Introduction

River sediments are the basic and integral components of fluvial ecosystems, not only providing nutrients for benthic organisms, but also serving as sinks and sources for pollutants, such as heavy metals and metalloids (Akcay et al. 2003; Jiang et al. 2013; Pejman et al. 2015). Globally, the river aquatic ecosystems have been continuously receiving significant amounts of anthropogenic heavy metals and metalloids from industrial, agricultural, and urban disposals for several decades (Feng et al. 2004; Jamshidi-Zanjani and Saeedi 2013; Yang et al. 2009b). Once entering in the rivers, heavy metals and metalloids, regardless of whether they are dissolved or not, are rapidly diluted and transported with hydrologic gradients for hundreds of kilometers and consequently deposited and accumulated in the sediments (Audry et al. 2004; Tam and Wong 2000). However, contaminated sediments are the potential nonpoint sources of heavy metals and metalloids, which may directly pollute overlying waters and pose environmental risk to the fluvial ecosystems, through altering their pH and redox potential, sediment perturbations, and migration of the contaminated benthic biota (Zhao et al. 2015b; Zhao et al. 2014). Therefore, it is a suitable tool to assess the negative influences of heavy metals and metalloids on a fluvial ecosystem by monitoring their contamination status and ecological risk in the sediments (Ciszewski et al. 2012; Heim and Schwarzbauer 2013; Li et al. 2013; Nazeer et al. 2014; Wang et al. 2012).

In the past decades, a series of geochemical and statistical indices has been developed to assess the pollution status and ecological risk of heavy metals and metalloids in the sediments (Dung et al. 2013; Jamshidi-Zanjani and Saeedi 2013; Li et al. 2013; Mamat et al. 2016; Yi et al. 2016). So far, geoaccumulation index (I geo), enrichment factor (EF), potential ecological risk index (RI), modified degree of contamination (mCd), contamination factor (CF), pollution load index (PLI), sediment quality guidelines (SQG), ratio pollution index (RPI), and risk assessment code (RAC) have been extensively employed to assess the contamination and ecological risk of heavy metals and metalloids in the sediments of marine and fluvial ecosystems (Bednarova et al. 2013; Cheng et al. 2015; Dung et al. 2013; Ghrefat et al. 2011; Hu et al. 2013; Jamshidi-Zanjani and Saeedi 2013; Jiang et al. 2013; Mamat et al. 2016; Nazeer et al. 2014; Pejman et al. 2015; Wang et al. 2015; Yi et al. 2016; Zhao et al. 2015a; Zhu et al. 2013b). The I geo is a widely utilized classical geochemical criterion to evaluate the pollution degree of heavy metals and metalloids in sediments (Cheng et al. 2015; Mamat et al. 2016; Yi et al. 2016). The EF is commonly used to assess the contamination status by evaluating the natural or anthropogenic sources of metals in sediments (Bednarova et al. 2013; Mamat et al. 2016; Yi et al. 2016). However, they only focus on individual pollutant, while multi-metal contamination in sediments is common. The mCd produces an overall average value for a range of pollutants to comprehensively assess the extent of contamination in sediments (Cheng et al. 2015). Unlike I geo, EF, and mCd, the RI is a combination ecological risk assessment of toxicity, migration, and transformation of heavy metals and metalloids in sediments (Jamshidi-Zanjani and Saeedi 2013; Pejman et al. 2015; Yi et al. 2016). The bivariate and multivariate statistical methods, Pearson’s correlation coefficient (PCC) analysis, principal component analysis (PCA), factor analysis (FA), and hierarchical cluster analysis (HCA) are also increasingly applied to identify the origin and evaluate the contaminated status of heavy metals and metalloids in rivers and sediments (Jiang et al. 2013; Liu et al. 2003; Pejman et al. 2015; Peluso et al. 2013; Varol 2011; Zhao et al. 2015a; Zhu et al. 2013b).

The Xiangjiang River is the main river in Hunan province, China, with major cities, industry, agriculture, and population of the province, as well as with one of the Chinese top 5 lead (Pb) and zinc (Zn) ore zones in the basin. The river was historically and dramatically polluted by heavy metals and metalloids, due to mining and smelting nonferrous metals for hundreds of years in the basin. Heavy metal and metalloid pollutants in the river have been a critical environmental concern in Hunan province in the last decades and caused increasing attention on the contamination status and ecological risk (Han et al. 2014; Jiang et al. 2013; Mao et al. 2013; Zhu et al. 2013a). However, the study to systematically and comprehensively investigate the contamination status and the ecological risk of heavy metal contamination in the whole Xiangjiang River is limited. Here, we aim to (1) determine the distribution of heavy metals and metalloids (Pb, Zn, copper (Cu), chromium (Cr), cadmium (Cd), nickel (Ni), arsenic (As), cobalt (Co), manganese (Mn), and vanadium (V)) in the surface sediments of the whole Xiangjiang River; (2) assess the contamination status of the heavy metal and metalloid pollutants from different angles with geoaccumulation index (I geo), enrichment factor (EF), and modified degree of contamination (mCd) and the ecological risk by potential ecological risk index (RI); and (3) identify the potential sources of the heavy metals and metalloids by Pearson’s correlation coefficient (PCC) analysis, principal component analysis (PCA), and hierarchical cluster analysis (HCA).

Materials and methods

Study area

The Xiangjiang River, one of the major tributaries of the Yangtze River, is the chief river in Hunan province, China, deriving from the Haiyang Mountain, Lingui county, Guangxi province (Fig. 1). Flowing entirely in Hunan province with typical subtropical monsoon climate, the river rises in southwestern Yongzhou and meanders slowly northward for 856 km to the Dongting Lake, successively passing through Yongzhou, Hengyang, Zhuzhou, Xiangtan, Changsha, and Yueyang districts. The basin totally covers 44.66 % (94,660 km2) of the Hunan province’s area and holds the major cities, industry, agriculture, and population of the province. The basin, deposits abundant nonferrous metal mineral resources, such as Pb, Zn, Cu, and Mn, and as a result, mining and smelting nonferrous metals for hundreds of years have caused severe heavy metal and metalloid contamination in the fluvial ecosystem, especially in the Hengyang and Zhuzhou sections. To systematically and comprehensively investigate the contamination status and ecological risk of the heavy metal and metalloid pollutants in the Xiangjiang River, we, therefore, selected different sampling locations: the origin of the river in Hunan, areas clustering mining and smelting factories, cities, dams, and the mouth of the river (Fig. 1).

Fig. 1
figure 1

Map displaying the study area and sampling sites in the Xiangjiang River

Full size image

Sediment sampling

The sampling sites are located in both sides of the Xiangjiang River. Here, the 16 sampling sites were selected in seven sections, as follows: The sites of Lvbutou, Yongzhou, are located in the origin of the Xiangjiang River of Hunan province (LBTB and LBTN). The sites of Shuikoushan, Hengyang (SBSB, SBSN, SBB, and SBN); Xiawan, Zhuzhou (XWD and XWZ); and Yijiawan, Xiangtan (YJWD) were in the areas clustering mining and smelting factories. The sites in Sanchaji (SCJD, SCJZ, and SCJX) were in Changsha city, and the site in Yijiawan (YJWD) was also in Xiangtan city. The sites in Wangcheng, Changsha (WCX and WCZ), were located before a dam. The sites in Xiangying, Yueyang (XYX and XYD), are the mouth of the Xiangjiang River to the Dongting Lake. As the sampling strategy, four surface sediment (depth:0–15 cm) samples for each sampling site were collected by a grab sampler (ZYC-200B, Hangzhou Yijie Technology Co. Ltd., Hangzhou, China) and stored in clean plastic bags prior to shipping to the laboratories.

Heavy metal and metalloid analyses

The heavy metals and metalloids in the sediment samples were determined following the protocols previously published (Liao et al. 2014; Lin et al. 2013; Wang et al. 2013). Briefly, the sediment samples were desiccated at 60 °C for 14 days, followed by desiccation at 110 °C for 2 days. The dry samples were grounded, homogenized, and sieved with 100-mesh nylon sieves (150 μm) for chemical analyses. Of well-prepared sediment sample, 40.00 mg was completely digested by 1 ml HNO3 plus 3 ml HF at 130 °C for 72 h sealed in a Teflon shaker. Next, 0.5 ml of HClO4 was added after cooling of the treated samples and heated with cap opened to 120 °C for 12 h until the white smoke was gone. Then, the residue was dissolved with 1 ml HClO4 and 1 ml ddH2O in a sealed shaker for 12 h. Finally, the solution was diluted to 40 ml by ddH2O in a Teflon bottle at room temperature for the next step. The concretions of Al, Pb, Zn, Cu, Cr, Cd, Ni, Co, Mn, and V were simultaneously determined with an inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7500 series, USA). In addition, 300.00 mg of the identical sediment sample was digested with 10 ml aqua regia at 95 °C for 2 h, followed by adding 5 ml HCl, 5 ml thiocarbamide, and aqua regia to 50 ml. The concentration of As was determined by an atomic fluorescence spectrophotometer (AFS-810, Beijing Titan Instrument Corp., Beijing, China). The average value of each metal in the four samples of a sampling site was considered as the concentrations of the metal for the site. In order to guarantee the quality of the analysis, laboratory quality assurance and control methods were implemented, including the standard operating procedures, standard calibration, reagent blank analysis, and repeated analysis. Blanks and China Stream Sediment Reference Materials (GBW07309 (GSD-9) and GBW07311 (GSD-11)) were performed for quality assurance and quality control, and recoveries of the standard reference metals were 92–104 %. In addition, each analysis was performed in triplicate and the SD was within ±5 % of the mean value.

Evaluating contamination status of heavy metals and metalloids in the surface sediments

Geoaccumulation index

Since formulated by Müller in 1969, the geoaccumulation index (I geo), a classical assessment model, was widely utilized to quantify the heavy metal and metalloid contamination in soils and sediments by the following formula (Audry et al. 2004; Cheng et al. 2015; Dung et al. 2013; Ghrefat et al. 2011; Jiang et al. 2013; Liu et al. 2009; Müller 1969; Shi and Wang 2013):

$$ {I}_{geo}={ \log}_2\left(\frac{{\mathrm{Me}}_{\mathrm{sediment}}}{{1.5\mathrm{Me}}_{\mathrm{baseline}}}\right) $$
(1)

where Mesediment is the concentration of the examined heavy metal and metalloid, and Mebaseline is the geochemical background value (average shale) or the preindustrial level for the identical element. The constant value 1.5 is the background matrix correction factor due to lithogenic effects. Here, the background concentration of heavy metals and metalloids in the soils of Hunan province in 1980s was denoted as Mebaseline (Table 5) (China National Environmental Monitoring Center 1990). The classification of I geo shown in Table 1 was proposed by Müller (Jiang et al. 2013).

Table 1 Standard of contamination levels by geoaccumulation index (I geo)
Full size table

Enrichment factor

The enrichment factor (EF) was also utilized to assess the degree of heavy metal and metalloid contamination in sediments (Acevedo-Figueroa et al. 2006; Audry et al. 2004; Bednarova et al. 2013; Dung et al. 2013; Ghrefat et al. 2011; Jiang et al. 2013; Zhu et al. 2013b). The formula was as follows:

$$ EF=\frac{{\left(\mathrm{Me}/Al\right)}_{\mathrm{sample}}}{{\left(\mathrm{Me}/Al\right)}_{\mathrm{baseline}}} $$
(2)

where (Me/Al)sample is the concentration ratio of a heavy metal or a metalloid to Al in the tested sample and (Me/Al)baseline is the ratio of the same element to Al in the geochemical background. Aluminum (Al) generally serves as the normalizing element to calculate the EFs of heavy metal and metalloid. The background values of the soils in Hunan province mentioned previously were used as the reference values. The categories of EF are summarized in Table 2 (Acevedo-Figueroa et al. 2006).

Table 2 Contamination categories based on enrichment factor (EF)
Full size table

Modified degree of contamination

The modified degree of contamination (mCd) developed by Abrahim and Parker (Abrahim and Parker 2008; Cheng et al. 2015) was applied to evaluate the comprehensive contamination of multiple heavy metals and metalloids in each sediment sample. The equations to calculate mCd are as follows:

$$ {C}_f^i=\frac{{\mathrm{Me}}_{\mathrm{sample}}^i}{Me_{baseline}^i} $$
(3)
$$ \mathrm{m}{\mathrm{C}}_{\mathrm{d}}=\frac{\sum_{i=n}^{i=1}{C}_f^i}{n} $$
(4)

where \( {C}_f^i \) is the contamination factor of the element i and \( {\mathrm{Me}}_{\mathrm{sample}}^i \) and \( {\mathrm{Me}}_{\mathrm{baseline}}^i \) are the concentrations of the element i in the sample and in the background, respectively; n is the number of the analyzed elements. The mCd is categorized in Table 3 (Abrahim and Parker 2008).

Table 3 Classifications for modified degree of contamination (mCd)
Full size table

Assessing potential ecological risk of heavy metals and metalloids in the surface sediments

The potential ecological risk index (RI) proposed by Hakanson (1980) was utilized to quantitatively assess the ecological risk degree of heavy metals and metalloids in aquatic sediments (Jamshidi-Zanjani and Saeedi 2013; Pejman et al. 2015; Zhu et al. 2013b). RI is defined as follows:

$$ {E}_r^i={T}_r^i\times {C}_f^i={T}_r^i\frac{{\mathrm{Me}}_{\mathrm{sample}}^i}{{\mathrm{Me}}_{\mathrm{baseline}}^i} $$
(5)
$$ RI=\sum {E}_r^i, $$
(6)

where \( {E}_r^i \) is the potential ecological risk factor of the heavy metal or metalloid i, \( {T}_r^i \) is the toxic response factor of the element i, \( {\mathrm{Me}}_{\mathrm{sample}}^i \) is the concentration of the element i in the examined sediments, and \( {\mathrm{Me}}_{\mathrm{baseline}}^i \) is the background concentration of the element i. The \( {T}_r^i \)s for Pb, Zn, Cu, Cr, Cd, Ni, As, Co, Mn, and V are 5, 1, 5, 2, 30, 5, 10, 5, 1, and 2, respectively (Cao et al. 2009; Jamshidi-Zanjani and Saeedi 2013). The evaluated criteria of RI are classified in Table 4.

Table 4 The potential ecological risk criteria for heavy metal and metalloid contamination
Full size table

Statistical analysis

Principle component analysis (PCA) was performed as the previous protocols (Jiang et al. 2013; Peluso et al. 2013; Zhao et al. 2015a; Zhu et al. 2013b). Briefly, the Kaiser-Meyer-Olkin (KMO) and Bartlett sphericity tests were performed to examine the validity of PCA and the covariance matrix was built based on the measured values of all heavy metals and metalloids in sediment samples followed by eigenvalue decomposition. According to the values and contribution rates of the eigenvalues (>1), the principle components were selected and the principle component values of the samples were calculated by the measured values of the elements with varimax rotation. Hierarchical cluster analysis (HCA) was generated on the standardized data sets (Z-cores) by the Euclidean distance method and the Ward method, and as a result, the dendrogram visually demonstrated the cluster relationship between the ten heavy metals and metalloids (Jiang et al. 2013; Tang et al. 2014; Zhu et al. 2013b). The relationship among metals and metalloids was also analyzed by Pearson’s correlation matrix to verify the results by multivariate analysis. The PCA, HCA, and Pearson’s correlation analyses were processed by SPSS 18 for Windows (SPSS, Inc., USA).

Results and discussion

Contamination distribution of the ten heavy metals and metalloids in the surface sediments of the Xiangjiang River

The concentrations of metals and metalloids in the surface sediments of the Xiangjiang River are summarized in Table 5. Generally, the concentrations of Pb, Zn, Cu, Cr, Cd, Ni, As, Co, Mn, and V ranged from 24.45 to 672.3, 30.7–1009.65, 24.55–250.05, 67.9–170, 4.25–31.2, 15.95–187.2, 13.55–122.1, 7.85–53.9, 748.05–3412, and 43.95–210.6 mg/kg dry weight, respectively. The mean concentrations of these metals in the sediment samples followed the order Mn > Zn > Cr > V > Pb > Cu > As ≈ Ni > Co > Cd (Table 5). Geographically, the highest concentrations of Pb, Zn, Cu, Cd, and As were clustered in the sites SBN of Shuikoushan, Hengyang, and XWD and XWZ of Xiawan, Zhuzhou, which were consistent with those of the major Pb and Zn mining and smelting factories located in the two areas. Similar results were shown in the previous researches on the Xiangjiang River (Han et al. 2014; Jiang et al. 2013; Mao et al. 2013; Zhu et al. 2013a). Compared with other rivers and lakes in China, the concentrations of the ten heavy metals and metalloids in the surface sediments of the Xiangjiang River caused serious concern and indicated that, therefore, anthropogenic activities possibly had caused heavy metal and metalloid contamination in the river (Fu et al. 2014; Hu et al. 2013; Li et al. 2013; Lin et al. 2013; Liu et al. 2003; Tang et al. 2014; Wang et al. 2015; Zhao et al. 2015a; Zhu et al. 2013b). Interestingly, the highest concentrations of Cr, Ni, Co, and V were in the site LBTB (Lvbutou in Yongzhou), while the concentrations of heavy metals and metalloids in the site LBTN were much lower than in the site LBTB. It is possibly that these heavy metals were from the mining and smelting activities in Guangxi province and were transported with hydrologic gradients (Audry et al. 2004; Liu et al. 2014; Tam and Wong 2000).

Table 5 Concentration of heavy metals and metalloids in the surface sediments of the Xiangjiang River (mg/kg)
Full size table

Contamination status

Geoaccumulation index

According to the standard of geoaccumulation index (I geo) in Table 1, Cd had caused severe pollution (class 6) in all sampling sites, while Pb, Zn, Cu, As, and Mn had caused moderate contamination (classes 2 and 3) (Fig. 2a). The contamination status of the ten heavy metals and metalloids were dramatically variant in different geographical sampling sites. Strong contaminations of Pb, Zn, Cu, and As were clustered in the sampling sites SBN, XWD, XWZ, and YJWD in which several large-scaled Pb–Zn mining and smelting factories were distributed, while Mn was moderately polluted in the river randomly (Fig. 2a). Moreover, the contaminations of the heavy metals and metalloids were more severe in urban and industry areas (SBB, SBN, XWD, XWZ, YJWD, SCJD, SCJZ, and SCJX) than in agriculture areas (LBTB, LBTN, XYX, and XYD) (Fig. 2a). All sampling sites were rarely polluted by Cr, Ni, Co, and V (I geo < 1).

Fig. 2
figure 2

The geoaccumulation indexes (I geo) (a) enrichment factors (EF) (b), modified degree of contamination (mCd) (c), and potential ecological risk index (RI) (d) of heavy metals and metalloids (Pb, Zn, Cu, Cr, Cd, Ni, As, Co, Mn, and V) in the surface sediments of 16 sampling sites in the Xiangjiang River

Full size image

Enrichment factor

To explore the potential sources of heavy metals and metalloids (anthropogenic and/or natural) in the surface sediments of the Xiangjiang River, enrichment factor (EF) was applied by calculating the degree of metal contamination compared to the background (Acevedo-Figueroa et al. 2006; Bednarova et al. 2013; Dung et al. 2013; Jiang et al. 2013). Al was selected as the conservative reference element because it is rich in the earth’s crust and assumed inert to anthropogenic activities (Acevedo-Figueroa et al. 2006; Wang et al. 2015). Generally, an EF value between 0.5 and 1.5 suggests that the given heavy metal or metalloid may be entirely from crustal materials or natural weathering processes. However, it is considered as a significant portion from anthropogenic sources when the EF value is greater than 1.5 (Zhang and Liu 2002; Zhang et al. 2009). The EF values of Cd (range 39.08–247.87; average 119.42) were highest in this study, indicating an extremely severe enrichment mainly from anthropogenic mining and smelting activities (Table 2; Fig. 2b). The following metals and metalloids were Pb (range 1.91–19.35), Zn (range 0.54–8.75), Cu (range 1.70–8.02), and As (range 0–8.27), and the mean values of them were 7.96, 5.17, 4.09, and 3.85, respectively, which suggested that they also originated from anthropogenic sources in most samples, especially in Shuikoushan of Hengyang, Xiawan of Zhuzhou, and Yijiawan of Xiangtan (Fig. 2b). Only V (range 0.95–2.45; average 1.22) was minor enriched, indicating that they are entirely from crustal materials or natural weathering processes (Table 2; Fig. 2b).

Modified degree of contamination

Different from I geo and EF revealing the contamination status of single element, the modified degree of contamination (mCd) modified from the degree of contamination (C d) was applied to calculate the average contamination value of all heavy metals and metalloids in sediment samples (Abrahim and Parker 2008). The contamination status shown by mCd in the Xiangjiang River was similar with the results by I geo and EF mentioned previously: the whole river was, at least, contaminated in high degree (class 3, Table 3) by all heavy metals and metalloids and very high degree of contamination (class 4, Table 3) clustered in Shuikoushan (SBN), Xiawan (XWD and XWZ), and Yijiawan (YJWD) (Fig. 2c).

Potential ecological risk

The potential ecological risk index (RI) is a comprehensive assessment of toxicity, migration, and transformation of heavy metals and metalloids in sediments (Hakanson 1980; Zhu et al. 2013b). As shown in Fig. 2d, according to the values of the average of potential ecological risk factor (\( {E}_r^i \)), the ecological risk of the ten heavy metals and metalloids ranked as the following order: Cd (3256.32) ≫ Pb (36.18) > As (30.59) > Cu (18.56) > Ni (8.96) > Co (6.95) > Zn (4.70) > Cr (3.37) > Mn (3.29) > V (1.11) in all sampling sites of the Xiangjiang River. The results indicated that Cd caused most severe potential ecological risk in the river, although its concentration was the lowest in all heavy metals and metalloids (Fig. 2d). Conversely, Mn induced low potential ecological risk (class 1, Table 4) in the Xiangjiang River, while its concentration was the highest in all the ten heavy metals and metalloids in the sampling sediments (Fig. 2d). Geographically, Pb induced a considerable potential ecological risk (class 3, Table 4) in the sampling sites XWD, XWZ, and YJWD, and Pb, Cu, and As induced a moderate risk (class 2, Table 4) in SBN, XWD, XWZ and YJWD, although Pb, Cu, and As were of low risk in other sites (Fig. 2d). The results of RI indicated that there was a very high potential risk in all sampling sites of the Xiangjiang River due to extremely severe potential ecological risk of Cd and the trend in the whole river was similar with the I geo, EF, and mCd. The results also suggested that long-term mining and smelting nonferrous metals are tightly associated with severe heavy metal and metalloid contamination and posed incredibly severe potential ecological risk in the Xiangjiang River (Mao et al. 2013; Zhu et al. 2013a).

Based on the results of total concentration, I geo, EF, mCd, and RI described previously, three groups of the ten metals and metalloids could be distinguished with different contamination behaviors: (1) Cd had a unique behavior that it caused severe contamination and posed a very high ecological risk in the whole river, but its concentration in all samples was the lowest among the metals. It is possibly that Cd is more sensitive to anthropogenic activities, as well as that Cd has a major chemical form more easily to dissolve and transport in sediments (Yang et al. 2009a). (2) Pb, Zn, Cu, and As showed highly variable distributions related to local pollution where major Zn–Pb mining and smelting factories are located. Such contamination distribution is also consistent with the pollutants emitted from the current techniques of Zn–Pb mining and smelting in China (Qi et al. 2016). (3) Cr, Ni, Co, Mn, and V showed homogeneous distributions among all sampling sites with low contamination level.

Potential resources of the ten heavy metals and metalloids

To further investigate the relationship and identify the potential sources of heavy metals and metalloids in sediments, bivariate and multivariate statistical analyses, such as Pearson’s correlation coefficient (PCC) analysis, principle component analysis (PCA), and hierarchical cluster analysis (HCA), were frequently applied (Jiang et al. 2013; Peluso et al. 2013; Tang et al. 2014; Zhao et al. 2015a; Zhu et al. 2013b). In all of the metal pairs, the relationship was positive at 99 % confidence level, except the pair Pb–Mn in the Pearson’s correlation matrix (Table 6). As shown in Table 6, Pb, Zn, Cu, and Cd had a significant positive correlation (>0.66) with each other (P < 0.01), while Ni, Co, and V composed another group (>0.866), indicating that two groups possibly originated from different resources. PCA was performed based on the concentrations of all heavy metals and metalloids with varimax rotation (Table 7). The values of Kaiser-Meyer-Olkin and Bartlett were 0.761 (>0.6) and 148.226 (d f  = 45, P = 0.000 < 0.001), respectively, indicating that PCA can be applied in dimensionality decompositions. The PCA results revealed that the variability of heavy metals and metalloids can be expressed by three principal components that explained 87.309 % of the total variance (Table 7). The first principal component (PC1) accounting for 54.625 % of the total variance had strong positive loadings of all of the heavy metals and metalloids (>0.6), except Mn, which indicated that the PC1 should possibly be anthropogenic activities, including industrial, agricultural, and urban (Han et al. 2014; Jiang et al. 2013; Liang et al. 2015; Mao et al. 2013; Zhu et al. 2013a). The second principal component (PC2) explained 21.006 % of the total variance and has negatively loaded Pb, Zn, Cu, and Cd. It was consistent with the results of I geo, EF, mCd, and RI that the four heavy metals severely contaminated the sampling sediments especially in Shuikoushan of Hengyang, Xiawan of Zhuzhou, and Yijiawan of Xiangtan and with that they are the main heavy metal pollutants of the current Zn–Pb mining and smelting technology in China (Qi et al. 2016). The third principal component (PC3) (11.678 % of total variance) only had a strong positive loading of Mn, implying that Mn originated from nature resource and/or Guangxi province (Liu et al. 2014). Hierarchical cluster analysis (HCA) generated similar results with PCC and PCA. As shown in Fig. 3, it was divided into two major clusters: (1) Pb, Cu, Zn, and Cd and (2) the rest of metals were further split into two subcluster: (i) Co, V, Ni, Cr, and Al and (ii) Mn and As, which indicated that the three groups of heavy metals and metalloids originated from different potential resources. The group (1) (Pb, Cu, Zn, and Cd) possibly originated from anthropogenic activities, especially from Pb–Zn mining and smelting industry. In the group (2), the subgroup (ii) (Mn and As) may be derived from Guangxi province (Liu et al. 2014) and the subgroup (i) (Co, V, Ni, Cr, and Al) possibly from nature resource because Al was the major natural element in the crust.

Table 6 Pearson’s correlation matrix for the heavy metal and metalloid concentrations in the surface sediments of the Xiangjiang River
Full size table
Table 7 Eigenvalues, variables and rotation of principal component analysis (PCA)
Full size table
Fig. 3.
figure 3

Dendrogram of hierarchical cluster analysis (HCA) to demonstrate the relationship between the ten heavy metals and metalloids in all sampling sediments of the Xiangjiang River, Hunan. Al was the reference element. Distance metrics were based on the Euclidean distance method and the Ward method

Full size image

Conclusions

Heavy metal and metalloid pollutants in the Xiangjiang River have been a critical environmental concern in Hunan province in the last decades. To systematically and comprehensively investigate the contamination status and ecological risk of heavy metal and metalloid contamination in the river, the geoaccumulation index (I geo), enrichment factor (EF), modified degree of contamination (mCd), and potential ecological risk index (RI) were employed to assess the contamination status and potential ecological risk of heavy metals and metalloids in the surface sediments of the Xiangjiang River and the Pearson’s correlation coefficient (PCC) analysis, principal component analysis (PCA) and hierarchical cluster analysis (HCA) were processed to identify the potential sources of the pollutants. The results showed that the mean concentrations of the ten heavy metals and metalloids in the sediment samples followed the order Mn > Zn > Cr > V > Pb > Cu > As ≈ Ni >Co > Cd. I geo, EF, mCd, and RI revealed that Cd, followed by Pb, Zn, and Cu, caused severe contamination and posed very high potential ecological risk in the Xiangjiang River, especially in Shuikoushan of Hengyang, Xiawan of Zhuzhou, and Yijiawan of Xiangtan. PCC, PCA, and HCA indicated that Cd, Pb, Zn, and Cu possibly originated from Pb–Zn mining and smelting industry and clustered in the sampling sites Shuikoushan, Xiawan, and Yijiawan. Given that the contamination status and the potential ecological risk of Cd plus Pb, Zn, and Cu were severe in the Xiangjiang River, it is urgent to further research on their ecotoxicity to benthic organisms and efficient remediation to the contaminated sediments in the future.