns among them (Aruga et al., 1993). Also, a varimax rotation was performed to address the problem of variables loading moderately on one or more axes. To evaluate the eﬀect of anthropogenic activities on surface water quality, multivariate CA (Q-CA) was performed to group 20 sample sites in four seasons based on each signiﬁcant principal factor obtained by PCA. The between-group linkage method was applied to normalized data using Euclidean distance to quantify similarity between sites. The results obtained were represented in a dendrogram. RESULTS Physicochemical parameters The mean and standard deviation of water quality parameters in the six categories of water bodies are presented in Table II. Mean water temperature was strongly seasonal and ranged from a minimum of 8.50 ◦ C in winter to a maximum of 31.60 ◦ C in summer. There was no signiﬁcant diﬀerence in temperature (P < 0.05) between sampling sites. Throughout the year, mean transparency ranged from 0.25 to 2.25 m. The sites in aquaculture (category 3), livestock farming (category 4), and urban residential areas (category 5) showed lower water transparency than those in agricultural (categories 1) and rural residential areas (category 2). The highest water transparency was recorded in the sites located in drinking water protection area (category 6). Seasonally, the highest value of water transparency occurred in winter in all sampling sites. Mean pH of water bodies, ranging from 7.45 to 8.87, showed a signiﬁcant spatial variation. In spring, summer, and autumn, the highest pH was recorded in the category 3, whereas in winter, it was in category 5. No clear seasonal trend was found at any of the sampling sites. Throughout the year, EC varied between 0.30 and 0.71 dS m−1 . The distribution of EC among sample sites showed spatial variability that it was lower in categories 6 and 3 relative to the other categories. Seasonally, EC was lower in summer and autumn than in spring and winter seasons. The DOC concentrations showed ﬂuctuations between seasons, with the lowest value (less than 12.95
(0.33)d) (0.67) (0.85) (0.66) (0.57) (0.07) (0.34) (0.30) (0.40) (0.23) (0.14) (0.00) (1.08) (0.49) (0.46) (0.40) (0.42) (0.35) (0.52) (0.30) (0.72) (0.46) (0.07) (0.71)
20.48 20.80 20.83 20.80 21.30 20.15 31.42 31.55 31.23 31.27 31.10 31.60 23.75 22.88 23.27 22.63 22.70 22.95 8.55 8.65 8.50 9.27 8.95 8.50
1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6
Season (0.08) (0.10) (0.06) (0.10) (0.01) (0.35) (0.15) (0.04) (0.13) (0.05) (0.04) (0.35) (0.33) (0.00) (0.13) (0.24) (0.08) (0.14) (0.22) (0.10) (0.06) (0.05) (0.04) (0.35)
m 0.68 0.65 0.33 0.36 0.36 1.75 0.82 0.55 0.35 0.25 0.38 1.25 0.73 0.90 0.30 0.38 0.44 1.30 0.83 0.88 0.57 0.35 0.38 2.25
Transparency dS 0.69 0.71 0.44 0.66 0.70 0.37 0.37 0.38 0.43 0.56 0.54 0.30 0.49 0.56 0.44 0.63 0.59 0.33 0.54 0.60 0.43 0.69 0.61 0.37 (0.08) (0.03) (0.06) (0.05) (0.01) (0.01) (0.04) (0.03) (0.11) (0.09) (0.02) (0.01) (0.06) (0.03) (0.11) (0.09) (0.06) (0.01) (0.06) (0.07) (0.05) (0.17) (0.01) (0.02)
ECa) 2.35(0.37) 0.64 (0.19) 0.01 (0.00) 1.01 (0.36) 0.06 (0.01) 0.58 (0.17) 0.43 (0.10) 0.00 (0.01) 0.01 (0.01) 0.60 (0.10) 0.21 (0.01) 0.31 (0.01) 0.50 (0.17) 0.24 (0.11) 1.91 (0.52) 0.57 (0.26) 0.46 (0.08) 0.58 (0.25) 4.97 (0.98) 4.40 (0.99) 1.56 (0.47) 2.58 (1.27) 0.68 (0.37) 0.58 (0.03)
NO3 -N 0.37 0.72 0.02 4.49 21.30 0.00 0.14 0.50 0.04 1.59 6.71 0.03 0.19 0.87 0.24 3.16 9.77 0.04 0.36 0.10 0.24 10.26 14.35 0.03
mg (0.17) (0.14) (0.02) (1.68) (3.78) (0.00) (0.04) (0.08) (0.02) (0.47) (2.41) (0.01) (0.08) (0.18) (0.06) (0.54) (0.51) (0.03) (0.13) (0.01) (0.06) (2.05) (4.55) (0.01)
NH4 -N 3.45 1.38 0.72 6.32 22.40 0.63 0.70 0.88 0.72 2.67 7.50 0.36 1.35 1.34 3.01 3.93 10.50 0.64 6.01 4.40 2.62 16.37 16.29 0.61
L−1 (0.46) (0.51) (0.34) (1.20) (3.39) (0.19) (0.14) (0.24) (0.26) (0.59) (2.12) (0.01) (0.18) (0.40) (0.43) (0.32) (0.71) (0.26) (1.03) (1.40) (0.82) (1.52) (4.02) (0.04)
in parentheses are standard deviations.
(0.85) (1.39) (2.04) (0.64) (8.46) (0.75) (2.31) (1.20) (3.00) (2.21) (1.34) (0.87) (1.83) (2.57) (1.08) (2.04) (0.33) (0.68) (0.68) (0.21) (0.77) (2.08) (0.35) (0.03) d) Data
10.73 14.57 24.79 17.27 13.50 8.46 12.97 16.71 13.88 14.90 10.75 7.62 11.85 19.79 16.23 27.02 11.23 5.23 4.60 5.24 6.38 12.95 5.70 3.27
(0.26) (0.29) (0.15) (0.26) (0.14) (0.07) (0.19) (0.22) (0.39) (0.23) (0.13) (0.07) (0.23) (0.25) (0.36) (0.64) (0.64) (0.07) (0.65) (0.21) (0.23) (0.44) (0.07) (0.07)
8.45 8.18 8.87 8.20 7.80 7.85 7.79 7.41 8.73 7.72 8.09 7.95 8.03 7.63 8.60 8.27 8.15 7.55 7.67 7.55 8.13 8.70 8.65 7.45
Water quality of the six categories of water bodies in Yili River watershed of Taihu Lake during diﬀerent sampling seasons
0.01 0.02 0.01 0.51 1.12 0.02 0.00 0.11 0.00 0.43 0.56 0.00 0.00 0.01 0.01 0.32 0.51 0.00 0.01 0.03 0.01 0.63 0.92 0.00
(0.01) (0.03) (0.00) (0.12) (0.23) (0.01) (0.00) (0.03) (0.00) (0.08) (0.08) (0.00) (0.00) (0.00) (0.01) (0.07) (0.13) (0.00) (0.01) (0.02) (0.00) (0.06) (0.31) (0.00)
0.05 0.06 0.02 0.74 1.30 0.04 0.12 0.23 0.12 0.50 0.66 0.03 0.04 0.08 0.03 0.51 0.65 0.00 0.02 0.03 0.02 0.73 1.03 0.02
(0.01) (0.03) (0.01) (0.15) (0.28) (0.01) (0.01) (0.10) (0.01) (0.10) (0.08) (0.01) (0.01) (0.03) (0.01) (0.18) (0.14) (0.00) (0.01) (0.02) (0.01) (0.13) (0.33) (0.01)
ANTHROPOGENIC IMPACT ON SURFACE WATER QUALITY 769
H. XU et al.
mg L−1 ) in winter. In spring, the DOC concentration (24.79 mg L−1 ) was the highest in category 3, whereas in summer and autumn, it was the highest in categories 2 and 4 (16.71 and 27.02 mg L−1 , respectively). Throughout the year, the lowest concentration of DOC occurred in category 6, with mean concentrations ranging from 3.27 mg L−1 in winter to 13.88 mg L−1 in summer. The sites inﬂuenced by livestock farming and urban domestic sewage (categories 4 and 5) showed high SRP concentrations during all the year, with relatively high values in spring and winter and low values in summer and autumn. Mean SRP concentrations in other sites were very low along the year except for a high value of 0.11 mg L−1 occurred in category 2 in summer. TP distribution showed similar spatial trend to SRP, with high values in categories 4 and 5 and low values in others categories. Seasonally, however, the highest TP concentrations occurred in summer in all of the categories except for categories 4 and 5. In general, the highest and the lowest (less than 0.60 mg L−1 ) NO3 -N concentrations occurred in winter and summer, respectively. The sites located in drinking water protection area (category 6) showed steadily low NO3 -N concentrations (less than 0.58 mg L−1 ) in all samplings during the year, whereas the streams in agricultural area (category 1) showed higher NO3 -N concentrations in spring and winter, which were mainly inﬂuenced by agricultural activities. The streams in rural residential area (category 2) showed higher mean NO3 -N concentration (4.40 mg L−1 ) only in winter. The NH3 -N concentration showed similar temporal pattern to NO3 -N concentration, with high values in spring and winter and low values in summer and autumn. The high NH3 -N concentrations were recorded mainly in the streams that were inﬂuenced by livestock farming and urban domestic sewage (categories 4 and 5). The NH3 -N concentrations in other sites were extremely low (less than 0.72 mg L−1 ). The distribution of TN showed signiﬁcant variation. The lowest concentration of TN was observed in category 6. Similar to NH4 -N, the high values of TN occurred in the categories 4 and 5. Seasonal variation was also observed, with lower values in rainy season (August and October) and higher values in dry season (February and April). 15
N natural abundance
+ 15 The δ 15 NO− 3 and δ NH4 determined at some sampling sites and Taihu Lake are shown in Table + 15 III. The reservoirs in drinking water protection area showed δ 15 NO− 3 below 10 , whereas δ NH4 was − + 15 15 below detection limit. The streams in agricultural area had δ NO3 around 10 , but δ NH4 was − 15 undetectable because of low NH+ 4 concentrations. The δ NO3 in ﬁsh ponds was as high as 31.77 . The streams in livestock farming and urban residential areas showed high δ 15 NH+ 4 around 20 , higher − + − 15 15 15 than δ NO3 . The Taihu Lake had a δ NH4 value of 16.86 and a δ NO3 value of 11.61 .
TABLE III + − − 15 15 natural abundance of NH+ 4 (δ NH4 ) and NO3 (δ NO3 ), sampled in April 2005, in the water bodies with diﬀerent land use types
Water body/land use type
Sampling site No.
δ15 NH+ 4
δ15 NO− 3
Reservoir/drinking water protection area
19 20 1 4 17 18 12 13 14 15 16 Center of lake
-a) 23.92 19.95 21.25 21.97 25.00 16.86
8.73 7.19 13.23 9.99 16.09 12.67 13.73 31.77 14.32 10.00 18.75 11.61
Stream/agricultural area Stream/urban residential area Fish pond/ﬁsh farming area Stream/livestock farming area
Taihu Lake a) Below
detection limit of 0.2 mg N L−1 .
ANTHROPOGENIC IMPACT ON SURFACE WATER QUALITY
PCA on parameters of water quality PCA was performed to identify characteristics of water quality variables of all studied sites. Bartlett’s sphericity test showed the signiﬁcance level was 0 (less than 0.05), indicating that the variables were not orthogonal but correlated, therefore, allowing to explain the data variability with a lesser number of variables by PCA. Scree plot (Fig. 2) showed the sorted eigenvalues from large to small as a function of the PC number. Three PCs having eigenvalues greater than 1 were extracted.
Scree plot of the eigenvalues.
The variable loading matrix, eigenvalues, and total and cumulative variance values are presented in Table IV. Three combined PCs explained 80.84% of the total variation. The ﬁrst factor accounted for 38.91% of the total variance and signiﬁcantly correlated with NH4 -N, TN, SRP, and TP, representing NH4 -N, TN, SRP, and TP pollution from strong anthropogenic impacts. The second factor accounted for 22.70% of the total variance and had strong positive loading on NO3 -N and strong negative loading on temperature. It represented the NO3 -N pollution. The third factor explained 19.23% of the total variation and had moderate positive loadings on pH and DOC and moderate negative loading on transparency, which denoted the organic pollution. High pH and low transparency may be taken as an ecological response due to growth of phytoplankton. Using PCA, the 10 original variables were reduced to three key independent factors, representing three diﬀerent dimension of the water quality. TABLE IV Loadings of experimental variables, eigenvalues, and total and cumulative variance values on the ﬁrst three rotated principal components (PCs) PC
Dissolved organic carbon
PC1 PC2 PC3
−0.132 0.912 −0.023
0.973 0.034 0.095
0.904 0.351 0.120
0.973 −0.037 0.154
0.955 −0.117 0.186
−0.029 −0.541 0.672
−0.137 −0.868 0.142
PC1 PC2 PC3
0.055 −0.001 0.689
−0.306 0.074 −0.798
0.366 0.498 0.509
3.891 2.270 1.923
38.91 22.70 19.23
38.91 61.61 80.84
Classiﬁcation of sampling sites based on pollution factors To examine anthropogenic impact on the water quality, multivariate CA was applied to classify the 20 sampling sites based on each signiﬁcant pollution factor obtained by PCA during the study period.
H. XU et al.
For NH4 -N, TN, SRP, and TP, the 20 sampling sites were classiﬁed into three groups in spring, summer and fall, and two groups in winter (Fig. 3). The sampling sites 14–18 showed higher concentrations of NH4 -N, TN, SRP, and TP compared with other sites in all years, with the highest and lowest values in spring and summer, respectively (Table V).
Fig. 3 Dendrogram of Q-hierarchical cluster analysis for 20 sampling sites based on pollution factor 1, including NH 4 -N, total N, soluble reactive phosphorus, and total P.
Based on the pollution factor 2 (NO3 -N), the 20 sampling sites were classiﬁed into two groups in spring, summer, fall, and winter (Fig. 4). Table VI shows the average values of NO3 -N for each classiﬁed group. In spring, the sampling sites 1–6, inﬂuenced by agricultural runoﬀ, were classiﬁed into one group and showed higher mean NO3 -N concentration (2.35 mg L−1 ) than other groups. In summer, all sampling sites showed low NO3 -N concentrations. The average value in group 2 (sites 7–13) was low to 0 mg L−1 . In fall, group 2 consisting of sampling sites 11–13 in ﬁsh farming area showed higher NO3 -N concentration (1.91 mg L−1 ) than group 1 consisting of the rest 17 sampling sites. In winter, sampling sites 1–10 and 14 were classiﬁed into one group and showed high average NO3 -N concentration (4.74 mg L−1 ), which are on streams in agricultural and rural residential areas except for site 14 that is in livestock farming area. The other sampling sites were classiﬁed into one group, with average NO3 -N concentration of 1.21 mg L−1 . Dendrograms of CA based on pollution factor 3 (i.e., DOC, pH, and transparency) are presented in Fig. 5, and the average values of related variables are listed in Table VII. The sampling sites were classiﬁed into three groups by multivariate CA in spring, summer, and fall and two groups in winter.
ANTHROPOGENIC IMPACT ON SURFACE WATER QUALITY
TABLE V Average values of pollution factor 1, including NH4 -N, total N, soluble reactive phosphorus (SRP), and total P, for each group of sampling sites computed by hierarchical cluster analysis Season
Total P mg
1 2 3 1 2 3 1 2 3 1 2
1.98 6.32 22.49 0.70 2.67 7.50 1.58 3.93 10.50 4.18 16.33
± 1.34a) ±1.20 ± 3.39 ± 0.23 ± 0.59 ± 2.12 ± 0.83 ± 0.32 ± 0.71 ± 2.17 ± 2.88
0.34 4.49 21.30 0.20 1.59 6.71 0.36 3.16 9.77 0.22 11.90
± 0.31 ±1.68 ± 3.78 ± 0.20 ± 0.47 ± 2.47 ± 0.34 ± 0.54 ± 0.51 ± 0.16 ± 3.50
L−1 0.04 0.74 1.30 0.13 0.50 0.66 0.04 0.73 1.03 0.02 0.85
± ± ± ± ± ± ± ± ± ± ±
0.02 0.15 0.28 0.08 0.10 0.08 0.03 0.15 0.33 0.01 0.25
0.01 0.51 1.12 0.03 0.43 0.56 0.01 0.63 0.92 0.01 0.74
± standard deviation.
Dendrogram of Q-hierarchical cluster analysis for 20 sampling sites based on pollution factor 2, NO 3 -N.
± ± ± ± ± ± ± ± ± ± ±
0.01 0.12 0.23 0.05 0.08 0.08 0.01 0.06 0.31 0.02 0.23
In spring, group 1 (sampling sites 1–6, 19, and 20) corresponds to relatively less polluted sites, with the lowest value of DOC (10.17 mg L−1 ) and the highest value of transparency (0.95 m). Group 2 (sites
H. XU et al.
TABLE VI Average values of NO3 -N for each group of sampling sites computed by hierarchical cluster analysis Season
Group 2 mg
Spring Summer Fall Winter
0.49 0.42 0.42 1.21
± ± ± ±
0.43 0.15 0.20 0.69
L−1 2.35 ± 0.37 0.00 1.91 ± 0.52 4.74 ± 0.97
Fig. 5 Dendrogram of Q-hierarchical cluster analysis for 20 sampling sites based on pollution factor 3, including transparency, pH and dissolved organic carbon.
7–10 and 14–18) corresponds to moderately polluted sites, with higher value of DOC (15.72 mg L−1 ) and lower value of transparency (0.53 m) than group 1. Group 3 included three sites inﬂuenced by ﬁsh farming and yielded the highest average values of DOC (24.79 mg L−1 ) and pH (8.87) and the lowest average value of transparency (0.33 m). In summer, group 1, consisted of all sampling sites inﬂuenced by rural residential subsistence except for sampling site 11, showed the highest average values of DOC (16.75 mg L−1 ) and pH (7.99). Group 2 included 13 sampling sites aﬀected by agriculture, residential subsistence, and livestock farming, with average values of transparency, pH, and DOC being 0.49 m, 7.61 and 12.98 mg L−1 , respectively. Group 3 included sampling sites 19 and 20, with the highest value of transparency (1.25 m) and the lowest value of DOC (7.62 mg L−1 ). In fall, group 3 (sites 13–15)
ANTHROPOGENIC IMPACT ON SURFACE WATER QUALITY
TABLE VII Average values of pollution factor 3, including transparency, pH and dissolved organic carbon, for each group of sampling sites computed by hierarchical cluster analysis Season
1 2 3 1 2 3 1 2 3 1 2
m 0.95 0.53 0.33 0.55 0.49 1.25 0.65 1.30 0.38 0.91 0.35
8.30 8.19 8.87 7.99 7.61 7.95 8.05 7.55 8.27 7.81 8.70
± ± ± ± ± ± ± ± ± ± ±
0.52 0.18 0.06 0.28 0.14 0.35 0.31 0.14 0.24 0.56 0.05
Dissolved organic carbon ± ± ± ± ± ± ± ± ± ± ±
0.35 0.25 0.15 0.46 0.48 0.07 0.45 0.77 0.64 0.54 0.44
mg L−1 10.17 ± 15.72 ± 24.79 ± 16.75 ± 12.98 ± 7.62 ± 14.76 ± 5.23 ± 27.02 ± 5.04 ± 12.95 ±
1.30 1.79 2.04 1.04 2.32 0.87 3.99 0.68 2.04 1.05 2.08
showed the highest DOC (27.02 mg L−1 ) and pH (8.27) and the lowest transparency (0.38 m). Group 2 consisted of sites 19 and 20, which showed the highest value of transparency (1.3 m) and the lowest value of DOC (5.23 mg L−1 ). In winter, DOC concentrations in all sampling sites were relatively low. However, the sampling sites inﬂuenced by livestock farming showed relatively high pH (8.70) and DOC (12.95 mg L−1 ). DISCUSSION Anthropogenic impact on surface water quality PCA is useful for considering several related random environmental variables simultaneously and thus for identifying a new, smaller set of uncorrelated variables that accounts for a large proportion of the total variance in the original variables (Lau and Lane, 2002). Using PCA, the 10 original variables were reduced to three key independent factors that inﬂuence water quality in Yili River watershed (Fig. 2). Factor loadings are the projection of the original variables on the subspace of the PCs, which show the mutual relationships among the variables. Factor 1 accounted for the majority of total variance and had high loadings on NH4 -N, TN, SRP, and TP (Table IV), reﬂecting that over-loadings of NH4 -N, TN, SRP, and TP are responsible for the heaviest pollution problem in the watershed. Factor 2 was positively related to NO3 -N and negatively related to temperature, suggesting that NO3 -N pollution is also an important environmental pollution to aquatic systems. Water temperature showed seasonal transition characters, so NO3 -N concentrations in stream ﬂow were highly variable and showed seasonal pattern during the sampling period. A positive relationship between factor 3 and DOC suggests that organic pollution is another principal environmental problem in this watershed. The pH and transparency were negatively related to DOC, suggesting that pH and transparency are sensitive to organic pollution and reﬂect a feedback of ecology. The CA is useful in solving classiﬁcation problems, whose objective is to place cases or variables into groups such that the degree of association is strong between members of the same cluster and weak between members of diﬀerent clusters (Brogueira, and Cabe¸cadas, 2006). In this study, CA showed strong spatial variations of principal pollution factors between sampling sites, which indicates that diﬀerent human activities have diﬀerent eﬀects on water quality. The dendrogram provides a visual summary of the clustering processes, presenting a picture of the groups and their proximity. For NH4 -N, TN, SRP, and TP pollution, three very well-diﬀerentiated clusters in spring, summer, and fall and two clusters in winter can be observed (Fig. 3). Sites 14–16, 17 and 18 showed separate proximity in dendrogram with high concentrations of NH4 -N, TN, SRP, and TP during all the year
H. XU et al.
(Table V). These sampling sites are located in urban residential area and livestock farming area, indicating that rapid development of urbanization and livestock farming is the principal reason for the deterioration of water quality in rivers (Li et al., 2000; Chen et al., 2003). Consequently, it is necessary to control pollutions from domestic sewage and livestock farming in the Taihu watershed. For NO3 -N, the sites 1–6 were in a same group in spring and winter with high NO3 -N concentration (Fig. 4 and Table VI). Considering that these sampling sites are located in an agricultural area, it can be inferred that agricultural activities signiﬁcantly contribute to NO3 -N pollution in streams. Nitrate is more associated with the use of organic and inorganic fertilizers (Mattikalli and Richards, 1996; Basnyat et al., 1999). It has been reported that N fertilizer applied to farmland is the main source of N pollution of the surface water in China (Yan et al., 1999; Si et al., 2000). In winter, sites 7–10 located in rural residential area near to Taihu Lake were classiﬁed into one group together with sites 1–6, showing similar NO3 -N concentration in streams. Because water ﬂow can run oﬀ from the lake to nearby streams in the dry season, high NO3 -N concentration in the lake is the principal reason for it. In summer and fall, despite two groups were recognized, all sampling sites showed low NO3 -N concentration, reﬂecting low NO3 -N pollution stress on water quality during that period. According to the dendrograms based on organic pollution (Fig. 4), three principal sources of organic pollution were recognized in Taihu Lake region. In fall, sites 14–16, located in livestock farming area, were classiﬁed into one group with the highest DOC concentration (Table VII), suggesting that livestock farming produced heavy organic pollution. In spring, sites 11–13, located in aquiculture area, showed proximity in dendrogram, which had a high DOC concentration. The aquaculture in Taihu Lake region accounts for about 25% of all aquaculture activity in China. The DOC from residual ﬁsh food and ﬁsh excreta increased the organic pollution of the water bodies. In summer, sites 7–10, located in rural residential area, showed similar trend, with high DOC concentrations. It can be inferred that domestic wastewater from rural residential area is also an important source of organic pollution in the watershed. In winter, all sampling sites showed low DOC concentrations; however, livestock farming still produced relatively heavier organic pollution than other anthropogenic activities. N pollution sources of Taihu Lake The NO3 -N is the principal N form in the lake during all the year except for the summer (Table VIII). From the above discussion, it can be concluded that streams in agricultural area were the most vulnerable water bodies to NO3 -N pollution by runoﬀ or leaching, suggesting that NO3 -N in Taihu Lake was originated mainly from farmland. In summer, NH4 -N concentration in Taihu Lake was higher than NO3 -N concentration and TP concentration also increased. Because streams in livestock farming and urban residential areas were heavily polluted by NH4 -N and TP, it can be concluded that domestic sewage and animal wastes contribute signiﬁcantly to NH4 -N and TP pollution of the lake by runoﬀ during the summer rainy season. TABLE VIII Seasonal concentrations of NO3 -N, NH4 -N, total N, total P, and soluble reactive phosphorus (SRP) in Taihu Lake Season
Spring Summer Autumn Winter
3.21 0.47 1.53 6.52
1.39 0.90 0.38 0.00
mg 6.26 2.28 1.97 6.76
0.07 0.118 0.0032 0.01
0.03 0.04 0.02 0.00
Mean Standard deviation
The sources of NO3 in ground water were distinguished based on the diﬀerence between the δ 15 NO− 3 originating from chemical N fertilizer and that from animal wastes (Gormly and Spalding, 1979). This
ANTHROPOGENIC IMPACT ON SURFACE WATER QUALITY
method, however, is now being questioned due to the recent evidence that denitriﬁcation in ground water can result in increased 15 N natural abundance of the residual NO− 3 (Bates and Spalding, 1998). We 15 attempt to cite another diﬀerence in N natural abundance between NH4 -N and NO3 -N originating from chemical N fertilizer and that from other pollution sources as an indicator to distinguish sources of nitrogen in Taihu Lake. The δ 15 NH+ 4 was much higher in Taihu Lake than in streams of agricultural area under the inﬂuence of chemical fertilizers, and it was close to that in streams polluted by the human and animal wastes (Table III). These results suggest that urban domestic sewage and livestock farming contributed mainly to NH4 -N pollution in lake water. Xing et al. (2001) reported that high δ 15 NH+ 4 and PO4 -P could serve as an evidence of domestic sewage and animal waste pollution. Xie et al. (2007) found that in to 14.99 . The δ 15 NO− the wheat growing season, δ 15 NO− 3 in leachate from soil ranges from 9.22 3 in Taihu Lake was in this range. As a result, we suggest that farmland fertilizer contributed mainly to NO3 -N pollution in Taihu Lake. CONCLUSIONS The surface water quality inﬂuenced by various human activities was investigated on 20 sampling sites in the Taihu Lake region. The PCA suggested that at least three latent factors were necessary for interpreting the water quality of Taihu Lake region. CA based on each pollution factor obtained by PCA suggested that the urban residential subsistence and livestock farming contributed signiﬁcantly to NH4 -N, total N, SRP, and total P pollution of the surface water. The agricultural activities contributed mainly to NO3 -N pollution of the surface water. The organic pollution of surface water resulted mainly from livestock farming, but aquaculture and rural residential subsistence were also important sources in diﬀerent seasons. Further analysis showed that farmlands contributed signiﬁcantly to N pollution in Taihu Lake, while urban residential subsistence and livestock farming also polluted water in the summer rainy season. This study identiﬁed the major factors that aﬀected water quality, and it might help to improve the design of the monitoring network for eﬀective management of water sources. REFERENCES Adams, S., Titus, R., Pietersen, K., Tredoux, G. and Harris, C. 2001. Hydrochemical characteristic of aquifers near Sutherland in the Western Karoo, South Africa. J. Hydrol. 241: 91–103. Aruga, R., Negro, G. and Ostacoli, G. 1993. Multivariate data analysis applied to the investigation of river pollution. Fresen. J. Anal. Chem. 346: 968–975. Basnyat, P., Teeter, L. D., Lockaby, B. G. and Flynn, K. M. 1999. The use of remote sensing and GIS in watershed level analyses of non-point source pollution problems. Forest Ecol. Manag. 128: 65–73. Bates, H. K. and Spalding, R. F. 1998. Aquifer denitriﬁcation as interpreted from in situ microcosm experiments. J. Environ. Qual. 27: 174–182. Brogueira, M. J. and Cabe¸cadas, G. 2006. Identiﬁcation of similar environmental areas in Tagus Estuary by using multivariate analysis. Ecol. Indic. 6: 508–515. Cai, Q. M., Gao, X. Y., Chen, Y. W., Ma, S. W. and Dokulil, M. 1996. Dynamic variations of water quality in Taihu Lake and multivariate analysis of its inﬂuential factors. Chinese Geogr. Sci. 6: 364–374. Chen, H. S. 2001. Restoration project of the ecosystem in Tai Lake. Resour. Environ. Yangtze Basin (in Chinese). 10: 173–178. Chen, Y. W., Fan, C. X., Teubner, K. and Dokulil, M. 2003. Changes of nutrients and phytoplankton chlorophyll-a in a large shallow lake, Taihu, China: An 8-year investigation. Hydrobiologia. 506: 273–279. Gao, C., Zhu, J. G., Zhu, J. Y., Gao, X., Dou, Y. J. and Hosen, Y. 2004. Nitrogen export from an agriculture watershed in the Taihu Lake area, China. Environ. Geochem. Hlth. 26: 199–207. Guo, H. Y., Wang, X. R. and Zhu, J. G. 2004. Quantiﬁcation and index of non-point source pollution in Lake Taihu region with GIS. Environ. Geochem. Hlth. 26: 147–156. Gormly, J. R. and Spalding, R. F. 1979. Sources and concentrations of nitrate-nitrogen in ground water of the central platte region, Nebraska. Ground Water. 17: 291–301. Huang, Y. P. and Zhu, M. 1996. The water quality of Lake Taihu and its protection. GeoJournal. 40: 39–44. Lau, S. S. S. and Lane, S. N. 2002. Biological and chemical factors inﬂuencing shallow lake eutrophication: A long-term study. Sci. Total Environ. 228: 167–181.
H. XU et al.
Lee, J. Y., Cheon, J. Y., Lee, K. K., Lee, S. Y. and Lee, M. H. 2001. Statistical evaluation of geochemical parameter distribution in a ground water system contaminated with petroleum hydrocarbons. J. Environ. Qual. 30: 1548–1563. Li, R. G., Xia, Y. L., Wu, A. Z. and Qian, Y. S. 2000. Pollutants sources and their discharging amount in Taihu Lake area of Jiangsu Province. J. Lake Sci. (in Chinese). 12: 147–153. Liang, Y. Z. and Yu, R. Q. 2000. Chemometrics (in Chinese). Central South University Press, Changsha, China. Lu, R. K. 2000. Analytical Methods of Soil Agrochemistry (in Chinese). China Agricultural Science and Technology Press, Beijing. Mattikalli, N. M. and Richards, K. S. 1996. Estimation of surface water quality changes in response to land use change: Application of the export coeﬃcient model using remote sensing and geographical information system. J. Environ. Manage. 48: 263–282. Qin, B. Q., Hu, W. P. and Chen, W. M. 2004. Process and Mechanism of Environmental Changes of the Taihu Lake (in Chinese). Science Press, Beijing. Qin, B. Q., Xu, P. Z., Wu, Q. L., Luo, L. C. and Zhang, Y. L. 2007. Environmental issues of Lake Taihu, China. Hydrobiologia. 581: 3–14. Quan, W. M. and Yan, L. J. 2002. Eﬀects of agricultural non-point source pollution on eutrophication of water body and its control measure. Acta Ecol. Sin. (in Chinese). 22(3): 291–299. Reghunath, R., Murthy, T. R. S. and Raghavan, B. R. 2002. The utility of multivariate statistical techniques in hydrogeochemical studies: an example from Karnataka, India. Water Res. 36: 2437–2442. Shen, L., Lin, G. F., Tan, J. W. and Shen, J. H. 2000. Genotoxicity of surface water samples from Meiliang Bay, Taihu Lake, Eastern China. Chemosphere. 41: 129–132. Si, Y. B., Wang, S. Q. and Chen, H. M. 2000. Water eutrophyication and losses of nitrogen and phosphrus in farmland. Soils (in Chinese). 32: 188–193. Simeonov, V., Simeonova, P. and Tsitouridou, R. 2004. Chemometric quality assessment of surface waters: two case studies. Chem. Eng. Ecol. 11(6): 449–469. Sun, G. Q., Xing, G. X. and Cao, Y. C. 1991. Study on variation in 15 N natural abundance. I. Characteristics of variation in 15 N natural abundance of forest soils. Pedosphere. 1: 277–281. Wang, S. M. and Dou, H. S. 1998. China Limnological Chorography (in Chinese). Science Press, Beijing. Wang, X. L., Lu, Y. L., Han, J. Y., He, G. Z. and Wang, T. Y. 2007. Identiﬁcation of anthropogenic inﬂuences on water quality of rivers in Taihu watershed. J. Environ. Sci. 19: 475–481. Wunderlin, D. A., Diaz, M. P., Ame, M. V., Pesce, S. F., Hued, A. C. and Bistoni, M. A. 2001. Pattern recognition techniques for the evaluation of spatial and temporal variations in water quality. A case study: Suquia River Basin (Cordoba, Argentina). Water Res. 35: 2881–2894. Xie, Y. X., Xiong, Z. Q., Xing, G. X., Sun, G. Q. and Zhu, Z. L. 2007. Assessment of nitrogen pollutant sources in surface water of Taihu Lake region. Pedosphere. 17: 200–208. Xing, G. X., Cao, Y. C., Shi, S. L., Sun, G. Q., Du, L. J. and Zhu, J. G. 2001. N pollution sources and denitriﬁcation in waterbodies inTaihu Lake region. Sci. China (Series B). 44: 304–314. Yan, W. J., Yin, C. Q. and Zhang, S. 1999. Nutrient budget and biogeochemistry in an experimental agricultural watershed in Southeastern China. Biogeochemistry. 45: 1–19. Ye, C., Xu, Q. J., Kong, H. N., Shen, Z. M. and Yan, C. Z. 2007. Eutrophication conditions and ecological status in typical bays of Lake Taihu in China. Environ. Monit. Assess. 135: 217–225. Zhang, H. C., Cao, Z. H., Wang, G. P., Zhang, H. A. and Wong, M. H. 2003. Winter runoﬀ losses of phosphorus from paddy soils in the Taihu Lake Region of South China. Chemosphere. 52: 1461–1466. Zhang, W., Wang, X. J., Jiang, Y. C. and Zhou, X. 2001. Eﬀect of emission control on water quality of the Taihu Lake. Rural Eco-Environ. (in Chinese). 17(1): 44–47. Zhang, Z. J., Zhang, J. Y., He, R., Wang, Z. D. and Zhu, Y. M. 2007. Phosphorus interception in ﬂoodwater of paddy ﬁeld during the rice-growing season in TaiHu Lake Basin. Environ. Pollut. 145: 425–433. Zuo, Q., Lu, C. A. and Zhang, W. L. 2003. Preliminary study of phosphorus runoﬀ and drainage from a paddy ﬁeld in the Taihu Basin. Chemosphere. 50: 689–694.