Introduction

Sediment is an ecologically important part of aquatic habitats and is key to the dynamics of water quality (Yi et al. 2011). Metal ions (e.g., Zn2+, Cd2+) are easily adsorbed on the surface of suspended particles after entering a water body and ultimately accumulate in the sediment (Deng et al. 2016; Xu et al. 2018). When the environmental conditions (pH, oxidation-reduction potential, etc.) at the sediment-water interface change, the metals in the sediment are released into the overlying water (Chow et al. 2005; Gupta et al. 2015), especially in shallow estuaries and lakes (ten Hulscher et al. 1992; Superville et al. 2014). Huang et al. (2012) found that sediment resuspension led to increases in the concentrations of the Ni, Cu, and Pb in water to 13.2 μg/L, 28.8 μg/L, and 20.5 μg/L, respectively. An increase in the metallic ion concentrations in water causes enrichment of these metals in aquatic organisms. Salem and Ayadi (2016) demonstrated, for example, that the livers, gills, and muscle tissues of fish exhibit different enrichment capacities for Cd, Cu, Ni, and Zn; Pal and Maiti (2018) found that the accumulative rate of metal ions in benthic-feeding fish is significantly higher than that in nonbenthic-feeding species, indicating that the metallic ion concentrations in various tissues of fish are different but are all significantly higher than those in ambient water (Velusamy et al. 2014).

It is essential to assess the risk of metallic contamination in sediments for water pollution prevention and control (Pravdić 1995; Sharifi et al. 2016). In recent decades, certain methods for evaluating metallic pollution risks have been proposed (Birch 2018), and the geoaccumulation index (Igeo), pollution load index (PLI), sediment quality guidelines (SQGs), and potential ecological risk index (RI) have been widely used. Loska et al. (1997) and Zhang et al. (2014) found that Pb and Cd are the primary pollutants in sediments of reservoirs and lakes according to Igeo. Suresh et al. (2012) and Islam et al. (2015) considered that metallic ions in sediments interact and that Cr, Cd, and Pb are the primary contributors to sediment pollution based on the PLI threshold. Aguiar et al. (2016) noted that Pb, Cu, Zn, and Cr in sediments are potential ecological risks for benthic organisms, especially commercial fish, according to the RI index. Although the results obtained by the different methods mentioned above are inconsistent, the results nevertheless reflect pollution from metals and ecological risk in sediments combined with the sediment properties and regional background values of metallic ions (Deng et al. 2016).

Determining the sources of metallic ions in sediments is important for understanding the environmental behaviors of the metals and reducing human exposure to metallic pollution (Gonzalez-Macias et al. 2014) and is a basic requirement of policy development in pollution control, treatment, and environmental management (Hepburn et al. 2018). Source apportionment of metallic ions in environmental media can be addressed by different means, including factor analysis, principal component analysis, and correlation analysis (Thuong et al. 2013; Xia et al. 2018). Brady et al. (2014) considered Ni and V to be identifying elements for ship transportation and petroleum combustion and Cu, Te, and Hg to be identifying elements for antifouling coatings. Dou et al. (2013) believed that total organic carbon (TOC), Hg, Cd, and As are correlated with agricultural activity sources. Principal component analysis and correlation analysis can be used to determine certain elements and compounds associated with pollution sources. For example, the combination of Te, Cu, Sn, V, and Ni is related to anthropogenic activities in ports, Pb and Sb are indicative of coal combustion, the combination of Cu, Pb, and Zn is representative of traffic emissions, Cu is associated with aquaculture activities, and As and Cd are generally related to the use of fertilizers and pesticides in agricultural activities (Boyle et al. 1998; Zhu et al. 2016).

Lake Taihu Basin (LTB), a typical region with high population density in China, is experiencing rapid urbanization, which plays a pivotal role in the economic development of China, and sediment pollution has become increasingly severe. For instance, Niu et al. (2015) investigated the potential ecological risk of Cu, Zn, Pb, Ni, and Cr in surface sediments from the Wujingang River and Caoqiao River in the LTB of China and found that metallic pollution is quite serious in the two rivers. Bian et al. (2016) demonstrated that the mean concentrations of metals in riverine surface sediments from the LTB are moderate ecological risks and that As and Cu are primary factors adversely influencing the benthic community. These studies preliminarily clarified the extent of pollution and ecological risks from metallic ions in the surface sediments of certain rivers. However, systematic research on metallic pollution in the riverine surface sediments of the entire LTB is still lacking. Therefore, the surface sediment samples from six water systems in the LTB were collected, and the metallic ion contents were measured to clarify the spatial distributions of the metallic contents in the riverine surface sediments. The metallic pollution levels in the riverine surface sediments were evaluated with the Igeo, PLI, SQGs, and RI, and the metallic sources were identified by statistical analysis.

Materials and methods

Study area

The LTB (30° 7′ 19″–32° 14′ 56″ N, 119° 3′ 1″–121° 54′ 26″ E) is located in a subtropical monsoon climate zone and has a total drainage area of approximately 36,800 km2. The rivers in the LTB, with a cumulative length of more than 120,000 km, can be divided into the Taoge water system (TG), Nanhe water system (NH), Tiaoxi water system (TX), Yanjiang water system (YJ), Huangpu water system (HP), and Hangzhou Bay-Yangtze estuary water system (HY) (Fig. 1). TX, NH, and TG are the primary inflow water systems of Lake Taihu, accounting for approximately 20%, 25%, and 50% of the total lake inflow, respectively (Wu et al. 2018). The TX and NH water systems both originate from the western hilly area of the basin, and the TG water system flows through the cities of Zhenjiang, Changzhou, and Wuxi. YJ, HP, and HY are the outflow water systems of Lake Taihu, primarily flowing through urban regions.

Fig. 1
figure 1

Schematic diagram of 94 sampling sites in 6 water systems of Lake Taihu Basin, China. YJ, Yanjiang river system; TX, Tiaoxi river system; TG, Tiaoge river system; NH, Nanhe river system; HP, Huangpu river system; HY, Hangzhou Bay and the Yangtze Estuary river system

Full size image

Sample collection, laboratory analysis, and quality control

According to riverine hydrological changes, sediment distributions, and metallic pollution sources, 94 sites were sampled in the 6 water systems of the LTB: 10 sites at TG, 10 sites at NH, 16 sites at TX, 20 sites at YJ, 25 sites at HP, and 13 sites at HY (Fig. 1). Riverine surface sediment was collected at each sampling site with a Petersen grab sampler in September 2014, placed in clean self-sealing polyethylene bags with marked labels, and immediately transported to the laboratory for further analysis. The samples from riverine surface sediments in the LTB were freeze-dried, ground with an agate grinder, passed through a 100-mesh nylon sieve, and stored at 4 °C prior to analysis.

The total nitrogen (TN) and total phosphorus (TP) concentrations were simultaneously measured by the potassium persulfate digestion method, and the organic matter (OM) was measured by the K2Cr2O7–H2SO4 wet oxidation method (Jin and Tu 1990). The Zn, Cr, Ni, Cu, Pb, As, Cd, Fe, and Mn concentrations were measured with an inductively coupled plasma mass spectrometry system (ICP-MS, Agilent 7700cx, Santa Clara, USA), and the Hg concentrations were measured with atomic fluorescence spectrometry (AFS, AFS-3100, Beijing Kechuang Haiguang Instrument Co., Beijing, China). To ensure the accuracy and precision of the measurement results, repeat analyses (repeatability, 10%) and standard sample analyses (recovery rate, between 85.0 and 115.0%) were performed on the samples. The analytical error for all the samples was less than 10%, which was consistent with the quality control requirements. The above analysis results were based on the dry weight of the sediment.

Assessment methods for sediment pollution

Igeo introduces factors that can cause changes in background values, such as anthropogenic factors, geochemical background values, and natural diagenesis, and quantitatively evaluates the pollution levels of metallic ions in sediments (Loska et al. 1997). The equation for calculating Igeo is as follows:

$$ {I}_{\mathrm{geo}}={\log}_2\left[{C}_i/\left(k\times {B}_i\right)\right] $$
(1)

where Ci is the content of element i in the sediment, Bi is the geochemical background value of element i, and k is the coefficient used in consideration of the diagenesis that may cause a change in the background values for metallic ions, for which a value of 1.5 is typically used.

The PLI reflects the spatial variation in metallic pollution and the contributions of various metals to sediment pollution. A PLI value of < 1 implies no pollution, and a PLI value of > 1 represents metallic pollution in the sediment (Suresh et al. 2012). The equations for calculating PLI are as follows:

$$ {\mathrm{P}}_{ij}={\mathrm{C}}_{ij}/{\mathrm{B}}_i $$
(2)
$$ {\mathrm{P}\mathrm{LI}}_j=\sqrt[m]{{\mathrm{P}}_{1j}\times {\mathrm{P}}_{2j}\times \cdots \times {\mathrm{P}}_{mj}} $$
(3)
$$ {\mathrm{PLI}}_{\mathrm{zone}}=\sqrt[n]{{\mathrm{PLI}}_1\times {\mathrm{PLI}}_2\times \cdots \times {\mathrm{PLI}}_n} $$
(4)

where Pij is the pollution coefficient of metal i at site j, Cij is the content of metal i at site j, Bi is the geochemical background value for metal i, PLIj is the PLI of site j, PLIzone is the PLI of a specific region (basin), m is the total number of metals involved in the assessment, and n is the number of sites sampled.

The SQGs proposed by Macdonald et al. (2000) include effect range-low (ERL), effect range-median (ERM), threshold effect level (TEL), and probable effect level (PEL). A metal concentration lower than the ERL and TEL in sediment indicates that benthic organisms are not influenced, a metal concentration between the ERL and ERM and between the TEL and PEL indicates that adverse effects on benthic organisms occasionally occur, and a metal concentration higher than the ERM and PEL indicates that toxic impacts on benthic organisms frequently occur (Hahladakis et al. 2013; Wang et al. 2014).

The RI is the most commonly used method for risk assessment of metallic pollution in sediment. The background values for heavy metal regions and toxic factors are considered, and the sensitivity of benthic organisms to metallic pollution in an assessment area is evaluated. The equations for the calculations are as follows:

$$ {\mathrm{C}}_{\mathrm{f}}^{\mathrm{i}}={\mathrm{C}}_0^{\mathrm{i}}/{\mathrm{C}}_{\mathrm{n}}^{\mathrm{i}} $$
(5)
$$ {\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}}={\mathrm{T}}_{\mathrm{r}}^{\mathrm{i}}\times {\mathrm{C}}_{\mathrm{f}}^{\mathrm{i}} $$
(6)
$$ \mathrm{RI}={\sum}_{\mathrm{i}=1}^{\mathrm{n}}\kern0em {\mathrm{E}}_{\mathrm{r}}^{\mathrm{i}} $$
(7)

where \( {C}_f^i \) is the pollution index of heavy metal i, \( {C}_0^i \) is the measured concentration of metal i, \( {C}_n^i \) is the geochemical background value for metal i, \( {E}_r^i \) is the RI for metal i, \( {T}_r^i \) is the toxic response factor for metal i, and n is the number of metals at the site. The toxic response factors for Hg, Cd, As, Pb, Cu, Ni, Cr, and Zn are 40, 30, 10, 5, 5, 5, 2, and 1, respectively. The potential ecological risk levels for metals and their thresholds are described in detail by Håkanson (1980).

Statistical analysis

The statistical parameters included the mean, minimum, maximum, standard deviation (SD), and coefficient of variation (CV). The applicability of the experimental data was tested with the Kaiser-Meyer-Olkin (KMO) and Bartlett’s sphere tests. In this study, the KMO and the Bartlett’s sphere test probability values were 0.77 and 0, respectively, suggesting that the principal component analysis results were reliable. A principal component analysis was performed with the physiochemical parameter values for the sediment and a standardized dataset of metallic concentrations, and the principal components were obtained after coupling the rotation of the maximum value of the correlation matrix (Lee et al. 2006). Ward’s method was adopted for the cluster analysis (Han et al. 2006), the Euclidean distance was used as the distance between clusters to couple with the principal component analysis, and individual parameters or variables with similar distance values were grouped (Li et al. 2013). Pearson correlation coefficients were determined to analyze the correlations between various elements. The spatial distribution and the pollution levels of heavy metals in the riverine surface sediment of the Taihu Basin were plotted by ArcGIS 10.2.

Results and discussion

Metal contents and spatial distributions in riverine surface sediments

The mean concentrations of Zn, Cr, Ni, Cu, Pb, As, Cd, and Hg in the riverine surface sediments in the LTB were 163.6, 102.5, 45.5, 44.7, 37.0, 13.3, 0.5, and 0.1 mg/kg, respectively, which were 2.8, 1.3, 2.3, 2.4, 2.4, 1.4, 1.8, and 0.99 times those of the respective background values given by Qu et al. (2001). The concentrations of Ni, Cu, Zn, and Pb in the riverine surface sediments from more than 95% of the sites sampled were higher than the corresponding background values, and the proportions of the sites with concentrations of As, Cd, Cr, and Hg in the riverine surface sediments exceeding the corresponding background values were 81.9%, 76.6%, 63.8%, and 47.9%, respectively.

The concentrations of Cr, Ni, Zn, Hg, and Cu in the riverine surface sediments of the LTB had similar spatial distributions and progressively decreased from north/east to south/west. The highest concentrations of the five metals were found in the surface sediments of TG, the lowest concentrations of Ni, Zn, and Cu all occurred in TX, and the lowest concentrations of Cr and Hg were detected in NH (Fig. 2). The Cd concentrations in the riverine surface sediments of the LTB gradually decreased from northwest to southeast; the highest value was detected in NH, and the lowest value was in HY. The As and Pb concentrations in the riverine surface sediments of the LTB were generally characterized by high concentrations in the north and west and low concentrations in the south and east. The highest concentrations of As and Pb appeared in TX and TG, respectively, and the lowest concentrations of As and Pb were found in HY. The concentrations of As, Cd, and Cu in the surface sediments of the six water systems were significantly different (p < 0.05). The means of As in the surface sediments of the six water systems were ordered TX (16.21 mg/kg) > TG (15.6 mg/kg) > NH (14.1 mg/kg) > YJ (13.7 mg/kg) > HP (13.6 mg/kg) > HY (10.7 mg/kg), the means of Cd were ordered NH (1.27 mg/kg) > TG (0.72 mg/kg) > TX (0.66 mg/kg) > YJ (0.44 mg/kg) > HP (0.36 mg/kg) > HY (0.28 mg/kg), and the means of Cu were ordered TG (83.6 mg/kg) > HP (68.0 mg/kg) > YJ (60.9 mg/kg) > NH (52.5 mg/kg) > HY (50.5 mg/kg) > TX (31.4 mg/kg). In addition, the contents of As and Cd in the surface sediments of the inflow water systems were significantly higher than those of the outflow water systems.

Fig. 2
figure 2

Spatial distributions of the metallic contents in surface sediments from Lake Taihu Basin

Full size image

Metals dominated by natural sources have low CVs, while the CVs of elements influenced by anthropogenic sources are expected to be quite high (Han et al. 2006; Yang et al. 2011). The CVs of Cr, Ni, Zn, and As in the riverine surface sediments ranged from 30 to 90%, whereas the CVs of Cu, Cd, Pb, and Hg exceeded 110%. In particular, the CVs of Cu, Pb, and Hg were 378.14%, 457.1%, and 291.7%, respectively, indicating significant spatial heterogeneity in metallic ion concentrations, which may be mainly triggered by anthropogenic activities (Wang et al. 2014). The TG water system flows through urban development and highly intensive agricultural production areas. The metallic concentrations in the surface sediments of TG are directly related to the discharges of industrial wastewater, urban domestic sewage, and agricultural and livestock husbandry wastes (Wu et al. 2018). The HP, HY, and YJ water systems not only flow through urban areas but also have well-developed water transport, and the metallic concentrations in the surface sediments of these water systems are closely related to urban domestic sewage and ship activities. The TX and NH water systems primarily flow through forested areas but are also affected by domestic sewage point sources and agricultural area sources (Cui et al. 2018). Therefore, frequent human activities in highly developed regions primarily contribute to spatial differences in the metallic concentrations of sediments (Yang et al. 2011). The CV value of the Hg concentrations in riverine surface sediments of the LTB was 291.7%, but the mean Hg concentration was approximately equal to the background value, indicating that the Hg concentrations in riverine surface sediments are not only affected by the physicochemical properties of sediments (Chakraborty et al. 2014; Zhao et al. 2018) but also directly related to the pollution source (Mil-Homens et al. 2013). The CV values of the Cr, Ni, Cu, Zn, As, Cd, and Pb concentrations in the riverine surface sediments were all higher than 30%, and the mean concentrations were greater than the corresponding background values, indicating that the metallic concentrations are considerably affected by anthropogenic input.

Assessment of metallic pollution in riverine surface sediments

According to the pollution threshold Igeo given by Nilin et al. (2013), the means of Pb, Ni, Zn, Cu, and Cd in the riverine surface sediments were at moderate pollution levels, and the proportions of sampling sites with concentrations of these five metals at and above moderate pollution levels were 7.45%, 15.95%, 8.5%, 37.24%, and 9.57%, respectively. The means of Cr, As, and Hg in the riverine surface sediments were at low pollution levels, and more than 50% of the sampling sites were not polluted (Table 1). Ni, Cu, Zn, Cd, and Pb were the primary pollutants of the surface sediments at most sampling sites (> 70%) of NH and TG. Most sampling sites of YJ (> 60%) and HY (> 80%) were polluted by Ni, Cu, Zn, and Pb. Ni, Zn, As, Cu, Cd, and Pb were the primary pollutants in the surface sediments of HP and TX. The Igeo values of the metals in the riverine surface sediments of the LTB were lower than those of economically developed South China (Liao et al. 2017) and were higher than those of the Liaohe River protected area (Ke et al. 2017), indicating that anthropogenic activity, including industrial and agricultural processes (Bastami et al. 2012; Tepanosyan et al. 2018), is an important driving force leading to the enrichment and pollution of metals in riverine surface sediments.

Table 1 Pollution evaluation of the heavy metals in surface sediments from Lake Taihu Basin
Full size table

The comprehensive PLI value of metals in the riverine surface sediments of the LTB ranged from 0.75 to 8.80, and the proportions of sampling sites with no pollution, low pollution levels, and medium and high pollution levels were 8.5%, 51.1%, and 40.4%, respectively. As indicated by the spatial distribution of the PLI values, the comprehensive pollution load of metallic ions in the riverine surface sediments in the northern LTB was the highest, and those in the eastern and southern regions were relatively low (Fig. 3). The mean PLI values of metals in the surface sediments of the six water systems were ordered TG (2.6) > HP (2.0) > YJ (1.9) > NH (1.85) > TX (1.7) > HY (1.5), all of which were at the moderate pollution level. Although the differences in the PLI values of the metallic ions in the surface sediments of the six water systems were not significant (p > 0.05), the metallic pollution in the surface sediments of these water systems was quite severe (Fig. 4), indicating that the metal pollutants in the riverine surface sediments interact, as evidenced by previous studies (Angulo 1996; Jahan and Strezov 2018). The mean PLI value of the metals in the riverine surface sediments of the LTB was 1.91, a moderate pollution level, similar to that of the Tigris Basin (Varol 2011); however, the value was lower than those of urban rivers in northern Bangladesh (Islam et al. 2015).

Fig. 3
figure 3

Spatial distribution of the PLI values of the heavy metals in the surface sediments from Lake Taihu Basin

Full size image
Fig. 4
figure 4

The PLI values of the heavy metals in the surface sediments from 6 water systems of Lake Taihu Basin. The box represents the 10th through to the 90th percentiles (thus, 80% of the data values). The solid line in the box represents the median value and the dotted line, the mean value

Full size image

The SQGs reflect the possibility of toxic effects of metals in sediments on benthic organisms (Smith et al. 1996; Farkas et al. 2007). The mean concentrations of Cd and Hg in the riverine surface sediments of the LTB were lower than those of the corresponding ERLs and TELs; the Cd concentrations at 98.9% and 64.9% of the sites sampled were lower than the ERL and TEL, respectively, and the Hg concentrations at 68.1% and 78.7% of the sites sampled were lower than the ERL and TEL, respectively, indicating that the amounts of Cd and Hg at most of the sites sampled do not adversely affect benthic organisms (Mataba et al. 2016). The mean concentrations of As and Cu in the riverine surface sediments were lower than the corresponding ERLs and were between the corresponding TELs and PELs, and the concentrations of As and Cu, at 97.9% and 71.3% of the sites sampled, respectively, were lower than the corresponding ERLs. However, the concentrations of As at all the sites sampled were higher than the TEL, and the concentrations of Cu at 35.0% of the sites sampled were lower than the TEL. The mean concentrations of Zn and Pb were between the TEL/ERL and PEL/ERM, respectively. The proportions of sites sampled with Zn concentrations lower than the ERL and TEL values were 28.7% and 42.6%, respectively, and the proportions of sites sampled with Pb concentrations lower than the ERL and TEL values were 30.9% and 42.6%, respectively, indicating that the As, Cu, Zn, and Pb in the riverine surface sediments of certain sites sampled occasionally have toxic effects on benthic organisms, as was found in the sediments of the Tajan River in Iran (Alahabadi and Malvandi 2018). The mean concentrations of Cr and Ni in the riverine surface sediments were between the corresponding ERL and ERM and higher than the PEL, respectively. The proportions of the sites sampled where the concentrations of Cr in the riverine surface sediments were lower than the TEL and ERL values were 2.1% and 1.1%, respectively, and the proportions of the sites sampled where the concentrations of Ni in the riverine surface sediments were lower than the TEL and ERL values were 37.2% and 15.3%, respectively, which suggests that Cr and Ni in the riverine surface sediments of most of the sites sampled frequently have toxic effects on benthic organisms, as observed in the surface sediments of the Buyuk Menderes and Gediz rivers in Turkey (Akcay et al. 2003).

The RI values of the metals in the riverine surface sediments of the LTB varied from 55.0 to 2128.4, with a mean of 266.6. The proportions of sites sampled with low potential risk, medium potential risk, and high potential risk were 45.7%, 35.1%, and 19.2%, respectively. In terms of the spatial distribution of the RI values, the potential ecological risk of the metals in the riverine surface sediments was high in the northern LTB and was low in the southern LTB (Fig. 5). The RI values of TG and NH were higher than those in the other water systems, whereas the metallic concentrations of more than 70% of the sites sampled in the HY and TX water systems exhibited low to medium potential ecological risk (Fig. 6). In general, the mean RI values of each water system were not significantly different (p > 0.05); however, the comprehensive potential ecological risk for each water system was at a moderate level. The comprehensive potential ecological risk of metals in the riverine surface sediments in the LTB was lower than that of the Liaohe River protected area, where the RI values ranged from 126.14 to 780.38 with a mean of 358.4 (Ke et al. 2017; Zhang et al. 2017), and was similar to that of Lake Qinghai Basin, where RI values varied from 22.1 to 352.6 with a mean of 150.6 (Li et al. 2016). Overall, metallic ions in the riverine surface sediments have imposed considerable ecological risks to biota in the LTB due to anthropogenic activities. Consequently, attention should be paid to controlling anthropogenic sources of metallic ions in the LTB.

Fig. 5
figure 5

Spatial distribution of the RI values of the heavy metals in the surface sediments from Lake Taihu Basin

Full size image
Fig. 6
figure 6

The RI values of the heavy metals in the surface sediments from 6 water systems of Lake Taihu Basin. The box represents the 10th through to the 90th percentiles (thus, 80% of the data values). The solid line in the box represents the median value and the dotted line the mean value

Full size image

In general, the pollution assessment of the metals in the riverine surface sediments in the LTB using the Igeo, PLI, and SQG methods was basically consistent. Pb, Ni, Zn, and Cu were the primary pollutants in this basin. The Igeo and PLI evaluations indicated that Cd, Cr, and As were all at low pollution levels. The SQG assessment demonstrated that the Cr concentrations of most sites sampled frequently have potential toxic effects on benthic organisms, and the As concentrations occasionally have toxic effects on benthic organisms. The Cd and Hg concentrations in the riverine surface sediments are indicative of potential moderate ecological risk, which may be related to the relatively high toxic response factors of Cd and Hg (Håkanson 1980; Fu et al. 2009; Deng et al. 2016).

Identifying the sources of the metals in the riverine surface sediments

The Mn predominantly originated from bedrock erosion and pedogenesis (Neal et al. 2006; Cai et al. 2015). The Mn concentrations were significantly linearly correlated only with the Fe and As concentrations (Table 2), suggesting that most of the metals in the riverine surface sediments in the LTB are mainly influenced by highly intensive human activities. Significant linear correlations were also observed among the Fe, OM, TP, TN, Cr, Ni, Cu, Zn, As, Cd, Pb, and Hg contents in the surface sediments from the river systems (Table 2), indicating that iron oxide or ferric hydroxide may be the primary binding form of the metals in the surface sediments in the LTB, as evidenced in previous studies (Yin et al. 2011; Yao et al. 2014). The principal component analysis divided the 13 component data of riverine surface sediments into the following four principal components (PCs): PC1, explaining 34.5% of the total variance, was dominated by Cr, Ni, Zn, As, TP, Cu, and Fe; PC2, accounting for 20.0% of the total variance was related to Pb, OM, and TN; PC3, explaining 12.9% of the total variance, was dominated by Mn and Cd; and PC4, explaining 8.0% of the total variance, had strongly positive loading of Hg (Table 3). Cluster analysis was performed to group the 13 parameters analyzed from the riverine surface sediments of the LTB; the parameters were grouped into four clusters (Fig. 7). Cluster 1 included OM, TN, and Pb, identified as contaminants mainly derived from the wastewater discharges of agricultural production and management and waste gas discharges related to traffic. Cluster 2 contained Cd, Mn, and Hg, dominantly stemming from the wastewater discharges of cadmium battery manufacturing plants and agricultural activities related to pesticide use. Cluster 3 included Cu and As, originating from wastewater discharges associated with copper processing and treatment. Cluster 4 contained Cr, Ni, Zn, Fe, and TP, primarily derived from the wastewater discharges of nickel-zinc battery manufacturing plants and industrial and domestic wastewaters.

Table 2 Pearson correlation coefficients among the different variables after logarithm transition
Full size table
Table 3 The principal components with varimax rotation based on the data of the heavy concentrations and physicochemical parameters of surface sediments in Lake Taihu Basin
Full size table
Fig. 7
figure 7

Dendrogram of 13 variables from Lake Taihu Basin based on Ward’s hierarchical clustering method

Full size image

As discussed above, the metals in the surface sediments of the river systems in the LTB can be divided into four groups. The first group, including Pb and strongly correlated with OM and TN, indicates that Pb is primarily sourced from the discharges of domestic sewage and agricultural waste (Kumwimba et al. 2018). Additionally, due to the well-developed transportation infrastructure in the area, the traffic volume is high, and the combustion of Pb-containing gasoline is one of the sources (deMiguel et al. 1997). The second group, consisting of Cr, Ni, and Zn and with relatively high concentrations in the northern areas of the LTB, where electroplating and manufacturing industries are well developed, was strongly related to TP and Fe, suggesting that the contents of the three ions are affected by the wastewater discharges from electroplating and manufacturing plants (Fu et al. 2014). Cu and As belonged to the third group. The As concentrations were high in the surface sediments of NH and TX and were primarily derived from agricultural nonpoint source pollutions. Furthermore, the riverine surface sediments in the industrially developed northern LTB were severely polluted by Cu and As, indicating that the two ions mainly originate from the extensive use of pesticides and industrial wastewater discharges (Niu et al. 2015; Bian et al. 2016). The fourth group included Cd and Hg. Cd is the primary component of electrolytes in storage batteries, and Hg is often used in automobile parts such as thermostats. There are well-established nickel-cadmium battery manufacturing industries and a large number of automobile parts manufacturing plants along the TG water system, and industrial wastewater discharges are the primary source of Cd and Hg (Fu et al. 2014).

The analysis of source apportionment indicates that the metallic ion concentrations and the extent of pollution in the riverine surface sediments are closely related to the surrounding industrial production, agricultural activities, and land use patterns. The most severe metallic pollution of the surface sediments was found in the TG water system in the northern part of the LTB, which is a priority water system for pollution control. A series of physical and chemical steps should be taken to reduce the metallic concentrations in industrial wastewater (He et al. 2016; Sobik-Szołtysek et al. 2017), decrease the total amount of metals entering rivers, and reduce the impact of external inputs on metallic ion enrichment in riverine surface sediments. In addition, the effect of the internal release of metals from the surface sediments of TG should be addressed by sediment dredging (Wu et al. 2014; Choppala et al. 2018).

Conclusions

Toxic metallic environmental behaviors and pollution risks in aquatic ecosystems have become increasingly concerning. The mean concentrations of Zn, Cr, Ni, Cu, Pb, As, Cd, and Hg in the riverine surface sediments of the LTB exceeded the corresponding background values. Significant spatial heterogeneity was observed in which the metallic concentrations in the surface sediments from the northern rivers of the LTB were higher than those from the southern rivers, and the metallic concentrations in the surface sediments from the western rivers were lower than those from the eastern rivers in the LTB. Pollution assessments by the Igeo, PLI, and SQGs methods indicated that Pb, Ni, Zn, and Cu were the primary pollutants and that the Cd and Hg concentrations posed moderate ecological risks to the riverine surface sediments in the LTB. The metallic concentrations and pollution levels in the surface sediments from the river systems were closely related to the surrounding industrial production, agricultural activities, and land use patterns. Discharges of industrial and agricultural wastewater and domestic sewage, extensive use of pesticides, and petroleum combustion were the primary sources of the metals in the riverine surface sediments of the LTB. The TG water system should be prioritized for pollution control and treatment of the metals in the riverine surface sediments. Therefore, certain measures, including improvement of environmental protection laws, should be taken to reduce wastewater discharges containing metallic ions from agricultural and industrial production activities.