Pronounced transition of heavy metal pollution sources in Chinese agricultural surface waters: The rising prominence of non-point source pollution

Summary

Despite strict government controls on pollutant discharges, heavy metal (HM) levels in China’s surface waters remain elevated above background values. Accurate source identification of HM pollution is essential for effective environmental management and public health protection. This study collected and analyzed water samples from the southwestern North China Plain to assess HM contamination levels, sources, and health risks, employing the absolute principal component score-multiple linear regression (APCS-MLR) model for robust source apportionment and quantification of pollution source contributions. Surface water HMs remained at “clean” levels but exceeded background values by 1–50 times. Source apportionment identified three primary sources: livestock/poultry (48.3%) > industrial (31.6%) > hybrid sources (20.1%), demonstrating a transition from point to non-point source (NPS) dominance. Monte Carlo simulation revealed serious carcinogenic risks for 1.1% of children and 19.5% of adults. These findings highlight evolving HM pollution patterns in China’s agricultural regions, offering important implications for developing nations.

Graphical abstract

Highlights

• •Livestock and poultry sources which account for 76.22% are the main HMs pollution
• •Sources of HMs have shifted from previous point source pollution to NPS pollution
• •As and Cr⁶⁺ were the main reasons for constituting a higher risk of cancer
• •Provide a reference for developing countries in control of HMs pollution from NPS

Environmental science; Pollution; Aquatic science

Introduction

Heavy metals (HMs) in water have attracted worldwide attention due to their potential toxicity and bioaccumulation.¹ Surface water, as an open space, has emerged as a significant medium for transporting and transforming HMs, where it is more susceptible to contamination.^1,2 Nearly 80% of China’s water environment has already suffered from varying degrees of HMs pollution.³ These HMs in water pose hazards to human health through the food chain, drinking water, and other processes.^4,5,6 Long-term consumption of water contaminated with Cr⁶⁺ increases the risk of lung and stomach cancer^7,8,9; more than 40 million people are affected by As drinking water in Bangladesh and about 270,000 of them have died of cancer.^10,11,12 Therefore, investigating the sources of HMs and pollution status in surface water is significant to health risk assessment and HM prevention and control.

The source of HMs in surface water are extensive, including natural and anthropogenic activities.¹³ Natural sources determine the background values of HMs, which mainly originate from rock weathering, soil erosion, groundwater dissolution, and other processes that have a relatively small impact on the water environment.^14,15 Anthropogenic sources of pollution are mainly classified into point source (PS) pollution and non-point source (NPS) pollution. PS mainly refer to industrial production, traffic emissions, metal extraction, mining, and other pollution sources with fixed discharge points. The high concentration and complex composition of pollutants discharged from these sources have a significant impact on the water environment, which are important drivers of the exceedance of HMs and the decrease in the density and diversity of organisms in watersheds.^16,17 NPS pollution mainly includes the application of fertilizers and pesticides, livestock and poultry farming, and other sources of pollution that do not have a fixed point of pollution emission,¹⁸ with characteristics of decentralization, uncertainty and lagging, whose impacts on the water environment are not easy to detect, which is less influential compared to the impacts of HM contamination generated by point source pollution. Moreover, under the active regulation of governments, HMs pollution in river and lake waters around the world has changed from single-metal pollution to mixed-metal sources,¹⁹ but it remains unclear whether the sources of pollution have changed. Therefore, accurate identification of HM sources in surface water not only allows targeted measures to be taken to control the input of HMs but also prevents and reduces the pollution of surface water with HMs, reducing the hazards to the ecological environment and human health.^20,21

The North China Plain is a highly densely populated, urbanized, highly developed industrial and agricultural area in China, with per capita water resources less than 1/8 of the national average, making the issue of water resource protection critical.^22,23With the rapid advancement of urbanization and industrialization, the environmental quality problems with surface water have become increasingly prominent. For example, mine smelting and industrial wastewater discharges have led to HM contamination of 18% of the surface water in the Shaying River Basin, with Hg and Mn exceeding Class III of China’s surface water quality standards²⁴; mining activities led to the enrichment of HMs Cu, Zn, Mo, and Cd in the river basin of Weihe in the North China Plain, increasing the potential ecological risks of the basin.²⁵ Timely investigation of HM pollution sources is crucial for controlling and reducing HM pollution in surface waters of the North China Plain and assessing potential human health risks.²¹ In this study, the surface water of eight typical agricultural areas, Zhengzhou, Xinxiang, Nanyang, Shangqiu, Zhoukou, Zhumadian, Kaifeng, and Hebi, which are located in the southwestern part of the North China Plain in China, were selected to evaluate the contamination status of seven toxic HMs in water, including Cu, Zn, As, Hg, Cd, Cr⁶⁺, and Pb.²⁶ The principal component analysis (PCA) and absolute factor score-multiple linear regression (APCS-MLR) receptor model were used to reveal the main sources of HMs in surface water quantitatively, and the potential risks posed by HMs in surface water to human health were evaluated. Preventive and curative measures against HMs in surface water were proposed to provide a theoretical basis for realizing the sustainable development of the water environment.

Results

Characterization of HMs distribution in surface water

The concentrations of seven HMs Cu, Zn, As, Hg, Cd, Cr, and Pb in surface water in the study area are shown in Table 1. The average concentration of HMs in surface water in the North China Plain decreased in the order of Zn > Cu > Cr > Pb > As > Cd > Hg, with the average values of 3.48E-02, 8.91E-03, 5.39E-03, 2.92E-03, 2.33E-03, 3.41E-04, and 4.11E-05 mg/L, respectively. None of them exceeded the Class III standard of China’s Environmental Quality Standard for Surface Water GB3838-2002 and the drinking water standard of WHO, while 1.91% of the points had exceeded the concentration of Hg. Compared to the previous study,^21,27,28 which suggests that HMs in surface waters of the study area have been better controlled. Background value refers to the geochemical properties of the element itself in the absence of anthropogenic pollution, as show in Table 2^27,29,30 The average values of HMs in water exceeded the background values as a whole, with the average values of Cu, Zn, As, Hg, Cd, Cr⁶⁺, and Pb exceeding the limits by multiples of 14.00, 51.00, 9.00, 8.22, 46.00, 1.19, and 2.38, respectively. Although the HMs in the water meet the prescribed thresholds for HMs, it is still heavily influenced by anthropogenic activities. The coefficients of variation for Cu, Zn, Cd, and Pb exceeded 100%, indicating that the HMs concentrations varied considerably among the sites.

Table 1.

Concentration of HMs in surface water of the study area (mg/L)

	max	min	mean	sd	cv	strand	background value	Exceedance factor
Cu	1.58E-01	1.35E-04	8.91E-03	1.99E-02	2.24E+00	1.00E+00	6.30E-04	14.00
Zn	1.29E-03	2.83E-01	3.48E-02	3.92E-02	1.13E+00	1.00E+00	6.80E-04	51.00
As	9.98E-03	3.00E-04	2.33E-03	2.24E-03	9.65E-01	5.00E-01	2.02E-04	9.00
Hg	1.60E-04	9.90E-07	4.11E-05	9.95E-06	2.42E-01	1.00E-04	5.00E-06	8.22
Cd	1.63E-03	3.50E-05	3.41E-04	3.97E-04	1.16E+00	5.00E-03	7.40E-06	46.00
Cr6+	1.93E-02	3.05E-03	5.39E-03	3.05E-03	5.65E-01	5.00E-02	4.47E-03	1.19
Pb	1.70E-02	9.00E-05	2.92E-03	3.33E-03	1.14E+00	5.00E-02	1.26E-03	2.38

SD: standard deviation; CV: coefficient of variation; strand: surface water class III standard; ND: not detected.

Table 2.

Contribution of principal components and factor loadings

	components
	1	2	3
Cu	0.87	0.15	−0.00
Zn	0.85	0.12	0.26
As	−0.18	−0.04	−0.58
Hg	−0.38	0.03	0.64
Cd	0.14	0.82	0.17
Cr6+	0.21	0.13	0.59
Pb	0.09	0.88	0.01
Eigenvalue	2.21	1.20	1.01
cumulative contribution rate	31.61%	48.73%	63.16%

Source analysis of HMs in surface water

Principal component analysis

The PCA uses correlation among HMs to determine whether the sources are the same; the more significant the correlation, the better the homology.³¹ The Pearson correlation coefficients (p < 0.05) were investigated to reveal the linear correlation between the HMs, and the results are shown in Figure 1. Among them, there was a significant positive correlation between Zn-Cu and Cd-Pb (correlation coefficients in the range of 0.213–0.663, p < 0.05), suggesting that the sources of Zn and Cu may be the same, and the sources of Cd and Pb may be similar.

Figure 1.

Heavy metal correlation thermogram of the study area

According to the KMO and Bartlett’s test of sphericity, where the KMO is 0.692 > 0.5, which indicates that there is a correlation between the variables. Bartlett’s result has a significance of 0.00 less than 0.05 indicating that it is suitable for factor analysis. Three potential pollution sources with eigenvalues greater than one were screened by PCA, as shown in Table 2. The first component accounted for 31.61% of the total variance, with Cu and Zn being the higher contributing elements and the high correlation between these two elements suggesting that they are of similar origin. Cu and Zn are widely used in China for agricultural production in the form of fertilizers and pesticides.^15,32,33 In addition, the North China Plain is not only an essential grain-producing area in China but also a primary livestock and poultry farming area where livestock and poultry farming are closely integrated with agriculture, making the return of manure to the field a typical means of utilization and increasing the probability of HMs in manure entering the soil and surface water. HMs compounds (e.g., copper, zinc, arsenic, chromium, lead, and cadmium) are commonly added to feeds as growth promoters, anti-microbial agents, and phosphorus-containing supplements to enhance the growth performance of animals in livestock farming.^34,35 However, up to 95% of the HMs ingested into the animal’s body are not fully absorbed by the livestock, ultimately being excreted in the feces³⁶ which enter surface water with the agricultural runoff after being returned to the field.^37,38,39,40 Some studies have shown that about 40% of the Zn and Cu in agricultural land comes from animal manure,⁴¹ while 40% of the HMs in the soil leach into surface water via soil leaching,⁴² which leads to the enrichment of Cu and Zn in surface water. Thus, the first component is mainly influenced by livestock farming.

The second component accounted for 17.13% of the total variance, with Pb and Cd being the main elements contributing. Existing studies have demonstrated that industrial activities such as battery production and the manufacture of electrical equipment are essential contributors to the enrichment of Cd and Pb in surface waters.^43,44 The southwestern part of the North China Plain is densely industrialized, accounting for 1/5 of the national GDP as of the end of 2019⁴⁵; however, the realization of economic benefits is accompanied by the destruction of the surrounding surface water environment. Therefore, the second component is judged to be mainly an industrial pollution source.

The third component accounted for 15.13% of the total variance; the primary HMs were Hg and Cr⁶⁺. Cr⁶⁺ showed a significant positive correlation with Cd and Pb, which are considered to have similar sources of contamination. Cr6+ and Hg can pollute the aquatic environment through pesticides and herbicides.^15,46 Furthermore, Cr⁶⁺ has been shown to originate from processes such as fuel combustion and steelmaking potentially,⁴⁶ while Hg may be related to winter coal burning. Most of the study area is rural and rich in agricultural activities. Coal-fired heating is still practiced in rural areas in most of the region, which may contribute to Cr⁶⁺ and Hg pollution in surface water. Accordingly, the third source can be summarized as a mix of agricultural, coal combustion, and industrial emissions.

Source resolution of HMs using APCS-MLR modeling

The use of PCA in source analysis can qualitatively characterize the sources of HMs in water, but it cannot make a quantitative analysis of each pollutant. The APCS-MLR receptor model was based on PCA and aims to differentiate and analyze the contribution of the three main sources to the pollution of HMs in surface water. The model was analyzed by converting the scores of the three main factors into absolute principal component scores and performing multiple linear regression analyses using the measured levels of HM elements in surface water as the dependent variable. In this way, the source contribution of HMs in water sources is quantified and determined. Results showed that except for As (0.366) and Cr⁶⁺ (0.408), the R² of the linear regression of all other surface water quality parameters were in good agreement with greater than 0.5, indicating that the source assignment is reliable.

The linear regression model coefficients and absolute factor scores were used to find the average absolute contribution of sources and the contribution of sources to each HM indicator, as shown in Table S3. The percentage of each PC of HMs in surface water in the three main sources were shown in Figure 2B: livestock and poultry farming sources (76.22%) > industrial activities (12.55%) > hybrid source (11.23%). The contribution of HMs to the source of pollution was shown in Figure 2A. Cu and Zn in surface water were mainly affected by livestock and poultry farming activities, with contributions of 85% and 67.31%, respectively. Industrial activities significantly affected Cd and Pb content, with contribution rates of 69.19% and 88.89%, respectively. As and Hg were more affected by hybrid source, with contribution rates of 60.92% and 74.85%, respectively.

Figure 2.

Percentage of surface water pollution sources

(A) The ratio of the contribution of the source to each HM.

(B) Percentage of surface water pollution sources.

Health risk assessment of HMs in surface water

Health risk based on Monte Carlo simulations

HMs in surface water mainly harm human health through oral intake and dermal contact. Still, since oral information is 3.20–5.50 orders of magnitude more harmful than dermal intake, the health risk from verbal communication is emphasized. The exposure of human beings to HMs in the environment will be calculated from two groups: adults and children.⁴⁷ Liu et. discovered that traditional health risk assessment methods might overstate or understate the health risks associated with HMs. However, Monte Carlo simulation can notably enhance the accuracy of health risk evaluation.⁴⁸ Thus, in this investigation, 10,000 simulations of HMs in surface water were executed with Crystal Ball software v11.1.24 (Oracle, USA), and the assessment outcomes are depicted in Figure 3 and Table S2.

Figure 3.

Carcinogenic and non-carcinogenic risks of HMs in surface water

(A) Carcinogenic risks of HMs in surface water.

(B) Non-carcinogenic risks of HMs in surface water.

For the non-carcinogenic risk of HMs, the contribution of HMs was in the order of Pb, Cr⁶⁺, Cd, Cu, Zn, As, and Hg, as shown in Figure S1. The non-carcinogenic risk indices of As, Cu, Hg, and Zn in surface water are less than 1, which is within the safe risk threshold and does not pose a non-carcinogenic risk to human health. However, some of the simulation results for Pb, Cr⁶⁺, and Cd were greater than 1, posing a non-carcinogenic risk to humans. The mean values for HQ in the population were all less than 1, but there was a 3.32% probability of posing a noncancer risk for children and a 2.33% probability of posing a noncancer risk for adult males and females. Overall, the risk of non-cancer was greater in children than in adult females and males, with Pb being the main contributor to the increased risk of non-cancer, mainly due to the greater environmental sensitivity of children,⁴⁹ a result that is consistent with previous studies.^50,51

Increased intake of inorganic HM elements such as As, Pb, Cd, and Cr⁶⁺ is an important contributor to the increased risk of carcinogenesis.⁵² Figure S2 shows the probability of carcinogenic risk for As, Cd, Cr⁶⁺, and Pb in surface water. The mean annual carcinogenicity risks (CRs) for children were 4.02E-06, 1.4E-07, 2.8E-06, and 2.9E-08; for adult females, 1.64E-05, 5.59E-07, 1.23E-05, and 1.19E-07; and for adult males, 1.53E-05, 5.23E-07, 1.17E-05, and 1.01E −07. The average carcinogenic risk of Cd and Pb was within the safe range and did not pose a risk to humans, whereas the average CR of As and Cr⁶⁺ was within the acceptable range for humans and should be considered a carcinogenic risk. The total carcinogenicity risks (TCR) for the population was 6.99E-06 (children), 2.94E-05 (adult females), and 2.76E-05 (adult males), which are all within the acceptable range for humans. As shown in Figure 4B, the probability distribution of the TCR in children showed a 1.1% probability of having a CR greater than 104; adult females and males had a 19.7% and 19.2% probability of having a TCR greater than 10-4, respectively, which would result in a serious cancer risk, with As and Cr⁶⁺ being the main contributors to the increase in the cancer risk, which suggests that there is a need to pay more attention to the carcinogenic risk of HMs in the study area.

Figure 4.

Distribution of sampling points in the study area

Sensitivity analysis

Results of the Monte Carlo sensitivity analyses are shown in Figure S3. The sensitivity of total carcinogenic risk and non-carcinogenic risk is mainly dependent on the intake of HMs in water (IR) by the population. Pb concentration and frequency of exposure are the secondary contributors to the non-carcinogenic risk; whereas, the frequency of exposure (EF) and the number of days (AT) are the secondary contributors to the sensitivity to the carcinogenic risk. Therefore, reducing intake of contaminated water is an important initiative for TCR and CR for all populations.

Discussion

Source transformation of HMs in surface waters of agricultural areas

The quality of China’s water environment has deteriorated from a localized deterioration in the 1980s to a full-scale deterioration in the 1990s, with the overall state of surface water characterized by “all rivers are polluted and all water is dirty”.⁵³ In particular, some enterprises have discharged massive amounts of wastewater in the course of long-term mineral extraction, processing and industrialization, posing a serious threat to the ecological environment and human health.¹⁹ For example, sewage discharge from a chemical plant in Yueyang City, Hunan Province, in 2006 caused the As content in the water to exceed the surface water quality standard by about 10 times, resulting in a threat to the drinking water of 80,000 residents⁵⁴; more than 300 children in Tai Po Township, Hengdong County, Hunan Province, had excessive blood lead levels as a result of sewage discharge from a factory in 2014.⁵⁵ Scholars have extensively studied the sources of HMs in the environment, as shown in Table 3, which confirms that point source pollution is the main factor affecting the distribution of HMs in the surrounding environment.¹⁵

Table 3.

Point source pollution of HMs in surface water

Heavy Metal	Point	Pollute source	Literatures
Hg, Mn	Shaying River Basin	Industrial wastewater	Ding et al. 24
Cr, Pb, Ni, Cu	Yellow River basin, China	Industrial pollution	Feng et al. 27
Cd, Co, Mo	Tao Wan North Ditch Basin, China	Mineral activities	Zhenyu et al. 25
Cu, Zn, Cd, Pb	Bushey River, China	Industrial pollution	Shuzhen 56
Pb	Songhua River, China	Paper mills	Yanru 57
Cd	Luhun Reservoir, China	Mine wastewater Discharge and tailings leaching	Mengmeng et al. 2
Cr	Henan section of the Yellow River, China	Leather industry	Jihong et al. 58
Cu, Pb, Zn, Hg	Huafei River, China	Industrial and transportation sources	Jin et al. 59
Cd, Cu, Pb	Batlagundu, Tamil Nadu, India	Anthropogenic industrial activities of urbanization	Khdary et al. 60
Cr, Cu, Ni	Dongbao River, China	Industrial activities	Wu et al. 61
As, Cu, Pb	Middle and Lower reaches of the Yangtze River, China	Mineral industry and other inputs	Liu et al. 62
Cd, Cr, Hg	Atlantic coast of Morocco	Coastal industrial activities	Maanan et al. 63

Long term mining and industrial activities have caused a wide range of pollution incidents that have attracted extensive attention from the governmental sector, resulting in increased control of PS pollution since the beginning of the 21st century, with marked improvements in the environmental quality of surface waters. During the period from the “Ninth Five-Year Plan” to the “Thirteenth Five-Year Plan”, the proportion of surface water sections of Class I to III increased by 56% and Class V sections decreased to 35.9%. The successive introduction of policy documents on emission standards for HMs, industrial restructuring, access conditions for HMs, etc., the relevant policy documents are shown in Table S4. The discharge of HMs in wastewater has been reduced by 65% from 2010 to 2019, and PS pollution is gradually being brought under control.^64,65

Ma et al. confirmed that the significant reduction of point source pollution discharges from industrial and domestic sources has improved inland water quality in China, while NPS pollution discharges from agriculture, livestock and poultry farming have caused increasing negative impacts on the water environment whose hazards have gradually come to the fore in the environment.⁶⁶ Globally 30–50% of surface waters are affected by NPS pollution.⁶⁷ For instance, 81% of N and 93% of P in Chinese waters come from NPS pollution; similarly, 94% of N and 52% of P in Danish rivers in the United States are caused by NPS.^68,69 Currently, studies on NPS pollution in surface water mainly focus on organic matter and nutrients,⁷⁰ while relatively few studies have been conducted on the accumulation of HMs and POPs in receiving waters. Studies by Chen Qiang et al. on the bottom water of lakes in China found that industrial sources are no longer the largest source of HMs pollution, and the agriculture have now become the main sources of pollution.⁷¹ Therefore, based on effective control of HMs pollution from PS, NPS pollution has gradually become the main hazardous factor of HM pollution in the Chinese water environment. In our analysis, the PCA-APCS-MLR receptor model was used to analyze the sources of HMs in surface water from agricultural areas in the southwestern part of the North China Plain, and it found that: livestock and poultry sources > industrial sources > hybrid source. Unlike previous studies in Zhumadian, the main sources of HMs in reservoir collected in 2015 were PS pollutions such as industrial activities and manufacturing production;⁷² earlier data on PS pollution in the southwestern watersheds of North China plains are also given in Table S4, which indicated that NPS pollution has become a major source of pollution in the area. Rothamsted, UK, found in the Broadbalk experiment that long-term application of farmyard manure increased the Zn and Cu content of the soil by about 60%,⁷³ livestock farming has become a major cause of HM contamination of soils.^74,75 According to the National Bureau of Statistics, the volume of fecal waste in China will reach approximately 4.20 billion tonnes by 2020, while the uncontrolled stacking of feces during the farming process and their treatment in return to the fields will lead to the pollution of the surrounding water as well as the soil environment, especially after the occurrence of the mega-floods in 2021.⁷⁶ Serious soil erosion in the region accelerates the input of pollutants into surface water from agricultural production and livestock farming in the surrounding area, making livestock farming the main source of HMs in surface water.^77,78 In addition, the curbs on industrial HM emissions by government departments have led to a continuous reduction in HM emissions from point sources, with HM emissions from key industries in Henan Province decreasing by 12% by the end of 2020 compared to 2013, while policies related to HM pollution from NPS are scarce, which has resulted in the status quo of HM pollution from NPS in the region. It can be seen that the heavy metal pollution of surface water bodies in the study area has changed from being dominated by PS pollution to being dominated by NPS pollution. Therefore, it is necessary to take measures to control NPS HM pollution in surface water. On the one hand, controlling the daily diet of livestock and reducing the use of HMs in feed during intensive rearing, thereby reducing the emission of HMs in livestock and poultry feces; on the other hand, the “Organic Fertilizer Agricultural Industry Standard (NY 525–2021)” issued by China’s Ministry of Agriculture sets the concentration thresholds of five highly toxic HMs (Pb, Cd, Cr, Hg, and As), yet Cu and Zn are not taken into account in the standard. Consequently, the government needs to strengthen the management of HMs in the process of livestock and poultry breeding and agricultural production, while further strengthening the means of controlling NPS pollution of HMs in surface water and setting appropriate discharge standards, to effectively minimize the hazards of HMs in surface water to the surrounding environment and human health.

As the largest developing country in the world, industrial development in China has shifted from early extensive development to refined governance, and has transformed its energy structure to a green and low-carbon one. As a result of the energy transition, various industries have taken action to reduce the use of fossil energy, thereby reducing the emission of HM pollutants from industrial as well as transport activities,⁷⁹ leading to a shift in the sources of HM pollution. In European countries, as early as 2003, it was found that 40–90% of HMs in freshwater systems were consistent with NPS such as agriculture, erosion or urban runoff⁸⁰; During 2008, LTC Bonten found that PS pollution in the Netherlands had been reduced by 50–90%.⁴² Evidently, the control of HM NPS pollution will become a necessary path for developing countries, and this study provides a reference for the transformation of HM pollution sources in other developing countries.

Prevention and control measures

The shift of surface water HMs pollution sources to NPS pollution provides new ideas for controlling surface water HMs pollution. First of all, it is necessary to adhere to the principle of NPS pollution management and advocate “source reduction, process control and end-of-pipe treatment”.⁸¹ The relevant departments should take measures to reduce the amount of HMs added to agricultural fertilizers and livestock and poultry feed additives, reducing the input of HMs from source; at the same time, measures should be taken to promote precise fertilization and scientific use of medicines, improve the utilization rate of fertilizers, pesticides and feed additives, and reduce the enrichment of HMs on the surface of the soil, which consequently reduces the migration of HMs into the runoff along with the process of precipitation and other processes. Furthermore, non-point source pollution typically exhibits varying changes during flood seasons, non-flood seasons, rainy seasons, and non-rainy seasons.⁸² Consequently, there is a pressing need to expedite the establishment of a dynamic monitoring network to oversee the real-time fluctuations of HM pollution, guarantee the efficacy and promptness of pollution control initiatives, and enhance policy coordination across different departments.

Health risks of HMs in surface water

Based on the health risk results, it can be seen that the probability of non-carcinogenic risk posed by HMs in surface water is low, but it poses a serious carcinogenic risk to human beings, with more than 88.9% probability of carcinogenic risk to human beings, and the main carcinogenic factors are Cr⁶⁺ and As. Based on the APCS-MLR receptor mode, it was concluded that livestock farming, agricultural production, fuel combustion and industrial pollution source were the primary sources of As and Cr⁶⁺ in surface water. On the one hand, the enrichment of HMs in surface water may be due to the use of pesticides and fertilizers during agricultural practices and the atmospheric deposition of HMs during fuel combustion; on the other hand, the HMs compounds may be added during livestock and poultry production, all of which produce HMs that can pollute the surrounding surface water environment through agricultural runoff. Wang Fei et al. collected livestock and poultry feces and feed samples from the North China Plain for their study and found that there were varying degrees of excess Cr, As, Cu and Zn in feces.⁸³ As in feces are reduced, oxidized and methylated by microorganisms into compounds that are highly soluble, volatile and readily taken up by plants^84,85 Although many developing countries have banned arsenic as a feed additive in pig and poultry production, it is still legal and widely used in China,⁸⁶ which undoubtedly increases the risk of disease in the surrounding population. Previous studies have found that the HMs with a high frequency of exceedance in pig manure are Cu, Zn, Cd and As, and As is second only to Cu and Zn⁸⁷; organic arsenic, degradation products derived from organic and inorganic As can also be detected in feces of poultry such as chickens^88,89,90; Jing Li et al. found that Cr⁶⁺ in animal feces has a low potential environmental risk of harming the surrounding water environment. More seriously, HMs in the soil can also migrate to surface water due to precipitation, which undoubtedly increases the potential health risk of HMs in surface water.⁹¹ Consequently, it is necessary to reduce the use of chemical fertilizers and pesticides in agricultural production and to strengthen the control of fuel combustion. It is also necessary to pay attention to the transport and transformation of HMs in feces in the surrounding environmental media and to strengthen the regulation of fecal emissions. In addition, improvements in vegetation conditions are also needed to increase soil reservoir capacity to decrease erosion caused by precipitation and irrigation and the migration of HMs from soil to surface water. In conclusion, the surface water meets the prescribed thresholds for HMs, yet, there is still a concern over its carcinogenic potential.⁹²

Limitations of the study

This study has several limitations that should be acknowledged. First, the geographical scope was restricted to surface water in agricultural areas of Henan Province, North China Plain. While the findings provide valuable insights into local pollution patterns, their broader applicability may be constrained by regional variations in environmental conditions and contamination sources across China’s diverse agricultural landscapes. Second, the temporal resolution of sampling, though covering four seasons, relied on annual averages for analysis. This approach could potentially mask short-term pollution fluctuations, such as seasonal spikes in fertilizer leaching or rainfall-induced runoff of livestock waste. To address these limitations, future studies should expand spatial coverage to enable interregional comparisons of HM pollution trends and mechanisms. Additionally, higher-frequency monitoring (e.g., monthly or event-based sampling) coupled with hydrological and meteorological data integration would facilitate the development of time-weighted models, thereby enhancing the temporal precision of pollution assessments.

Conclusion

This study provides a comprehensive assessment of the levels, sources, and health risks of HMs in surface water in eight municipalities in the southwestern part of the North China Plain. The results showed that the overall condition of surface water quality in the study area was good, and only 1.91% of the points had Hg exceeding the Class III standard in the surface water quality criteria. Although HMs in the water meets the standard, the mean values all exceeded the background values, which indicate that HMs are highly influenced by anthropogenic activities. In addition, the PCA and APCS-MLR receptor model analyses yielded that the HMs Cu and Zn in surface water mainly originated from livestock and poultry farming, Cd and Pb were mainly affected by industrial activities, As, Cr⁶⁺ and Hg were affected by the combined effects of hybrid source, and fuel combustion. The main sources of the HMs were livestock and poultry farming sources, industrial sources and hybrid source. Unlike the results of previous studies, the contribution of point source pollution to HMs in surface water is gradually decreasing, and NPS pollution has become the main cause of HM enrichment. Monte Carlo results showed that the average overall carcinogenic risk was CR(As) > CR(Cr⁶⁺) > CR(Cd) > CR(Pb), with As and Cr⁶⁺ constituting the main cause of higher carcinogenic risk. Children and adult males and females each had 1.1%, 19.2% and 19.7% TCR outputs at serious carcinogenic risk levels. The results of this study provide an important reference for relevant departments to further strengthen the control of non-point source HM pollution in surface water. On the one hand, it is necessary to strengthen the improvement of the monitoring system of NPS pollution by paying attention to the problem of HM addition during agricultural activities, livestock and poultry breeding, etc., which can further realize the reduction of HM pollution from the origin. On the other hand, it will serve as a valuable reference for other developing countries in the control of NPS HM pollution.

Resource availability

Lead contact

Further information and requests should be directed to and will be fulfilled by the lead contact, Mingshi Wang (mingshiwang@hpu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Date and code availability

• •All data have been deposited at Science date bank, and are publicly available as of the date of publication. DOI is listed in the key resources table.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact on request.

Acknowledgments

This work was financially supported by Henan Province Key R&D Special Project (241111320400).

Author contributions

S.S., Mingya W., Mingshi M.W.: Investigation, conceptualization, writing, formal analysis, data curation, editing, and original Draft; W.M.: Experimental analysis and data curation; L.J. and T.L.: Review and investigation; S.Y.: Experimental analysis and data curation; F.Z., W.M.: Investigation and experimental analysis; Mingshi W.: Conceptualization, methodology, supervision, and project administration.

Declaration of interests

The authors declare no competing interest.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Deposited data
Raw and analyzed data	This paper; Science date bank	https://www.scidb.cn/en/s/JNRZV3
Software
SPSS/window	Statistical software for data science	Version 20, IBM
Oracle Crystal Ball	Statistical software for data science	V11.1.24

Experimental model and study participant details

This study does not use experimental models.

Method details

Description of the study area

The study area is located in eight districts of Zhengzhou, Xinxiang, Nanyang, Shangqiu, Zhoukou, Zhumadian, Kaifeng, and Hebi in Henan Province in the southwestern part of the North China Plain of China, with 61.9% of the total livestock production in Henan Province in 2020.^93,94 It belongs to the continental monsoon climate, spanning the four major river basins of the Haihe River, Yellow River, Huaihe River and Yangtze River, in the transition climate from the northern subtropical zone to the warm temperate zone, with the average annual temperature ranging from 10.5°C to 16.7°C. The distribution of surface water resources within the year is highly concentrated due to atmospheric circulation and other climatic influences. Surface runoff during the flood season is abundant, accounts for 60-80% of the total annual runoff, and is concentrated in several major rainstorms and floods, especially during the heavy rainstorms of 2020, which accelerated the transport and dispersion of pollutants through surface runoff.⁹⁵

Sample collection and analysis

In this study, a total of 70 sampling sites were deployed along the rivers of eight districts along the southwestern part of the North China Plain in 2021, including Zhengzhou (6), Nanyang (11), Hebi (6), Shangqiu (12), Kaifeng (6), Xinxiang (5), Zhoukou (4), and Zhumadian (20). Four samples were collected at each point in different seasons, as shown in Figure 4.

The collection, preservation, transport, and quality control of water samples are in accordance with the "Standard Test Methods for Drinking Water Water Sample Collection and Preservation" (GB/T 5750.2-2006), "Standard Test Methods for Drinking Water Water Quality Analysis Quality Control" (GB/T 5750.3-2006), distilled water was added to each batch of samples as a blank control groups in older to eliminate the sample handling and testing process of contamination, which was performed randomly. Water samples were collected from well-mixed reaches of the river using a cylindrical hook with a rope, rinsing the high-density polyethylene (HDPE) bottles with deionized water and rinsing three times with the original sample water before sampling. The samples were filtered through a 0.45 μm microporous filter membrane in HDPE bottles. Metal precipitation was prevented by adding approximately 2 mL of 65% nitric acid (HNO₃) into the water samples collected at each sampling station, and the samples were stored frozen in a refrigerator at -20°C. The concentration of HMs in the samples was determined using an inductively coupled plasma mass spectrometer (ICPMS-7700X). The detection limits for Cu, Zn, As, Hg, Cd, Cr, and Pb were 0.08-0.32,0.67-2.68,0.12-0.48,0.001-0.01,0.05-0.20,0.11-0.44 and 0.09-0.36. The measurement results showed that the error was within the permissible error range (90-105%) and met the experimental requirements. Meanwhile, 20% repeatability test was performed on the measurement results, with an error within 5%, indicating that the analytical steps of the samples in this study meet the quality requirements.

Methods

Principal component analysis

PCA utilizes the idea of dimensionality reduction to filter out a few major composite indices from many variables which both preserves the original amount of information and simplifies the process, making the evaluation results look much more straightforward.⁹⁶ The PCA uses orthogonal transformations to transform observations of variables.

$$\left\{\begin{array}{c}{\mathsf{X}}_1\mathsf{=}{\mathsf{\partial }}_{11}{\mathsf{F}}_1\mathsf{+}{\mathsf{\partial }}_{12}{\mathsf{F}}_2\mathsf{+}\mathsf{·}\mathsf{·}\mathsf{·}\mathsf{+}{\mathsf{\partial }}_{1\mathsf{m}}{\mathsf{F}}_{\mathsf{m}}\\ {\mathsf{X}}_2\mathsf{=}{\mathsf{\partial }}_{21}{\mathsf{F}}_1\mathsf{+}{\mathsf{\partial }}_{22}{\mathsf{F}}_2\mathsf{+}\mathsf{·}\mathsf{·}\mathsf{·}\mathsf{+}{\mathsf{\partial }}_{2\mathsf{m}}{\mathsf{F}}_{\mathsf{m}}\\ \mathsf{·}\mathsf{·}\mathsf{·}\\ {\mathsf{X}}_{\mathsf{n}}\mathsf{=}{\mathsf{\partial }}_{\mathsf{n}1}{\mathsf{F}}_1\mathsf{+}{\mathsf{\partial }}_{\mathsf{n}2}{\mathsf{F}}_2\mathsf{+}\mathsf{·}\mathsf{·}\mathsf{·}\mathsf{+}{\mathsf{\partial }}_{\text{nm}}{\mathsf{F}}_{\mathsf{m}}\end{array}\right\}$$

$${\mathrm{X}=\text{AE}+\mathrm{E},\mathrm{A}=\left({\partial }_{\text{ij}}\right)}_{\mathrm{m}\times \mathrm{n}}$$

where Xn is the component score; a is the component loadings; F is the measured value of the variable; n is the number of components; m is the total number of variables. In PCA, the principal components of the eigenvalues are larger than the units, which are usually considered to be related to most of the variability in the original dataset.

APCS-MLR modeling

The APCS-MLR model is based on PCA to obtain the normalised heavy metal concentration factor scores APCS, which are then transformed into the source’s concentration contribution to each heavy metal.⁹⁷ The main analytical procedure was subtracting the factor scores at 0 concentration samples from the principal component factor scores to obtain the absolute central component score APCS value. Then, multiple linear regression was done with the absolute central component score as the independent variable and the surface water heavy metal concentration as the dependent variable:

$${\mathrm{C}}_{\mathrm{i}=}\left(b_0\right)i+\sum _{\mathrm{p}=1}^{\mathrm{K}}{\mathrm{b}}_{\text{pi}}\times {\text{APCS}}_{\mathrm{p}}$$

where APCS_p is the Absolute Principal Component Score for Factor p; b_pi is the regression coefficient for multiple linear regression; b_pi×APCS_p is the contribution of factor p to the content of C_i; the mean value of b_pi × APCS_p for all samples is the average absolute contribution of the source corresponding to factor p; the source contribution corresponding to factor p is the ratio of its average absolute contribution to the contribution of all sources.

Health risk assessment

Health risk assessment is an assessment methodology that relates environmental pollution to human health to quantitatively characterize the risk of harm to people when exposed to polluted environments.⁹⁸ The procedure consists of four steps: hazard identification, dose-effect analysis, exposure assessment and risk characterization.⁹⁹ Monte Carlo simulation (MCS) considers heavy metal concentrations and individual exposure differences. It considers uncertainty through a probabilistic approach to more comprehensively assess the probability distribution of potential health risks.¹⁰⁰ Exposure to HMs such as As, Cr, Pb, Hg, etc. may increase the accumulation of non-carcinogenic and carcinogenic risks. Hosein Alidadi,⁵⁰ Zhou⁴⁷ et al. found that human heavy metal intake through the oral cavity is 3.20-5.50 orders of magnitude greater than that through the skin and the health risks caused by HMs ingested through the oral cavity need to be taken very seriously. Therefore, this study focuses primarily on the health risks of oral ingestion of surface water, the average daily exposure (ADD) risk for HMs under this pathway is:

$$\text{ADD}\mathsf{=}\frac{\mathsf{C}\mathsf{\times }\text{IR}\mathsf{\times }\text{EF}\mathsf{\times }\text{ED}}{\text{BW}\mathsf{\times }\text{AT}}$$

Where C is the measured value of the indicator element (mg/L), IR is the daily oral water intake (L/d), EF is the frequency of exposure (d/a), ED is the duration of exposure (a), BW is the mean body weight (kg), and AT is the average time for non-carcinogenic and carcinogenic effects (d). Children, adults female and male were evaluated and the values for each parameter are shown in Table S1of the supplemental information. Assessment of non-carcinogenic and carcinogenic risks of HM using Hazard Index (HI) and Total Cancer Risk (TCR) methods.

$$\text{HI}\mathsf{=}\sum _{\mathsf{j}\mathsf{=}1}^{\mathsf{m}}{\text{HQ}}_{\mathsf{j}}\mathsf{=}\sum _{\mathsf{j}\mathsf{=}1}^{\mathsf{m}}\frac{{\text{ADD}}_{\mathsf{j}}}{{\text{RfD}}_{\mathsf{f}}}$$

where HQ_j is the non-carcinogenic risk for element j and RfD_j is the reference dose (mg/(kg·d)) for element j under this exposure route, as shown in Table 1. A Hazard Quotient (HQ) or HI value < 1 indicates no non-carcinogenic health risk, and the higher the value, the higher the risk.

$$\text{TCR}\mathsf{=}\sum {\text{CR}}_{\mathsf{j}}\mathsf{=}\sum {\text{ADD}}_{\mathsf{j}}\mathsf{\times }{\text{SF}}_{\mathsf{j}}$$

where CR_j is the carcinogenicity risk index for carcinogen j and SFj is the reference dose (mg/(kg·d)) of carcinogen j under this exposure pathway, whereas shown in Table 4. Consider carcinogenic risk when TCR or CR value > 10⁻⁴; a 10⁻⁶ < TCR or CR < 10⁻⁴ indicates an acceptable carcinogenic risk; while a TCR or CR < 10⁻⁶ indicates no carcinogenic risk.

Table 4.

Rfd_i and SF_i values for each heavy metal

Parameters	Cr	Cu	Zn	Cd	As	Hg	Pb
RfD [mg/(kg·d)]	3.00E-03a	5.00E-03a	3.00E-1a	5.00E-04a	3.00E-04a	3.00E-04a	1.40E-03a
SF [mg/(kg·d)]	5.00E-01b	–	–	6.10E-00b	1.50E-0b	–	8.35E-03b

^aAskari et al.¹⁰¹.

^bEPA US et al.¹⁰².

Monte Carlo simulation

Monte Carlo simulations were implemented using Crystal Ball software v11.1.24 (Oracle, USA), and in this study, the concentration data of Cu, Zn, As, Hg, Cd, Cr⁶⁺, and Pb were fitted to probability distributions to perform 10,000 simulations of the health risks of HMs.¹⁰³

Quantification and statistical analysis

There is no statistical analysis in article.

Additional resources

We have no relevant resource.

Published: April 23, 2025