Distribution and Contamination Assessment of Soil Heavy Metals in the Jiulongjiang River Catchment, Southeast China

A total of 63 soil samples were collected from three soil profiles (yellow soil, red loam, red soil) from Jiulongjiang river catchment to investigate the distribution, controlling factors, and toxic risks of heavy metals, including Cr, Mn, Fe, Cu, Zn, Cd, Pb, and Ni. The results showed that Cr and Cd in soils were enriched. The relationships between heavy metals and soil properties were assessed by principal component analysis. The results indicated that soil organic matter (SOM) played a fundamental role in controlling Cd and Pb in yellow soil and red loam sites. The Cd was significantly correlated with Pb and Cu, and Cr, Zn, Ni, Fe displayed strong correlations with each other, however, no statistical correlation was found between Cd and Cr. The enrichment factor and geoaccumulation index analyses showed that the soils in the study area were contaminated by Cd. Potential ecological risk analyses indicated that Cd posed a considerable ecological risk in yellow soils, and posed a moderate ecological risk in red loams and red soils.


Introduction
Heavy metals (HMs) have caused significant ecological environmental concern due to their toxicity and persistence. Moreover, heavy metal accumulation in soils is a serious potential threat to ecosystems [1,2]. Heavy metal pollution not only impairs soil's chemical/physical properties and leads to soil nutrient loss, but also affects the soil organisms and contaminates the food chain. Organisms in soil have the ability to accumulate HMs and finally to pose a threat to human health, such as hypophosphatemia, neurotoxicity, liver damage, and heart disease [3]. Examples of heavy metals include Pb, Cd. Tang et al. reported that Pb and Cd are two of the most common toxic HMs which are related to cancers and malfunction of the nervous system [4]. As a nonessential metal, Cd has been classified as a human carcinogen [5], and Cd can enter the brain parenchyma and neurons and lead to neurological alterations, and finally result in memory deficits, attention deficits, and olfactory dysfunction [6]. By studying the data on the national communique of soil pollution survey of China in 2014, soils are significantly polluted in some regions, and the quality of agricultural land soil is particularly concerning [7]. The rapid development of industrialization and the increasing application of agrochemicals have led to the accumulation of HMs in soils [8][9][10]. High concentrations of HMs in soils may lead to ecological damage and threaten the health of humans and animals [11][12][13][14][15].

Study Area
The study sites were located in Jiulong river catchment (24 • 13 -25 • 51 N, 116 • 47 -118 • 02 E), Fujian province, Southwest China ( Figure 1). The study area is controlled by the subtropical oceanic monsoon climate. The average annual temperature is 21 • C and the mean annual precipitation is 1200-2000 mm [33]. The forest coverage accounts for over 60% with varied vegetation. In this catchment, soil is characterized by red loam, red soil, yellow soil, and paddy soil. The area proportion of red loam is about 62%, and red soil is about 16% [34].

Sampling and Analysis
The sampling sites were selected from yellow soil (YS) at abandoned agricultural land, red loam (RL), and red soil (RS) at forest land along Jiulongjiang river catchment in January 2018. Soil profiles of a depth of 100 cm were selected to research the vertical distribution of HMs, because the variation extent of HM contents focuses on this soil layer. A total of 63 soil samples were collected from the three soil profiles, and each profile was cut into 5 cm sections, and the description of sampling sites can be seen in Table 1. Soil samples were air-dried at 25 °C and sieved through a 2 mm nylon sift to remove coarse debris. Then, soil samples were ground until all the particles would pass through a 200-mesh sift. The soils were digested with HNO3-HF-HClO4, and the concentrations of eight heavy metals (i.e., Cr, Mn, Fe, Cu, Zn, Cd, Pb, and Ni) in soils were determined using ICP-MS (Elan DRCe, Perkin Elmer). Soil samples, reagent blanks, and standard reference samples were synchronously analyzed. Total phosphorus (TP), soil particle distribution, and pH were measured following the methods as in our report [35], and soil organic carbon (SOC) and soil organic nitrogen (SON) were measured following the methods reported by Liu [36,37].

Data Analysis
In the present study, the degrees of HM contamination in soils were assessed using enrichment factor (EF) [38] and geo-accumulation (Igeo) [39], and potential ecological risk index (RI) [40] values

Sampling and Analysis
The sampling sites were selected from yellow soil (YS) at abandoned agricultural land, red loam (RL), and red soil (RS) at forest land along Jiulongjiang river catchment in January 2018. Soil profiles of a depth of 100 cm were selected to research the vertical distribution of HMs, because the variation extent of HM contents focuses on this soil layer. A total of 63 soil samples were collected from the three soil profiles, and each profile was cut into 5 cm sections, and the description of sampling sites can be seen in Table 1. Soil samples were air-dried at 25 • C and sieved through a 2 mm nylon sift to remove coarse debris. Then, soil samples were ground until all the particles would pass through a 200-mesh sift. The soils were digested with HNO 3 -HF-HClO 4 , and the concentrations of eight heavy metals (i.e., Cr, Mn, Fe, Cu, Zn, Cd, Pb, and Ni) in soils were determined using ICP-MS (Elan DRC-e, Perkin Elmer). Soil samples, reagent blanks, and standard reference samples were synchronously analyzed. Total phosphorus (TP), soil particle distribution, and pH were measured following the methods as in our report [35], and soil organic carbon (SOC) and soil organic nitrogen (SON) were measured following the methods reported by Liu [36,37].

Data Analysis
In the present study, the degrees of HM contamination in soils were assessed using enrichment factor (EF) [38] and geo-accumulation (I geo ) [39], and potential ecological risk index (R I ) [40] values were selected to evaluate potential ecological risks.
EF is defined as follows: where C m is the measured value of target element in soils, and B m is the background value (BV) of this element, and the reference background content was obtained from the Fujian soils in this study [41]. Aluminum (Al) was used as the geochemical normalizing element. I geo is calculated by the following formula: where C m and B m are the determined value and background value of target element in soils, respectively. In this study, B n denoted the content of heavy metals in the soils of Fujian province [41]. The calculation equation of R I is given as follows: where Er m is the monomial potential ecological risk factor, and C f is the contamination factor, and T m is the biological toxicity factor (i.e., Cd = 30, Cr = 2, Cu = 5, Pb = 5, Ni = 5 and Zn = 1) [40].

Soil Properties and Heavy Metal Concentrations
The natural concentration of HMs in soils largely depends on the parent materials, and its distribution is also influenced by soil properties [42]. Vertical distribution of soil properties, including SOC, SON, pH, TP and clay contents, in the three types soils (red loam, red soil, and yellow soil) are shown in Figure 2. The pH in soil profiles revealed strong acid with values ranging from 3.77 to 4.96, and pH at 30 cm depth decreased greatly in site YS, implying a likely influence of human activity. SOC contents, which ranged from 1.28 to 14.42 g kg −1 , decreased with the increase of soil depth with obvious change trends in upper 30 cm of profiles. SON contents decreased with the increase of soil depth with a range of 0.88-0.11 g kg −1 . Clay content ranged from 9.60% to 15.78% without obvious differences among the three sampling sites. TP contents at site YS were much higher than that at sites RL and RS, which was likely related to the previous application of fertilizer in abandoned agricultural land at site YS. The concentrations of HMs (Cr, Mn, Fe, Cu, Zn, Cd, Pb, and Ni) in soils are presented in Table A1. Compared with the references of BV in Fujian soils [41], the Cr and Cd concentrations in all the profile soils were much higher than BV, and the Mn, Fe, Cu, Zn, Pb, and Ni concentrations were much lower than BV. The distributions of HMs in topsoils were quite different at three sampling sites. Cu, Cd, and Pb contents were the highest at site YS, which was associated with human activities [43]. Zn and Mn contents were the highest at site RL. Cr, Fe, and Ni contents were the highest at site RS and were the highest in red soil, which may depend on soil types [41].

Vertical Profiles of Heavy Metals
Distributions of HMs, including Cr, Mn, Fe, Cu, Zn, Cd, Pb, and Ni, along soil profiles at the three sampling sites are shown in Figure 3. At site YS, the Cr, Fe, Ni, Cu, and Zn contents varied remarkably at 25-35 cm soil layer. Cd, Pb, and Mn content decreased with soil depth in the upper 30 cm, while obvious changes were not observed in the soils lower than 30 cm, which might be related to previous agricultural activities, such as plowing and fertilization [44]. The contents of most HMs (e.g., Cr, Cu, Zn, Pb, and Ni) at site RL showed an irregular variation and fluctuated remarkably, which might be influenced by soil properties. At site RS, the contents of all the eight HMs were the highest in topsoils and decreased slightly with the increase of soil depth, which might be affected by atmospheric deposition related to the combustion of coal and mining activities in Beixi region [31].

Vertical Profiles of Heavy Metals
Distributions of HMs, including Cr, Mn, Fe, Cu, Zn, Cd, Pb, and Ni, along soil profiles at the three sampling sites are shown in Figure 3. At site YS, the Cr, Fe, Ni, Cu, and Zn contents varied remarkably at 25-35 cm soil layer. Cd, Pb, and Mn content decreased with soil depth in the upper 30 cm, while obvious changes were not observed in the soils lower than 30 cm, which might be related to previous agricultural activities, such as plowing and fertilization [44]. The contents of most HMs (e.g., Cr, Cu, Zn, Pb, and Ni) at site RL showed an irregular variation and fluctuated remarkably, which might be influenced by soil properties. At site RS, the contents of all the eight HMs were the highest in topsoils and decreased slightly with the increase of soil depth, which might be affected by atmospheric deposition related to the combustion of coal and mining activities in Beixi region [31].

Pearson Correlation Analysis
Pearson correlation analysis was used to identify the relationship between HM concentrations and soil properties in different soil types (Table 2), such as yellow soil (YS), red loam (RL), red soil (RS). As shown in site YS, positive correlations were found between Cd and TP, pH, SON, and SOC. Pb also had significantly positive correlations with SON and SOC (p < 0.01). However, Cr, Fe, and Zn

Pearson Correlation Analysis
Pearson correlation analysis was used to identify the relationship between HM concentrations and soil properties in different soil types (Table 2), such as yellow soil (YS), red loam (RL), red soil (RS). As shown in site YS, positive correlations were found between Cd and TP, pH, SON, and SOC. Pb also had significantly positive correlations with SON and SOC (p < 0.01). However, Cr, Fe, and Zn had negative correlations with pH, SON, and SOC (p < 0.01). Ni had a negative correlation with SOC. One potential reason may be the adsorption of HMs by SOM, which has a greatly absorptive capacity for metals, such as Cd and Pb [45]. However, humic acid and humin may reduce the contents of some metals in soils [46], so the higher organic matter contents may not immobilize more metals, such as Cr, Fe, and Zn. In accordance with the results from our previous reports, Cd have a strong positive correlation with SOC, while Cr and Fe have negative relationships with SOC in karst soils from southwest China [35]. TP and pH showed a positive correlation with SON and SOC in soils at YS and RS sites. At site RL, Mn, Cd, and Pb maintained a remarkable correlation with TP, pH, SON, and SOC, and negative correlations were found between Ni and TP, pH, SON and SOC. These results can be also explained as the influence of SOM. At site RS, the relationships between the HMs and soil properties, including SOC, SON, TP, pH, and clay, were not close and only Mn showed correlations with TP, SON, and SOC. A weak positive correlation was found between clay with Fe and Zn in yellow soil, and no correlation was observed between clay and HMs in other soil types. Correlation relationships among HMs were determined using Pearson correlation analysis to provide information on their sources and transport [47]. The results of Pearson correlation analysis in soils at all sampling sites are displayed in Table 3. Cd was significantly correlated with Pb and Cu (p < 0.01), indicating that similar geochemical behavior or input sources are likely related to the use of pesticides and fertilizers. Cr, Zn and Ni, Fe displayed a strongly positive correlation (p < 0.01) with each other, suggesting the possibility of their common origin. However, no statistical correlation was found between Cd and Cr, likely indicating the different origins of these two elements.

Contamination Assessment of Heavy Metals
The EF values of selected HMs in soil profiles are depicted in Figure 4. The EF values of most HMs, including Mn, Cu, Zn, Fe, Pb, and Ni, were less than 1.0, and most of them were less than 0.5 at three sampling sites. The EF values of Cr and Cd were much higher than that of other HMs. The Cr was at minimal enrichment levels (1 < EF < 2.0) in most soils without significant changes in EF values along the profiles. The Cd was at significant enrichment level (EF > 5.0) in soils above 30 cm layer at YS site, indicating strong anthropogenic sources, including agrochemicals and chemical fertilizers [48]. Cd was at moderate enrichment level (2.0 < EF < 5.0) in soils above 30 cm layer at RL site, while was at minimal enrichment level (0.5 < EF < 1.5) at RS site, which might be influenced by industrial activities, including mining and fossil fuel, near RL site. layer at YS site, indicating strong anthropogenic sources, including agrochemicals and chemical fertilizers [48]. Cd was at moderate enrichment level (2.0 < EF < 5.0) in soils above 30 cm layer at RL site, while was at minimal enrichment level (0.5 < EF < 1.5) at RS site, which might be influenced by industrial activities, including mining and fossil fuel, near RL site.  in all soil profiles were less than 0, suggesting that soils in the study area were uncontaminated by these HMs. Igeo values of Cr were lower than 1.0 at the four depths without a significant decrease with the increase of soil depth. Combining with the results of EF analysis, the results indicated the soils were not contaminated by anthropogenic inputs, and Cr enrichment in soils mainly depended on bedrocks [2]. Igeo values of Cd were higher than 1.0 in all soils at YS site, while ranged from 0 to 1 in topsoils at RL and RS sites, indicating that Cd pollution level at YS site was more serious than at RL and RS sites, and only topsoils at RL and RS sites were contaminated by Cd.  I geo values of Mn, Cu, Zn, Fe, Pb, and Ni in all soil profiles were less than 0, suggesting that soils in the study area were uncontaminated by these HMs. I geo values of Cr were lower than 1.0 at the four depths without a significant decrease with the increase of soil depth. Combining with the results of EF analysis, the results indicated the soils were not contaminated by anthropogenic inputs, and Cr enrichment in soils mainly depended on bedrocks [2]. I geo values of Cd were higher than 1.0 in all soils at YS site, while ranged from 0 to 1 in topsoils at RL and RS sites, indicating that Cd pollution level at YS site was more serious than at RL and RS sites, and only topsoils at RL and RS sites were contaminated by Cd.
depths without a significant decrease with the increase of soil depth. Combining with the results of EF analysis, the results indicated the soils were not contaminated by anthropogenic inputs, and Cr enrichment in soils mainly depended on bedrocks [2]. Igeo values of Cd were higher than 1.0 in all soils at YS site, while ranged from 0 to 1 in topsoils at RL and RS sites, indicating that Cd pollution level at YS site was more serious than at RL and RS sites, and only topsoils at RL and RS sites were contaminated by Cd.

Assessment of Potential Ecological Risk
Potential ecological risk index (RI) was comprehensively introduced to assess the potential ecological risks caused by HMs, which was considered to be applied in various study domains, such as ecological environment, biological toxicology, and environmental chemistry [40,49]. The results of RI and Erm for Cr, Cu, Zn, Cd, Pb, and Ni in the soils at 0, 30, 60, and 100 cm depth in the three sites

Assessment of Potential Ecological Risk
Potential ecological risk index (R I ) was comprehensively introduced to assess the potential ecological risks caused by HMs, which was considered to be applied in various study domains, such as ecological environment, biological toxicology, and environmental chemistry [40,49]. The results of R I and Er m for Cr, Cu, Zn, Cd, Pb, and Ni in the soils at 0, 30, 60, and 100 cm depth in the three sites are listed in Table 4. Except for Cd, the Er m values of most HMs (<40) decreased as follows: Cd > Cr > Pb > Ni > Cu > Zn, indicating the pollution degree of these HMs. The highest Er m for Cd were found at site YS, which had a considerable ecological risk of Cd (80 ≤ Er m < 160) at 0-100 cm depth according to the description of risk classification [20]. Cd posed a moderate ecological risk (40 ≤ Er m < 80) in topsoils at sites of RL and RS. The R I values exhibited a moderate ecological risk (95 ≤ R I < 190) at site YS, and low ecological risk (R I < 95) at sites of RL and RS, which were associated with the degree of anthropogenic disturbance. The type of land-use at YS site was abandoned agricultural land, which used to be tea plantation several years ago. The sources of Cd pollution of this site were influenced by previous agricultural activities, including fertilizer and pesticide. The accumulation of Cd in topsoils at site RL and RS presumably results from the effects of atmospheric HM deposition from mining and fossil fuel exploitation.