The concentrations of eight heavy metals (Cr, Hg, As, Pb, Cd, Cu, Zn, Ni) in six river sediment samples were collected for evaluation of the degree of the heavy metals pollution distribution and ecological risk of three main rivers' sediments in Jinan. Multivariate statistical techniques were used to determine the most common pollution sources. The results illustrated that all of the metals in Damatou and Xinfengzhuang sections of the Xiaoqing River were much higher than the background value, and the level of potential ecological risk index was very high. The remaining four sections had a low or moderate degree of ecological risk. Principal component analysis (PCA) showed that all metals, with the exception of As, formed the first component explaining 86.85% of the total variance and industry sources could be considered as the first component, while As alone could be the second component, representing agricultural source. The elements Cr and Zn were grouped together while the remaining six metals formed a separate category. Among all heavy metals, Hg and Cd were the most significant contributors to the pollution. Therefore, the prevention of pollution should pay more attention to controlling the sources, especially Hg and Cd.

INTRODUCTION

As a consequence of industrialization and urbanization in advanced cities in China, a large amount of industrial wastewater and domestic sewage is discharged directly into rivers and other bodies of water with minimal treatment. River sediment pollution has attracted considerable attention over the past decades. Of all the issues that endanger ecosystems, heavy metals play a significant role (Fu et al. 2009; Zhang et al. 2012). With the characteristics of bioaccumulation, persistence, and environmental toxicity (Jiang et al. 2012; Bednarova et al. 2013), heavy metals can amalgamate different carriers to form hazardous pollutants to nature. Heavy metal contaminants enter the water system via water circulation and are adsorbed into suspended sediments and deposited on the riverbed (Yi et al. 2016). Sediments can store approximately 99% of the metals in water bodies in various forms under certain conditions (Salomons & Stigliani 1995). Moreover, they are capable of accumulating extremely low and undetectable concentrations of heavy metals. Therefore, with the higher heavy metal content in sediments than in water, the enrichment rate of heavy metals in river sediments is a generally preferred indicator of the rivers' contamination status (Soares et al. 1999). Instead of migration and biodegradation by spontaneous degeneration, heavy metals commonly undergo long-term accumulation and storage that cause them to harm aquatic ecosystems through a series of physical, chemical, and biological interaction (Tang et al. 2002). Bottom sediments create natural habitats and food sources for benthic fauna. In addition, pollutants can be directly or indirectly toxic to the aquatic flora and fauna (Lamin Daddy et al. 2015). When heavy metals in the water system accumulate to a high level, the sublethal effects of pollutants on local fish populations would increase, thereby resulting in a potential long-term implication on human health through food chains. (Almeida et al. 2002; Yujun et al. 2008; Yi & Zhang 2012).

Many previous studies focused mainly on the spatial analysis of a single river (Fu et al. 2009; Varol 2011), but only a few probe heavy metal pollution in river sediments across an entire city. Studies on river sediment pollution across a city-wide scale can provide a database for status analysis and take targeted measures for management integration and pollution control, as well as make rational use of water resources. This study focuses on the Yellow River, the Xiaoqing River, and the Tuhai River. Furthermore, this study aims to (a) analyze the spatial distribution patterns of heavy metals in sediments, (b) use Hakanson potential ecological risk analysis to evaluate the degree of the risk level, (c) identify possible pollution sources that cause contamination, and (d) provide a comprehensive description and scientific basis for the control and management of heavy metals, which could be beneficial for the protection of river sediments in Jinan.

MATERIALS AND METHODS

Study area

The study area is located in Jinan, one of the most developed cities in China and the capital of Shandong Province (situated at 36°01′N- 37°32′N, 116°11′E-117°44′E and covering 8,177.21 km2). The Yellow River, the Xiaoqing River and the Tuhai River are the main rivers flowing through Jinan. The Yellow River is one of the longest rivers in China with a catchment area of about 2,778 km2 and a length of 185.3 km within Jinan, providing both drinking water and domestic water. The basin area of the Xiaoqing River is 2,803 km2, and its length in Jinan is 70.3 km. Recently, the increasing development of industries along this river result in wastewater discharge. Industrial wastewater discharge mainly affects the urban areas of Jinan city and Zhangqiu city. Being the river that substantially acquires pollutants, the Xiaoqing River accounts for 68.8% of the total industrial wastewater in Jinan (Jinan Environmental Quality Reports 2014). The Tuhai River is located in the northern portion of Jinan, with a catchment area of approximately 1,452 km2 and a length within the city of 85.5 km.

The industrial enterprises of Jinan are relatively distributed centrally along the rivers, with the Yellow River flowing through an energy manufacturer and a food factory in the middle reaches. Several factories are also established along the Xiaoqing River banks. No industrial factory in the river source Mulizhuang is constructed. A pharmaceutical and a power plant are located in the upper sections, and an industrial complex named Wangsheren, including a steel manufacturer, a petroleum factory, a chemical fertilizer plant, and a refinery, is situated in the middle reaches – the area where Damatou is also located. A fertilizer also plant lies in the lower reaches of the Xiaoqing River, whereas two pharmaceutical plants and a dyeing plant are located in the middle section of the Tuhai River.

The sampling sites and data collection

The study collected the concentrations of eight types of heavy metals from six monitoring sections of three rivers: Luokou from the Yellow River, Mulizhuang, Damatou, and Xinfengzhuang from the Xiaoqing River, and Xiakouqiao and Shenqiao from the Tuhai River (Figure 1). Pb and Cd concentrations were determined by graphite furnace atomic absorption spectrophotometry (GFAAS). The samples were microwave digested with HCl-HNO3 to determine the content of Hg and As concentrations by atomic fluorescence spectrum (AFS). Cu, Ni and Cr concentrations were determined by flame atomic absorption spectrophotometry (FAAS). All data were collected from Jinan environment quality report (2014), the descriptive statistics were computed with Microsoft Excel and IBM SPSS Statistics 19.0. The specific distribution of the monitoring sites is shown in Figure 1 and the longitude and latitude of the sampling sites are shown in Table 1.
Table 1

The coordinates of the sampling sites in the three rivers

RiverPositionControl levelLongitudeLatitude
The Yellow River Luokou National control 116°59′20″ 36°44′00″ 
The Xiaoqing River Mulizhuang Provincial control 116°49′57″ 36°40′08″ 
Damatou City control 117°10′34″ 36°47′17″ 
Xinfengzhuang Provincial control 117°23′25″ 36°56′37″ 
The Tuhai River Xiakouqiao Provincial control 116°56′08″ 36°01′38″ 
Shenqiao Provincial control 117°24′40″ 37°12′06″ 
RiverPositionControl levelLongitudeLatitude
The Yellow River Luokou National control 116°59′20″ 36°44′00″ 
The Xiaoqing River Mulizhuang Provincial control 116°49′57″ 36°40′08″ 
Damatou City control 117°10′34″ 36°47′17″ 
Xinfengzhuang Provincial control 117°23′25″ 36°56′37″ 
The Tuhai River Xiakouqiao Provincial control 116°56′08″ 36°01′38″ 
Shenqiao Provincial control 117°24′40″ 37°12′06″ 
Figure 1

The distribution of the sampling sites in sediments of Jinan.

Figure 1

The distribution of the sampling sites in sediments of Jinan.

Assessment methods

Potential ecological risk index

Many other methods, such as the enrichment factor, are used to reflect the effects of human activities on the enrichment of a single heavy metal, with the limits due to minimal consideration on the bioavailability and the combined effects of heavy metals. This paper used the Hakanson potential ecological risk index (PERI) (Hakanson 1980) to quantitatively analyze the degree of potential ecological risk for aquatic pollution control, which was based on the hypothesis that the sensitivity of the aquatic system depends on its productivity (Yi et al. 2016) and is generally widely used in such research. The degrees of risk level are shown in Table 2.

Table 2

The degrees of contamination potential ecological risk

The range of The degree of riskThe range of RIThe degree of risk
< 30 Low RI < 110 Low 
30 ≤ < 60 Moderate 110 ≤ RI < 220 Moderate 
60 ≤ < 120 Considerable 220 ≤ RI < 440 High 
120 ≤ < 240 High RI ≥ 440 Very high 
≥ 240 Very high   
The range of The degree of riskThe range of RIThe degree of risk
< 30 Low RI < 110 Low 
30 ≤ < 60 Moderate 110 ≤ RI < 220 Moderate 
60 ≤ < 120 Considerable 220 ≤ RI < 440 High 
120 ≤ < 240 High RI ≥ 440 Very high 
≥ 240 Very high   

The computational formula is as follows:
formula
  • RI – the potential ecological risk index;

  • CF – the pollution parameter of heavy metal i;

  • – the toxic index of heavy metal i (Zn:1, Cr:2, Cu:5, Ni:5, Pb:5, As:10, Cd:30) (Hakanson 1980);

  • – the potential ecological risk parameter of heavy metal i;

  • – the determined content of heavy metal i;

  • – the background value of heavy metal i.

Multivariate analysis methods

Pearson correlation analysis was performed to identify the relationships among the concentrations of the heavy metals (Varol 2011). The paper chose principal component analysis (PCA) for explaining relationships and associations between objects and variables (Wold et al. 1987). Moreover, PCA is widely applied in data processing and reduction of the original dimensionality and an exploratory data analysis tool to complete the linear combination of original variables as well as to derive the maximal variance (Zou et al. 2006). Hierarchical cluster analysis (HCA) is always paired with PCA to confirm the results and identify the relationships among the analyzed parameters and their possible sources (Varol & Şen 2009, 2012). Microsoft Excel 2010, IBM SPSS Statistics 19.0, Photoshop 2015 and Origin Pro 9.1 were used for data processing, statistical analysis, and drawing, respectively.

RESULTS

Heavy metal concentration and spatial distribution

Background values play a significant part in evaluating the degree of heavy metal pollution. The heavy metals in soil can be washed down to streams and rivers and accumulate in riverbed sediments (Yi et al. 2016). Comparing the concentrations with the soil background values is the most widely used and approved method in the world and is commonly used in such research. Including specific background values associated with the assessed region to have a reference for testing is generally recommended (Chapman et al. 1999). The background values of soil heavy metals in Shandong Province were selected as the background values in surface sediments (Wang et al. 2012). Table 3 shows the contents of Cd, Hg, As, Pb, Cr, Cu, Zn and Ni in river sediments in the six sections. In the Luokou section of the Yellow River, Cd content was slightly higher than the background value, whereas the concentrations of the other seven metals were considerably low. In Damatou, the concentrations of all the aforementioned metals exceeded the limit of the background values and reached the maximum, except for As. Even the Hg concentration was approximately 47 times that of the background value (0.906 mg kg−1). The concentration of all the studied metals showed a clear decrease from the Damatou to the Xinfengzhuang areas, particularly for Hg, which decreased by 22%. As was the only metal that increased at Xinfengzhuang, reaching a maximum of 15.6 mg kg−1. However, all the metal concentrations in this area also exceeded the background values. In the Tuhai River, the metal content was close to background values, whereas that of Hg, As and Zn was significantly high in Xiaokouqiao. Ni was present in the amounts equal to the value, and the content of Cr, Pb, Cd, and Cu was substantially low. At Shenqiao, Hg, As, Cd, Zn and Ni contents exceeded the value. On the other hand, Cu content was equal to the background value, and Cr and Pb contents were considerably low. Generally, almost all elements showed higher levels in the Xiaoqing River than in the Yellow River and the Tuhai River. Figure 2 provides a visual representation of these data. The dotted lines in the figure below represent the background values in Jinan.
Table 3

Heavy metal concentrations in the sediment from six sections (mg kg−1)

 SectionsCrHgAsPbCdCuZnNi
The Yellow River Luokou 43 0.017 5.99 11.9 0.0952 11 46.3 16.9 
The Xiaoqing River Mulizhuang 62.1 0.037 3.68 8.7 0.0628 13.4 48.5 24.8 
Damatou 334 0.906 12.7 58.2 0.683 110 514 48.2 
Xinfengzhuang 110 0.199 15.6 30.6 0.112 38.6 128 38.3 
The Tuhai River Xiakouqiao 55.5 0.036 12.1 18.3 0.0697 23 70 25.8 
Shenqiao 59.8 0.042 11.6 17.8 0.13 24 73.8 31.2 
Background value  66 0.019 9.3 25.8 0.084 24 63.5 25.8 
SDa  103.4 0.33 4.123 16.66 0.224 34.318 169.37 10.29 
CVb  0.99 1.82 0.407 0.68 1.27 0.984 1.256 0.341 
 SectionsCrHgAsPbCdCuZnNi
The Yellow River Luokou 43 0.017 5.99 11.9 0.0952 11 46.3 16.9 
The Xiaoqing River Mulizhuang 62.1 0.037 3.68 8.7 0.0628 13.4 48.5 24.8 
Damatou 334 0.906 12.7 58.2 0.683 110 514 48.2 
Xinfengzhuang 110 0.199 15.6 30.6 0.112 38.6 128 38.3 
The Tuhai River Xiakouqiao 55.5 0.036 12.1 18.3 0.0697 23 70 25.8 
Shenqiao 59.8 0.042 11.6 17.8 0.13 24 73.8 31.2 
Background value  66 0.019 9.3 25.8 0.084 24 63.5 25.8 
SDa  103.4 0.33 4.123 16.66 0.224 34.318 169.37 10.29 
CVb  0.99 1.82 0.407 0.68 1.27 0.984 1.256 0.341 

aSD, standard deviation.

bCV, coefficient of variation.

Figure 2

Concentration of heavy metals in the sediment of three rivers in Jinan.

Figure 2

Concentration of heavy metals in the sediment of three rivers in Jinan.

Dividing the standard deviation by the expected value resulted in a dimensionless variance coefficient (Weber et al. 2004). The coefficient of variation (CV) reflects the average variation among sampling points. Among the six sampling sites, Cr content was 0.65–5.06 times higher than the background value, Hg content was 0.89–47.68, As was 0.40–1.67, Pb was 0.34–2.26, Cd was 0.75–8.13, Cu was 0.46–4.58, Zn was 0.73–8.09 and Ni was 0.66–1.87 times more than the background values. Hg, Cd, and Zn clearly had the greatest variation with a variance coefficient in excess of 1. The strong variation confirmed that the distribution of these three metals in Jinan river sediments was particularly uneven. Additionally, the concentrations of these metals were higher in Damatou than in other sites with about 47.68, 8.13, 8.09 times greater than the background values, respectively. The high concentrations also appeared in Xinfengzhuang with the contents at 10.47, 1.33, and 2.02 times greater than the background value, respectively. Hg content was lower in four other sites with the range of concentration from 0.017 to 0.042 mg kg−1. The strongest variability was shown in the distribution of Hg, suggesting that Hg was particularly distributed unevenly. With the exception of As and Ni, the degrees of dispersion of the other six heavy metals were relatively high, with a variance coefficient of more than 0.5. The fluctuation of the spatial variability of As and Ni was relatively low, indicating a consistent distribution of these metals.

Potential ecological risk assessment of heavy metals

Reference value is one of the key parameters used to calculate the potential ecological risk factor. This paper used the soil background value of Shandong province as the reference value (Wang et al. 2012). The calculated results (Table 4) corroborated that the risk index (RI) values for the six sites ranged approximately from 86 to 2,226, with those in Damatou and Xinfengzhuang having the highest RI values (2,226.7 and 502.4, respectively) and reaching a very high risk level. Mulizhuang in the Xiaoqing River and the two sites in the Tuhai River showed moderate ranges, whereas Luokou displayed a low degree of risk level. Hg showed the highest contribution to RI values in all the sites, with a particularly high value of up to 1,907 at Damatou, followed by Cd. Figure 3 shows the contribution of different heavy metals to potential ecological risk in river sediments.
Table 4

Values of ecological risk factor and RI of heavy metals in sediments

 MetalsLuokouMulizhuangDamatouXinfengzhuangXiakouqiaoShenqiao
Ei Cr 1.303 1.882 10.121 3.333 1.682 1.812 
Hg 35.789 77.895 1,907.368 418.947 75.789 88.421 
As 6.441 3.957 13.656 16.774 13.011 12.473 
Pb 2.306 1.686 11.279 5.930 3.547 3.45 
Cd 34 22.429 243.929 40 24.893 46.429 
Cu 2.292 2.792 22.917 8.042 4.792 
Zn 0.729 0.764 8.094 2.016 1.102 1.162 
 Ni 3.275 4.806 9.341 7.422 5.000 6.047 
RI  86.136 116.210 2,226.705 502.465 129.815 164.793 
Degree  Low Moderate Very high Very high Moderate Moderate 
 MetalsLuokouMulizhuangDamatouXinfengzhuangXiakouqiaoShenqiao
Ei Cr 1.303 1.882 10.121 3.333 1.682 1.812 
Hg 35.789 77.895 1,907.368 418.947 75.789 88.421 
As 6.441 3.957 13.656 16.774 13.011 12.473 
Pb 2.306 1.686 11.279 5.930 3.547 3.45 
Cd 34 22.429 243.929 40 24.893 46.429 
Cu 2.292 2.792 22.917 8.042 4.792 
Zn 0.729 0.764 8.094 2.016 1.102 1.162 
 Ni 3.275 4.806 9.341 7.422 5.000 6.047 
RI  86.136 116.210 2,226.705 502.465 129.815 164.793 
Degree  Low Moderate Very high Very high Moderate Moderate 
Figure 3

Contribution of different heavy metals to potential ecological risk in river sediments.

Figure 3

Contribution of different heavy metals to potential ecological risk in river sediments.

Multivariate statistical analysis

Correlation analysis

Table 5 shows the Pearson correlation coefficient of heavy metals in river sediments. Remarkably, no significant correlation was present between As and the other metals in either the 0.05 probability level or the 0.01 significance level. Among the remaining heavy metals, the correlations between each pair were significant, apart from Cd and Ni. Cu, Hg, Cr, Zn, and Pb had the strongest correlation, with the coefficient between each pair being close to 1.

Table 5

The pearson correlation coefficient of heavy metals in river sediments

Heavy metalsCrHgAsPbCdCuZnNi
Cr 0.999** 0.390 0.964** 0.980** 0.992** 0.997** 0.867* 
Hg  0.389 0.965** 0.983** 0.992** 0.998** 0.856* 
As   0.615 0.317 0.489 0.398 0.693 
Pb    0.929** 0.985** 0.965** 0.921** 
Cd     0.973** 0.990** 0.795 
Cu      0.994** 0.897* 
Zn       0.851* 
Ni        
Heavy metalsCrHgAsPbCdCuZnNi
Cr 0.999** 0.390 0.964** 0.980** 0.992** 0.997** 0.867* 
Hg  0.389 0.965** 0.983** 0.992** 0.998** 0.856* 
As   0.615 0.317 0.489 0.398 0.693 
Pb    0.929** 0.985** 0.965** 0.921** 
Cd     0.973** 0.990** 0.795 
Cu      0.994** 0.897* 
Zn       0.851* 
Ni        

*p < 0.05; **p < 0.01, levels of significance.

Principal component analysis

Table 6 shows the eigenvalues and loadings of components. The first two components with a cumulative variance contribution rate of 98.47% were extracted as principal component 1 (PC1) and principal component 2 (PC2). PC1 explained 86.86% of the total variance, and Cu, Pb, Cr, Ni, Zn, Hg, and Cd showed the maximum loading. PC2 accounted for 11.62% of variation in the data and was characterized by the high loading of As. The principal component loading matrix can be viewed clearly in Figure 4.
Table 6

Total variance explained for the heavy metals in the sediments and the component matrix

ComponentInitial eigenvalues
Extraction sums of squared loadings
Total% of varianceCumulative %Total% of varianceCumulative %
6.949 86.859 86.859 6.949 86.859 86.859 
0.929 11.615 98.474 0.929 11.615 98.474 
0.102 1.272 99.746 0.102 1.272 99.746 
0.018 0.223 99.969 0.018 0.223 99.969 
0.003 0.031 100 0.003 0.031 100 
Component matrix
Principal component loading matrix
MetalsPC1PC2MetalsPC1PC2
Cr 0.985 −0.164  Cr 0.374 −0.170 
Hg 0.984 −0.170  Hg 0.373 −0.176 
As 0.538 0.835  As 0.204 0.866 
Pb 0.992 0.089  Pb 0.376 0.092 
Cd 0.959 −0.251  Cd 0.364 −0.260 
Cu 0.998 −0.060  Cu 0.379 −0.062 
Zn 0.985 −0.164  Zn 0.374 −0.170 
Ni 0.921 0.274  Ni 0.349 0.284 
ComponentInitial eigenvalues
Extraction sums of squared loadings
Total% of varianceCumulative %Total% of varianceCumulative %
6.949 86.859 86.859 6.949 86.859 86.859 
0.929 11.615 98.474 0.929 11.615 98.474 
0.102 1.272 99.746 0.102 1.272 99.746 
0.018 0.223 99.969 0.018 0.223 99.969 
0.003 0.031 100 0.003 0.031 100 
Component matrix
Principal component loading matrix
MetalsPC1PC2MetalsPC1PC2
Cr 0.985 −0.164  Cr 0.374 −0.170 
Hg 0.984 −0.170  Hg 0.373 −0.176 
As 0.538 0.835  As 0.204 0.866 
Pb 0.992 0.089  Pb 0.376 0.092 
Cd 0.959 −0.251  Cd 0.364 −0.260 
Cu 0.998 −0.060  Cu 0.379 −0.062 
Zn 0.985 −0.164  Zn 0.374 −0.170 
Ni 0.921 0.274  Ni 0.349 0.284 
Figure 4

The loading plots for principal component of PCA analysis.

Figure 4

The loading plots for principal component of PCA analysis.

Hierarchical cluster analysis

The result of HCA is shown in the dendrogram in Figure 5. The elements connected in the first section indicated the origins from a similar source, and the length represented the degree of difference among the elements in the source (Zhang et al. 2013). Therefore, a short distance between the two elements indicated a significantly high probability of origins from the same source. The image suggested that the eight heavy metals can be divided into two categories when the distance is less than 10. The distance among Hg, Cd, As, Pb, Ni and Cu was the shortest; hence, they were in a single category, indicating the deep homology between them. Cr and Zn belonged to another cluster, and the result agreed with the correlation analysis, indicating a significant correlation between Cr and Zn.
Figure 5

Dendrogram from HCA of heavy metals in river sediments.

Figure 5

Dendrogram from HCA of heavy metals in river sediments.

DISCUSSION

Potential ecological risk in sediments

The river sediments in Jinan, especially in the Xiaoqing River, were found to be in serious condition according to the status of heavy metal contamination. The degree of risk of the eight metals in the three rivers, respectively, followed the order Hg > Cd > As > Ni > Pb > Cu > Cr > Zn in the Yellow River, Hg > Cd > As > Cu > Ni > Pb > Cr > Zn in the Xiaoqing River, and Hg > Cd > As > Ni > Cu > Pb > Cr > Zn in the Tuhai River. Overall, among the six heavy metals, Hg, Cd, and As made the greatest contributions to the potential RI, especially Hg, which had its RI contribution of 84.5% in the Xiaoqing River and 55.74% in the Tuhai River. Cd made the greatest contribution in the Yellow River, at 39.47%. Although Hg and Cd were both present in low levels, taking the background values, which were also of a low level, into consideration was crucial. The analysis of the measured concentrations and background values is necessary for the accurate judgment on pollution status and is helpful to highlight the supported data that Hg and Cd are the most important objects of heavy metal pollution prevention and control in Jinan.

Source analysis of heavy metals in sediments

Through a combination of correlation analysis, principal component analysis, and cluster analysis, the sources of the heavy metals can be determined (Zhao et al. 2015). The concentrations of most heavy metals increased progressively from Mulizhuang to Damatou, with contents of Hg at 24.5, Cd at 10.9, Zn at 10.9, Cu at 8.2, Pb at 6.7, and Cr at 5.4, which are several times more in Damatou than contents in Mulizhuang. This pattern was connected with the upstream factories and enterprises around Damatou, including a steel manufacturer, a petroleum factory, a chemical fertilizer plant, and an oil refinery. The steel smelting process tends to be associated with heavy metal pollutants, such as Hg, Zn, Cd Cu, Pb and Cr (Meng et al. 2016). The results of the correlation analysis verified that the significant correlations found among these six heavy metals also supported the likelihood of the case (Table 5). Some of the contributions to the Hg concentrations may also come as a result of a power plant in the upper reaches of Damatou. Hg and its compounds were emitted into the atmosphere because of the coal combustion in generating electricity (Yokoyama et al. 2000). Some studies also investigated the extent of Hg contamination due to Zn smelting process (Li et al. 2008), which caused the accumulation of Hg by atmospheric deposition into the river. Cd and Cu may have originated from the sewage of the fertilizer plants. Hg, Cd, As, Pb and Ni could have come from the combustion of oil and coal; hence, they were grouped together (Figure 5) with the oil refinery plant near the Xiaoqing River. Cr and Zn were far from the other six heavy metals in the dendrogram, indicating that the source of Cr is similar to Zn, with the correlation coefficient between Cr and Zn even reaching 0.997 at the 0.01 level of significance. Only As content increased from Damatou to Xinfengzhuang, which was suspected to have originated from pesticides and agricultural wastewater (Murphy & Aucott 1998). However, in Xinfengzhuang located in Zhangqiu, agriculture is the main economy of the area and accounts for 30.3% of the total industrial wastewater emissions (Jinan Environmental Quality Reports 2014), which reached agreement with As content increasing instead of decreasing. With PCA, the agricultural factor can be identified as the second principal component (PC2), whereas PC1 can be considered to represent the industrial sources due to a high capacity for receiving industrial pollutants, which was consistent with the correlation analysis result, thereby showing no significant correlation between As and the other seven heavy metals.

Measures and recommendations

Generally, the Xiaoqing River was seriously polluted, with a very high degree of ecological risk, whereas the Yellow River and the Tuhai River were slightly polluted. The most important purpose of this paper is to advise on the measures to manage the studied regions and control heavy metal pollution. These measures should be taken to achieve this practical significance, particularly the pollutant control in the Xiaoqing River, Hg and Cd resulted in a high ecological risk in sediments although their concentrations remained at a low level, along with the background values which were also low. Previous studies confirmed that the combustion of fuel by power generation and heating was the largest source of anthropogenic mercury emissions (Pacyna et al. 2010). Hg can easily be evaporated at room temperature as a silver white liquid, emitted into the atmosphere, and then entered the river via atmospheric deposition. Of all the Hg contents emitted from coal combustion in China, 51% were from industrial activities, 33.6% from power plants, 9.8% from residential use, and 5.6% from other uses (Streets et al. 2005). The industrial enterprises, such as chemical industries, non-ferrous metal mining-selection and metallurgy, chemical manufacturing industry, and the dry battery processing industry, may cause the Hg emission. Cd is a kind of toxic metal that is widely used in various industries such as smelting, electroplating, dyeing, chemical plant and battery manufacturing. Adjustments should be made to industrial structure, the smoke control function equipment should be improved, and heavy metal pollution control should be strengthened. Rendering heavily polluting technologies obsolete and upgrading lagging equipment to reach discharge or emission standards. The government should continue combining the consistency control and total control of pollution emissions, closing down factories that exceed the discharge standards, and employing considerably effective methods to regulate the pollution. The emission limits for Hg should be strictly performed, and the Hg removal technologies by adsorption should be modified especially in power plants. Meanwhile, strengthening the prevention and control of agricultural pollution and reducing the application of chemical fertilizers and pesticides can ensure the health of the water system.

CONCLUSIONS

The study aimed to determine the pollution status and sources of heavy metals by multivariate statistical and used the PERI to analyze the risk level at six monitoring sections in Jinan. Substantially practical and valuable suggestions on the control and management of heavy metal pollution were proposed. The results corroborated that Damatou and Xinfengzhuang had a very high ecological risk status and that the contents of Hg and Cd made significant contributions to the RI. The sources of the heavy metals were primarily contributed by anthropogenic industrial activity inputs, such as the steel work, petroleum factory, and power plant in the upper reaches of Damatou. Agricultural impacts, such as the use of pesticides, accounted for a small portion of the concentrations.

The overall situation of the environment assessed by river sediments has significantly improved through the implemented government measures during past decades, but heavy metal pollution status still remains a concern. This study suggested that high-efficiency measures to control the pollution and reduce the ecological risk in river sediments should be developed. The government should strictly execute standards and seriously take the task of dealing with pollution-contributing industries to meet the green development requirement in Jinan, particularly the Xiaoqing River.

ACKNOWLEDGEMENTS

We thank the anonymous reviewers for their critical review of the manuscript. We also appreciate the person and institution, which provided accurate and reliable data for original criteria. We also thank the professional institution (www.shineWrite.com), and Katherine Olson, a native English speaker who is studying in Virginia Polytechnic Institute and State University, to help us correct grammatical errors. This work was supported by China National Natural Science Foundation (no. 41301649).

REFERENCES

Almeida
J. A.
Diniz
Y. S.
Marques
S. F.
Faine
L. A.
Ribas
B. O.
Burneiko
R. C.
Novelli
E. L.
2002
The use of the oxidative stress responses as biomarkers in Nile tilapia (Oreochromis niloticus) exposed to in vivo cadmium contamination
.
Environment International
27
(
8
),
673
679
.
Bednarova
Z.
Kuta
J.
Kohut
L.
Machat
J.
Klanova
J.
Holoubek
I.
Jarkovsky
J.
Dusek
L.
Hilscherova
K.
2013
Spatial patterns and temporal changes of heavy metal distributions in river sediments in a region with multiple pollution sources
.
Journal of Soils and Sediments
13
(
7
),
1257
1269
.
Chapman
P. M.
Wang
F.
Adams
W. J.
Green
A.
1999
Appropriate applications of sediment quality values for metals and metalloids
.
Environmental Science & Technology
33
(
22
),
3937
3941
.
Jiang
X.
Wang
W.
Wang
S.
Zhang
B.
Hu
J.
2012
Initial identification of heavy metals contamination in Taihu Lake, a eutrophic lake in China
.
Journal of Environmental Sciences
24
(
9
),
1539
1548
.
Jinan Environmental Quality Reports
2014
The Environmental Protection Bureau of Jinan (in Chinese)
.
Lamin Daddy
M.
Hui
M.
Xue Hui
L.
Peng Yu
H.
Shu-Mei
Z.
Zhong-Xian
H.
Dian-Wu
L.
2015
Heavy metal accumulation in soils, plants, and hair samples: an assessment of heavy metal exposure risks from the consumption of vegetables grown on soils previously irrigated with wastewater
.
Environmental Science and Pollution Research
22
(
23
),
18456
18468
.
Li
G.
Feng
X.
Qiu
G.
Bi
X.
Li
Z.
Zhang
C.
Wang
D.
Shang
L.
Guo
Y.
2008
Environmental mercury contamination of an artisanal zinc smelting area in Weining County, Guizhou, China
.
Environmental Pollution
154
(
1
),
21
31
.
Meng
G.
Xuyun
C.
Tianfang
Z.
Dihan
L.
Yushang
L.
Lingyun
Y.
2016
Research on electro-coagulation for the treatment of heavy metal from smelting wastewater
.
World Nonferrous Metals
20
,
14
17
.
Murphy
E. A.
Aucott
M.
1998
An assessment of the amounts of arsenical pesticides used historically in a geographical area
.
Science of the Total Environment
218
(
218
),
89
101
.
Pacyna
E. G.
Pacyna
J. M.
Sundseth
K.
Munthe
J.
Kindbom
K.
Wilson
S.
Steenhuisen
F.
Maxson
P.
2010
Global emission of mercury to the atmosphere from anthropogenic sources in 2005 and projections to 2020
.
Atmospheric Environment
44
(
20
),
2487
2499
.
Salomons
W.
Stigliani
W. M.
1995
Biogeodynamics of Pollutants in Soils and Sediments
.
Springer-Verlag
,
New York, NY, USA
,
pp. 331–343
.
Soares
H. M.
Boaventura
R. A.
Machado
A. A.
Esteves da Silva
J. C. G.
1999
Sediments as monitors of heavy metal contamination in the Ave river basin (Portugal): multivariate analysis of data
.
Environmental Pollution
105
(
3
),
311
323
.
Streets
D. G.
Hao
J.
Wu
Y.
Jiang
J.
Chan
M.
Tian
H.
Feng
X.
2005
Anthropogenic mercury emissions in China
.
Atmospheric Environment
39
(
40
),
7789
7806
.
Wang
C. L.
Xia
X. Q.
Zhao
X. Q.
Zheng
W. J.
Zhou
G. H.
Hu
X. P.
Ji
S. L.
2012
Distribution and migration regularity of soil heavy metal pollution along the Xiaoqing watershed, Shandong province
.
Geology in China
39
(
2
),
530
538
.
Wold
S.
Esbensen
K.
Geladi
P.
1987
Principal component analysis
.
Chemometrics & Intelligent Laboratory Systems
2
(
1–3
),
37
52
.
Yi
Y. J.
Sun
J.
Tang
C. H.
2016
Ecological risk assessment of heavy metals in sediment in the upper reach of the Yangtze River
.
Environmental Science and Pollution Research
23
(
11
),
11002
11013
.
Yokoyama
T.
Asakura
K.
Matsuda
H.
Ito
S.
Noda
N.
2000
Mercury emissions from a coal-fired power plant in Japan
.
Science of the Total Environment
259
(
1–3
),
97
103
.
Yujun
Y. I.
Wang
Z.
Zhang
K.
Guoan
Y. U.
Duan
X.
2008
Sediment pollution and its effect on fish through food chain in the Yangtze River
.
International Journal of Sediment Research
23
(
4
),
338
347
.
Zhang
Q. H.
Lin
C. H.
Tan
H.
Lin
S. X.
Yang
H. B.
2013
Accumulation, distribution and pollution assessment of heavy metals in surface sediment of Caohai Plateau Wetland, Guizhou Province
.
Environmental Science
34
(
34
),
1055
1061
.
Zou
H.
Hastie
T.
Tibshirani
R.
2006
Sparse Principal Component Analysis
.
Journal of Computational and Graphical Statistics
15
(
2
),
265
286
.