Status of Sustainable Balance Regulation of Heavy Metals in Agricultural Soils in China: A Comprehensive Review and Meta-Analysis

Abstract

To control heavy metal pollution effectively, prevention of heavy metal accumulations in agricultural soils should be the priority rather than remediation of heavy metal contamination. In this research, papers which contained input and output fluxes of Cd, As, Cr, Hg, and Pb in topsoil (the plough layer) of agricultural lands in Hunan, Zhejiang, the Yangtze River Delta, Hainan, and China as a whole were explored. Fluxes of heavy metal species were recalculated, and future trends of pollution were predicted. Also, cases regarding the application of technologies to control the input and output of heavy metals were analyzed. Results indicated that atmospheric decomposition was the dominant input source of heavy metals in all study sites except Hainan. The relative contributions of fertilizers, irrigation water, and straw returning fluctuated greatly among different sites. Cd pollution in all sites was the most serious, followed by Cr and Pb. In Hunan, Cd has already exceeded the maximum limit value and needs to be controlled urgently. The input of heavy metals from irrigation water, fertilizers, and straw returning could be controlled by proposing more policies to manage their quality and application amounts. The amount of heavy metals absorbed by plants could be increased by cultivating crops with hyperaccumulators and high-biomass plants.

1. Introduction

Soils are one of the most basic and precious natural resources on Earth, especially agricultural soils, which form the basis of agricultural production [1,2]. The global area of agricultural soils is limited. It only occupies approximately 38% of the Earth’s land surface [3]. In recent years, soil heavy metal pollution has become an environmental issue that cannot be neglected. The accumulation of heavy metals in agricultural soils causes depletion and degradation of limited soil resources, which has raised the global attention of most countries worldwide [4].

In China, the problem of soil heavy metal pollution is especially widespread and serious [5]. As a country which has a population of more than 1.4 billion, China’s self-sufficiency rates of wheat and maize, and rice exceeded 95% and 80%, respectively [6]. Consequently, the environmental quality of agricultural soils in China affects food safety and thus influences the health of citizens significantly. In recent years, due to the high industrialization and urbanization rate, the heavy metal pollution of agricultural soil has become more serious, especially in large industrial areas like the Yangtze River delta and the Pearl River Delta [7,8]. According to the outcomes of meta-analysis research, the number of heavy metal pollution cases in China increased significantly over the past 20 years, while the distribution of these cases also expanded [9]. In addition, several previous studies also showed that the pollution of Pb, Hg, and Cd is now serious in China and should be controlled with high efficiency [9,10,11,12].

China attaches great importance to the regulation of soil heavy metal pollution [13,14]. At present, there are many studies working on the control of soil heavy metal pollution in agricultural soils. However, most of them focus on the remediation of polluted agricultural soils. Instead of controlling heavy metal accumulations that have already happened, the prevention of pollution should be the priority [4]. In order to prevent heavy metal accumulation, exploring and regulating the balance of heavy metals should be implemented [15,16]. The balance consists of input and output fluxes of heavy metals. For input sources of heavy metals, the direct control of them can fundamentally diminish the accumulation of heavy metals in soil and the occurrence of soil pollution [17]. The accumulation of heavy metals in soil can cause irreversible harm to the soil environment. If the heavy metals are treated before they enter the soil system, the difficulty of remediation and the time required for remediation can be significantly reduced [18]. In addition, if these sources are not controlled, to maintain the concentration of heavy metals within a suitable range, remediation strategies need to be implemented repeatedly and manually. The intensive application of these strategies can result in high costs and severe environmental impact, which make remediation work ineffective and non-sustainable [19,20]. For output sources, plant removal, leaching, and surface runoff are the main output sources of heavy metals in agricultural soils [21,22]. The total amount of heavy metals removed by plants can be significantly affected by humans. For example, the factitious introduction of hyperaccumulators and non-hyperaccumulator large biomass plants are the two main alternatives to increase plant uptake of heavy metals [23]. By adjusting the inputs and outputs of heavy metals, the dynamic equilibrium of heavy metal input and output fluxes can be reached, and the net fluxes of heavy metals in agricultural soils can be minimized over a long period.

The concept of soil heavy metal sources (balance) regulation was first proposed systematically by Moolenaar and Lexmond (2000) [24]. Its connotation can be summarized as carrying out quantitative material flow analysis (SFA) on agricultural soil and building corresponding input–output flux models. Material flow analysis refers to exploring the composition of the main input and output sources of heavy metals in agricultural soil. Even though the concept was proposed early, the importance of soil heavy metal sources (balance) regulation is only gradually receiving great attention in China [25]. There are a certain number of papers which build input–output flux models for specific regions, but the information is still insufficient and scattered [26]. In addition, there are fewer papers which analyze how to control fluxes of heavy metals in detail. These research gaps suggest that a comprehensive review which collects information about the input and output sources and their corresponding contributions to different areas, and the ways of managing heavy metal sources (both input and output sources), is required to promote sustainable and large-scale heavy metal balance regulation in agricultural soils in China.

In this study, previously published papers were collected to analyze and compare the overall status of heavy metal balance regulation in agricultural soils in some representative areas in China. The specific purposes of this research were to (1) explore the contributions of input and output sources to heavy metal fluxes and predict the future trend of pollution of five heavy metals: Cd, Hg, Pb, As, and Cr in five representative regions; (2) collect information about the application of remediation technologies on long-term and large-scale heavy metal balance regulation, and explore this information in details and innovatively. The results of this study will provide useful information about how to reduce the accumulation of heavy metals in farmland in China sustainably.

2. Materials and Methods

2.1. Input–Output Fluxes and Prediction Models

Papers with keywords “soil, heavy metals, input and output, inventory, flux, balance, source apportionment” were screened. The time range was from 2013 to 2022. The papers should meet the requirements as follows. Firstly, the soil samples in these papers should be collected at a depth of 0–20 cm. Second, the papers should provide information on the experimental location and the concentrations of at least one of the most toxic heavy metals: Cd, Hg, Pb, As, and Cr. Third, the papers should originate from China. Ultimately, the papers should contain the input and output pathways of heavy metals, and the tested soils should be agricultural soils. Based on the requirements, 25 papers were selected from the Web of Science (WoS) and China National Knowledge Infrastructure (CNKI) databases for recalculation and analysis. Five representative regions, Hunan, Yangtze River Delta Region, Zhejiang, and the whole of China, were selected for analyzing heavy metal fluxes and predicting future accumulation trends of heavy metals in detail. In this study, the whole of China contains all provinces in China, including the aforementioned four provinces. The available information on other regions will also be provided and listed.

Based on the papers collected, the main input sources of heavy metals in agricultural soils include atmospheric decomposition, fertilizers (including manure), and irrigation water. The output sources consist of leaching, surface runoff, and plant uptake. The formula of the net flux of heavy metals is shown as follows [27]:

$$\begin{array}{c} I_{\mathrm{total} }=I_{atm}+I_{ir}+I_{fer}+I_{re},\end{array}$$

(1)

$$\begin{array}{c}{I}_{\mathrm{total} }=I_{\mathrm{atm} }+I_{\mathrm{ir}}+I_{\mathrm{fer} },\end{array}$$

(2)

$$\begin{array}{c} O_{\mathrm{total} }=O_{\mathrm{p} }+O_{\mathrm{lea} }+O_{\mathrm{rf} },\end{array}$$

(3)

$$\begin{array}{c}\Delta =I_{\mathrm{total} }-O_{\mathrm{total}},\end{array}$$

(4)

where $I_{\mathrm{total} }$ and $O_{\mathrm{total} }$ are the total input and output fluxes of heavy metals, respectively (g/ha/year), $I_{\mathrm{atm} }$, $I_{\mathrm{ir} }$, $I_{fer}$, and $I_{re}$ are soil input fluxes of heavy metals from atmospheric deposition, irrigation water, fertilizers, and straw returning, respectively (g/ha/year). Manures are included in fertilizers. For output fluxes, $O_{\mathrm{p}}$, $O_{\mathrm{lea} }$, and $O_{rf}$ are soil output fluxes of heavy metals from plant removal, leaching, and surface runoff, respectively (g/ha/year). Δ is the net flux of heavy metals (g/ha/year). $I_{\mathrm{re} }$ is the amount of heavy metals returned to the soil (g/ha/year) and refers to the amount of heavy metals in straw being returned to the soil. It can be calculated using the following equation [27]:

$$\begin{array}{c}{I}_{re}=O_p\times Rf,\end{array}$$

(5)

where $Rf$ is the straw return factor (%). Based on the National Report on the Comprehensive Utilization of Crop Straw released by the Ministry of Agriculture and Rural Affairs, the comprehensive utilization rate of crop straw in China reached 88.1% in 2021. Therefore, in this research, the value of $Rf$ was taken as 88% [28].

By utilizing the measured net flux values in collected papers, temporal changes in soil concentration of five heavy metal species in agricultural soils of some areas can be calculated. The equations can be written as follows [15,29]:

$$\begin{array}{c}{W}_{dry}=A_i+F_i+M_i,\end{array}$$

(6)

$$\begin{array}{c}{C}_{t,i}=C_{0,i}+t\times \Delta C_i,\end{array}$$

(7)

$$\begin{array}{c}\Delta C_i=\left[\Delta +C_{0,i}\times W_{soil }\right]/\left(W_{soil }+W_{dry }\right)C_{0,i},\end{array}$$

(8)

where $W_{dry}$ is the increased average weight of dry matter (mg/m²), $C_{t,i}$ (mg/kg) is the concentration of element i at the temporal point t, $C_{0,i}$ (mg/kg) is the initial concentration of element i, and Δ$C_i$ (mg/kg) is the change in the concentration of soil element i. $A_i$ is the annual input of dry atmospheric deposition (mg/m²), $F_i$ is the annual amount of fertilizers applied (mg/m²), $M_i$ is the annual amount of manures applied (mg/m²), and $W_{soil }$ is the weight of the soil surface layer (0–20 cm).

For calculations, the assumptions were as follows: (1) the net flux values will remain constant over time; (2) the weight of surface soil (0–20 cm) will be taken as 2.3 × 10⁸ kg/km² [30]; (3) compared with W_soil, the A_i, F_i, and M_i are relatively small, so they are negligible in this analysis. Consequently, the weight of the soil top layer (0–20 cm) is assumed to be constant; (4) all input of heavy metals from different sources occur within the soil top layer (0–20 cm).

All recalculation work included in this part was conducted using Microsoft Excel 2019 MSO (Version 2401).

2.2. Strategies for Long-Term and Large-Scale Heavy Metal Regulation

Potential strategies which can be used for diminishing the input of heavy metals and increasing the uptake of heavy metals by plants will be provided and discussed in terms of their sustainability and feasibility. Typical application cases of some remediation technologies will also be analyzed. Phytoremediation and some sustainable chemical remediation technologies will be the main focus. In this section, more than 200 studies were explored and evaluated using the Web of Science (WoS) and China National Knowledge Infrastructure (CNKI) databases. In the plant uptake subsection, 165 papers (listed in Supplementary Material S4) were used for co-occurrence network analysis and visualization. CiteSpace software (6.2.R4) was used to generate co-occurrence network figures. The node type was set to Keyword for co-occurrence network analysis, while the other parameters were set to the default values.

3. Results and Discussion

3.1. Analysis of Heavy Metal Fluxes in Representative Areas

3.1.1. Input Fluxes

The proportion of each input source to the total inputs of heavy metals in China is provided in Figure 1. If crop straws were not returned to the fields (Figure 1a), in the whole of China, atmospheric decomposition was the dominant input source of all five heavy metals, contributing 60–93%. This result was close to the range (50–93%) reported by Peng et al. (2019) [31]. For Pb and Cr, the contribution of atmospheric decomposition was 86.5% and 90%, respectively, which was significantly higher than that of Cd, As, and Hg. Fertilizers were the second dominant input source of heavy metals, ranging from 5.2% to 37%. The contributions of irrigation water to the inputs of all heavy metals were low (<3%), except for Cd where 16.5% of the total input was contributed by irrigation water.

Also, based on Peng et al. (2019) [31], the relative contribution of sewage irrigation to Cd input was 13%, which was also similar to the result of this study. If straw was returned to the fields (Figure 1b), the structure of the relative contribution of Pb remained virtually unchanged. On the other hand, for Cd, the input percentage of atmospheric decomposition reduced from 60% to 48%. In China, for the straw removal case (Equation (1)), the annual input flux order is Pb (248.43 g/ha/year) > Cr (229.49 g/ha/year) > As (38.89 g/ha/year) > Cd (8.83 g/ha/year) > Hg (1.01 g/ha/year). For the straw returning case (Equation (2)), the annual input flux order is Pb (249.57 g/ha/year) > Cr (235.03 g/ha/year) > As (40.73 g/ha/year) > Cd (10.89 g/ha/year) > Hg (1.07 g/ha/year). It can be seen that the annual input of Cd is increased by 23% if straw is returned.

In Hunan, if crop straw was removed, approximately 51–95%, 4–48%, and 0.5–13% of Pb, Cr, As, Cd, and Hg in total annual input fluxes were derived from atmospheric decomposition, fertilizers, and irrigation water, respectively (Figure 2a). The relative contributions of each input source to the total input flux of heavy metals are provided in Figure 2b. In this case, the relative contributions of Hunan’s input sources were quite similar to those of China. However, the annual average input fluxes of heavy metals in Hunan were significantly higher than those of China as a whole. In Hunan, for the straw removal case (Equation (1)), the annual input flux order is Pb (770.03 g/ha/year) > As (437.41 g/ha/year) > Cr (312.19 g/ha/year) > Cd (33.82 g/ha/year) > Hg (1.2 g/ha/year). For the straw returning case (Equation (2)), the annual input flux order is Cr (1106.02 g/ha/year) > Pb (896.38 g/ha/year) > As (603.74 g/ha/year) > Cd (72.18 g/ha/year) > Hg (1.99 g/ha/year). Hunan is famous for its high production of non-ferrous metal smelting. In this province, approximately 77% of Cd in the atmosphere arose from emissions of non-ferrous metal smelting [32]. In addition, the intensive industrial activities in Hunan also result in the accumulation of heavy metals in agricultural soils, crop residues, and main water systems, including Zijiang, Yuanjiang, and the Lishui River [33]. Among all five heavy metal species, the total annual input flux of Cr increased most significantly by 250% due to the straw returning.

In Zhejiang, the heavy metals input from atmospheric decomposition, fertilizers, and irrigation water was 29–83%, 6–34%, and 11–38%, respectively, for the straw removal case (Figure 3a). However, if the straw was returned, and the return factor was taken as 88%, atmospheric decomposition, fertilizers, irrigation water, and crop straws were responsible for 26–77%, 5–30%, 10–34%, and 8–37% of the total input fluxes, respectively (Figure 3b). For the straw removal case (Equation (1)), the annual input flux order in Zhejiang was Pb (258.67 g/ha/year) > Cr (204.96 g/ha/year) > As (35.33 g/ha/year) > Cd (10.46 g/ha/year) > Hg(1.7 g/ha/year). For the straw returning case (Equation (2)), the annual input flux order was Pb (278.36 g/ha/year) > Cr (224.33 g/ha/year) > As (40.01 g/ha/year) > Cd (15.67 g/ha/year) > Hg (2.18 g/ha/year). Except for Cd (only for the straw returning case) and As, atmospheric decomposition was still the dominant input source of these heavy metal species.

Zhejiang is an important part of the Yangtze River Delta, which is a famous economic zone in China. When compared to most provinces in China, industrial activities in Zhejiang are frequent. In addition, it is well-known for its high-yield areas like Hangjiahu Plain and Ningshao Plain, where agricultural activities are very intensive [34]. Ultimately, the background values of soil heavy metals in Zhejiang were much higher than those of China as a whole [29]. This means that in Zhejiang, it would take less time to raise soil heavy metal concentrations from background values of topsoil to the maximum permissible limits.

The Yangtze River Delta is a comprehensive economic zone and consists of Zhejiang, Jiangsu provinces, and Shanghai. In this region, overall, the heavy metal pollution in Zhejiang was not as serious as in the other two areas (Jiangsu and Shanghai) of the Yangtze River Delta. For Cd and Pb, most pollution cases were located in Lin’an and Fuyang and the boundaries of Zhejiang, Jiangsu, and Shanghai [35]. The relative contributions of input sources to heavy metal fluxes are shown in Figure 4a (straw removal) and Figure 4b (straw returning). In both straw removal and straw returning cases, rather than atmospheric decomposition, irrigation water was the most essential input source of As, Cd, and Hg, the relative contributions of which all exceeded 56%. In the Yangtze River Delta, for the straw removal case (Equation (1)), the annual input flux order was Pb (186.72 g/ha/year) > Cr (156.55 g/ha/year) > As (47.48 g/ha/year) > Cd(8.42 g/ha/year) > Hg (0.55 g/ha/year). For the straw returning case (Equation (2)), the annual input flux order was Pb (186.34 g/ha/year) > Cr (157.78 g/ha/year) > As (48.32 g/ha/year) > Cd (8.82 g/ha/year) > Hg (0.59 g/ha/year).

Hainan is one of the main rice production areas in China. Based on the National Bureau of Statistics of China (2022) [36], the annual rice yield of Hainan was approximately 1.27 million tons, while the rice production area reached 2.29 × 10⁵ hectares. In addition, it is rich in biodiversity and has abundant germplasm resources of rice and some other commercial crops like sugarcane [37].

For Hainan, the atmospheric decomposition, irrigation water, fertilizers, and straw contributions to Pb, Cr, Cd, As, and Hg are provided in Figure 5a (straw removal) and Figure 5b (straw returning). In this place, for As and Hg, irrigation water was the most important source. The contributions of irrigation water in both straw removal and straw returning cases reached 96%. Atmospheric decomposition and fertilizers only accounted for 0.05–16% and 1–2% of As and Hg inputs, respectively. For Cd, atmospheric decomposition was also not the dominant input source; instead, fertilizers were the predominant pollution source in both cases. For Cr, total annual input is increased by 112% if straw is returned to fields. In Hainan, for the straw removal case (Equation (1)), the annual input flux order was As (481.3 g/ha/year) > Pb (167.7 g/ha/year) > Cr (133 g/ha/year) > Hg (30.55 g/ha/year) > Cd (5.21 g/ha/year). For the straw returning case (Equation (2)), the annual input flux order was As (487.02 g/ha/year) > Cr (282.6 g/ha/year) > Pb (221.38 g/ha/year) > Hg (30.66 g/ha/year) > Cd (5.77 g/ha/year).

3.1.2. Output Fluxes

The relative contributions of output sources to total output fluxes of heavy metals in the Yangtze River Delta and Hunan were calculated using Equation (3) and are shown in Figure 6a,b. For the Yangtze River Delta, leaching was the major output pathway of all heavy metals, and its relative contributions ranged from 71% to 98%. In this place, data on surface runoff was not available. In Hunan, plant removal was the major output source of all five heavy metals, which accounted for 87% to approximately 100% of total output fluxes. In the Yangtze River Delta, the annual output flux order was Pb (44.67 g/ha/year) > Cr (34.68 g/ha/year) > As (20.56 g/ha/year) > Cd (1.56 g/ha/year) > Hg (0.251 g/ha/year). In Hunan, the annual output flux order was Cr (915.35 g/ha/year) > As (214.98 g/ha/year) > Pb (183.43 g/ha/year) > Cd (44.14 g/ha/year) > Hg (0.9 g/ha/year).

For the other three regions, including Hainan, Zhejiang, and the whole of China, the only output data available were for plant removal, the details of which are provided in Table S1.

3.1.3. Net Fluxes and Prediction of Accumulation

The net fluxes of China, Hunan, Zhejiang, the Yangtze River Delta, and Hainan were calculated based on Equation (8) and are shown in Section 2.1 and provided in

Table 1. Net fluxes of heavy metals in representative areas (g/ha/year).

Source	Elements	China	Hunan	Zhejiang	Yangtze River Delta	Hainan	Elements
Balance	Straw removal	Pb	247.13327	586.6057	236.28527	141.05	106.7
		Cr	223.20229	603.1575	182.94731	121.87	−3.7
		Cd	6.4834825	10.64694	4.5475813	6.86	0.41
		As	36.789913	222.424	29.912599	26.92	44.5
		Hg	0.9444088	0.29388	1.1661539	0.299	3
Balance	Straw return	Pb	248.27603	712.9502	255.98173	141.6748	160.38
		Cr	228.73893	190.6696	202.31951	123.102	112.6
		Cd	8.5467546	27.71675	9.7529278	7.256	5.1332
		As	38.635851	388.757	34.681575	27.756	480.52
		Hg	1.0014687	1.08558	1.638969	0.34124	30.5356

. Also, the current increase rates of heavy metal concentrations in agricultural soils were utilized to predict the time required for rising soil heavy metal concentrations from the current average values (C_0,i) to maximum permissible limits. The results are shown in

Table 2. The predicted time in which heavy metal concentrations accumulated to the limit values in China.

China	Pb	Cr	Cd	As	Hg
Average values of topsoil (mg/kg) a	35.24	72.59	0.29	12.98	0.29
Background values (mg/kg) b	26	61	0.097	11.2	0.065
Maximum permissible limits (mg/kg) c,*	90	150	0.3	40	1.8
Time required (straw removal) (year)	510	798	4	1689	3678
Time required (straw return) (year)	507	778	3	1608	3468

^a Ren et al., 2022 [9]. ^b Wei et al., 1999 [38]. ^c Feng et al., 2020 [39]. * Maximum permissible limits are equivalent to the risk screening values (RSVs) for heavy metals. When concentrations of heavy metals exceed the RIVs, soil pollution and problems of food safety may occur, so specific investigations need to be conducted [40].

,

Table 3. The predicted time in which heavy metal concentrations accumulated to the limit values in Hunan.

Hunan	Pb	Cr	Cd	As	Hg
Average values of topsoil (mg/kg) a	42.78	59.97	0.381	13.39	0.201
Background values (mg/kg) a	29.7	71.4	0.126	15.7	0.116
Maximum permissible limits (mg/kg)	90	150	0.3	40	1.8
Time required (straw removal) (year)	185	/ b	/	275	12514
Time required (straw return) (year)	152	1086	/	157	3388

^a Chen et al., 2020 [41]. ^b “/” indicates that the average value of topsoil in this place has already exceeded the maximum permissible limits.

,

Table 4. The predicted time in which heavy metal concentrations accumulated to the limit values in Zhejiang.

Zhejiang	Pb	Cr	Cd	As	Hg
Average values of topsoil (mg/kg) a	47.14	47.42	0.23	7.27	0.12
Background values (mg/kg) b	33.14	67.29	0.2	8.47	0.13
Maximum permissible limits (mg/kg)	90	150	0.3	40	1.8
Time required (straw removal) (year)	417	1290	36	2517	3313
Time required (straw return) (year)	385	1166	16	2170	2358

^a Zhang et al., 2019 [42]. ^b Shi et al., 2018 [29].

,

Table 5. The predicted time in which heavy metal concentrations accumulated to the limit values in the Yangtze River Delta.

Yangtze River Delta	Pb	Cr	Cd	As	Hg
Average values of topsoil (mg/kg) a	37.63	74.52	0.226	7.8	0.14
Background values (mg/kg) b	24.3	70.94	0.14	8	0.16
Maximum permissible limits (mg/kg)	90	150	0.3	40	1.8
Time required (straw removal) (year)	854	1425	25	2751	12769
Time required (straw return) (year)	850	1410	23	2668	11189

^a Chen et al., 2020 [41]. ^b Xiao et al. 2010 [43].

and

Table 6. The predicted time in which heavy metal concentrations accumulated to the limit values in Hainan.

Hainan	Pb	Cr	Cd	As	Hg
Average values of topsoil (mg/kg) a	33.62	46.44	0.288	9.87	0.08
Background values (mg/kg) b	24	22	0.061	2	0.03
Maximum permissible limits (mg/kg)	90	150	0.3	40	1.8
Time required (straw removal) (year)	1215	/	67	1557	1319
Time required (straw return) (year)	808	2115	5	144	130

^a Chen et al., 2020 [41]. ^b Jiang et al., 2014 [44].

.

Based on the results in Table 2, Table 3, Table 4, Table 5 and Table 6, Cd is the heavy metal species of the most concern for agricultural soils in all five areas, followed by Pb and Cr. The average Cd concentrations in five sites have either exceeded or approached the maximum permissible value (0.3 mg/kg). The estimated time for increasing average top values to the maximum limits of heavy metals is 3–67 years. Straw return influences the increment rates of all heavy metals, especially for Cd and Hg fluxes in Hainan. Therefore, it is better to decrease the percentage of straw returned to agricultural soils in this place.

The available information about the balance of Cd in other regions in China is listed in Table S2. Information about other heavy metal species is not sufficient enough, so it is not provided in this study.

3.1.4. Limitations and Prospects of the Balance Model

The limitations of the heavy metal balance model used in this study can be summarized as follows. Firstly, the number of papers which were used in recalculating the net fluxes of heavy metals was limited. There were a few papers which included the balance model, while most of them focused on the balance of heavy metals in small-scale agricultural soils located near special areas (e.g., industrial areas). This kind of agricultural soil is not the majority of agricultural soils in most places and is unrepresentative. Therefore, many papers were excluded during the paper screening step.

Secondly, while considering plant (mainly straw) removal as a main output source of heavy metals, most researchers did not consider the straw returning case. In addition, even if the aboveground parts of plants were removed, the roots were still left in the soil. For most plants, the concentrations of heavy metals in roots were higher than in aboveground parts [45]. Therefore, overall, the relative contributions of plant removal to output fluxes were overestimated. On the contrary, the relative contributions of surface runoff and leaching were somewhat underestimated. In many studies, either leaching or surface runoff was neglected. However, as was shown in Figure 5 and Table 6, the contributions of leaching and surface runoff to total output fluxes of heavy metals in some places were significantly higher than plant removals.

Consequently, more long-term studies should be conducted to use the heavy metal balance model to measure input and output fluxes of large-scale agricultural soils. The parameters included in the model should be updated based on local conditions.

3.2. Heavy Metal Balance Regulation

In this part, how to control dominant input and output sources of heavy metals in agricultural soils will be described separately. The input sections include atmospheric decomposition, fertilizers (including manure), and irrigation water. The output section is the plant uptake. In each section, instead of listing all available remediation technologies, sustainable and advanced technologies which can be used to control the heavy metal fluxes will be analyzed and compared. The application of phytoremediation technologies to reduce the input or increase the output of heavy metals will be an important topic in this part. The phytoremediation technologies include hyperaccumulators and non-hyperaccumulator large-mass plants. The cases of replacement of flooding by drip irrigation and reduction of utilization of fertilizers, which are rich in heavy metals, will also be mentioned.

3.2.1. Atmospheric Decomposition

Unfortunately, there are no specialized strategies for controlling the input of soil heavy metals from atmospheric decomposition into agricultural soil. Atmospheric decomposition is a sophisticated non-point source of pollution which is difficult to monitor and control. The heavy metals in the atmosphere originate from diverse anthropogenic activities like non-ferrous metal smelting, coal-fired emissions from thermal power plants, transportation discharge, incineration of wastes, etc., [46,47,48]. For individual heavy metal species, Cr in the atmosphere is mainly derived from soil particles (soil parent material), Pb is mainly from the combustion of coals and emissions of transportation and construction industries, and Cd initiates from industrial emissions [49,50,51]. The only sustainable way to control the input of soil heavy metals from atmospheric decomposition is to control or even directly eliminate the sources of atmospheric pollution. Therefore, strategies such as controlling the discharge of waste gas and wastewater from industries around agricultural fields, reducing the use of poor-quality oil products by farmers, and mitigating the combustion of fossil fuels are important. However, due to the rapid industrialization and current energy structure (coal-based and energy-consuming) of China, the process of improvement will take a long time [52].

On the other hand, for heavy metal elements released from traffic emissions, cultivating green plants and building greenbelts with high tolerance and enrichment ability for heavy metals around roads is also an effective controlling method. In order to test the impact of green belts on the diffusion patterns of heavy metals Pb and Cr, Wang et al. (2010) [53] collected soil samples in some highway greenbelts in Shanxi Province. The results showed that firstly, the highway greenbelts had a significant effect on remediating soil pollution caused by heavy metals Pb and Cr. The tested greenbelts could effectively control soil pollution within a range of 50 m. Secondly, the width of green belts was closely related to their protective effects. A width of 40–70 m can provide good protection against soil heavy metal pollution caused by traffic emissions.

3.2.2. Irrigation Water

Based on the outcomes of the previous part, the overall contribution of irrigation water to heavy metal input fluxes in agricultural soil is not high. Nevertheless, in areas where crops are irrigated using wastewater, irrigation water can be one of the dominant input sources of heavy metals [54]. When compared to atmospheric deposition, heavy metal pollution from irrigation water is easier to control and can be treated in the following ways: first, to remediate polluted water, aquatic plants, which are either hyperaccumulators or biomass plants, are promising options [55,56]. More details of hyperaccumulators and biomass plants will be provided in the plant uptake section. Moreover, traditional chemical remediation technologies are mostly unsustainable since they require constant input of materials and may cause secondary pollution [57]. However, the introduction of clean materials and the improvement of the production process give rise to the presence of advanced and sustainable chemical remediation technologies. For example, biochar made using invasive species and biochar produced by introducing fallen leaves and natural bioaugmentation (decay) pretreatment can certainly treat polluted irrigation water while avoiding secondary pollution [58,59,60].

The second strategy is to directly replace the water source of irrigation water. According to the Standard for Irrigation Water Quality (GB 5084-2021) [61], for irrigation water (pH 5.5–8.5), the total concentration of Cd, Pb, Hg, Cr, and As, should not exceed 0.01, 0.2, 0.001, 0.1, and 0.1, respectively. Before the heavy metal content in sewage is reduced to the specified standard, other water sources should be selected for irrigation. The third solution is to adopt water-saving irrigation methods such as drip irrigation to replace traditional flooded irrigation. Drip irrigation, supplemented by natural rainfall, can reduce irrigation water consumption by more than 50%, thereby decreasing the input of heavy metals [62,63]. The wide application of drip irrigation also brings benefits such as effectively improving the utilization rate of rainfall resources, reducing the occurrence of diseases and pests, and improving the efficiency of fertilizer utilization.

3.2.3. Fertilizers (Including Manure)

The concentrations of heavy metals in most fertilizers, especially inorganic fertilizers, meet the limit of determination of arsenic, cadmium, chromium, lead, and mercury contents for fertilizers (GB/T 23349-2020) (State Administration for Market Regulation & Standardization Administration, 2021) [64]; however, the continuous and excessive application of fertilizers can still trigger heavy metal accumulations in some areas [65]. For example, based on Hu et al. (2013) [66], in a peri-urban vegetable farm near Nanjing, the average input contribution of Cd and Pb by fertilizers was approximately 56.0% and 43.7%, respectively. The population in China is large, so the demand for food is high. While the area of arable land is limited, it is difficult to control heavy metal pollution by completely stopping the use of fertilizers [67]. Therefore, it is important to develop authoritative theories to guide farmers to optimize the use of fertilizers and thus reduce the input of heavy metals by fertilizers [68].

By optimizing the application rates of fertilizers, the amount of heavy metals input can be reduced while crop yields can be maintained and even increased. Yin et al. (2021) [69] proposed a long-term steady-state N balance (SSNB) approach to optimize the N application range in 3824 counties for wheat, rice, and maize in China. The final outcomes indicated that when compared to traditional practices, the SSNB approach could decrease N fertilizer use by 21 to 28% while increasing yields by 6 to 7%. In addition, the Chinese government has encouraged each province to have their own standard limits for the application amount of fertilizers for different crops. On 7 April 2020, Zhejiang Province released the document “Limit Standard for Fertilizer Quota System for Major Crops” (DB330185/T 005-2020), which set the maximum amount of fertilizers for nine crops, including single cropping rice, mini sweet potato, eggplant, gourd, tea tree, peach tree, grape, bamboo, and hickory (Ministry of Agriculture and Rural Affairs of the People’s Republic of China, 2020) [70]. This is the first local standard for limited fertilizer application in Zhejiang Province and even the whole country, and it was officially implemented on 14 April. Consequently, more authoritative standards will be released to help farmers prevent overusing fertilizers.

To reduce the input of heavy metals, more attention needs to be paid to the application of fertilizers, which contain more heavy metals. In most cases, phosphorus (P) fertilizers and compound fertilizers contain significantly higher concentrations of heavy metals than nitrogen (N) and potassium (K) fertilizers [31]. Phosphorous rock, as a raw material of P fertilizers contains a certain concentration of Cd [71]. Compound fertilizers include organic fertilizers such as manure, crop residues like rice and wheat straw, sewage sludge, and solid wastes [72], which include a higher content of heavy metals than P fertilizers. The two most commonly used organic fertilizers are manure and crop residues. For manure, the concentrations of heavy metals are positively correlated to livestock feeds [73]. The quality of feed needs to be monitored more carefully by enhancing China’s feed standards; it is better to feed rare poultry on agricultural lands with natural food [74]. For crop residues, straw returning is an effective and sustainable method to improve soil fertility; however, in agricultural land where soil is moderately or highly polluted by heavy metals, the return of straw can deteriorate pollution [75].

Therefore, whether plant residues can be reused or not, how plant residues can be reused to prevent the accumulation of heavy metals should follow guidelines provided by reliable research outcomes or institutions. Environmental quality standards for agricultural land soil (GB15618-2018) [76] state that in soil with pH values higher than 7.5 and Cd concentration higher than 0.8 mg/kg, the return of straw should be forbidden.

To reduce the use of fertilizers, especially organic fertilizers, the improvement of soil fertility in sustainable ways is an important topic. In China, distinctive ecological methods have been developed. For paddy soils, raising aquatic animals like fish, shrimp, crabs, mud fish, turtles, frogs, and eels has gradually become popular in many areas. For dry land, as is mentioned in the previous paragraph, rearing poultry like chickens in open farmland can increase soil fertility while decreasing the heavy metals content in poultry manure. For these innovative methods, there are lots of practical cases and news stories in China; however, the number of journal papers is still not sufficient. The number of journal papers for these innovative methods in the last 10 years (2014–2023) is presented in Figure 7 below. There are 512 journal papers about raising fish in paddy fields, while there is only 1 journal paper focusing on raising earthworms in dry fields.

3.2.4. Plant Uptake

Plant uptake (phytoremediation) is one of the important remediation strategies for heavy metal pollution. This section will mention the concepts of hyperaccumulators and large-biomass plants and analyze how these plants can be applied to long-term and large-scale heavy metals balance regulation. In addition, crop rotation and its role as an important stage of phytoremediation will also be analyzed.

• Hyperaccumulator and biomass (non-hyperaccumulator) plants

Hyperaccumulators refer to plant varieties that can accumulate a greater amount of heavy metals in their aboveground parts compared to normal plants without affecting physiological activities [77]. This type of plant has specificity in the absorption of heavy metals and can be divided into Cd hyperaccumulators, As hyperaccumulators, Zn hyperaccumulators, etc., [78]. Non-hyperaccumulators and large biomass plants can absorb much more heavy metals than normal plants because they are better biomass producers and grow faster [79]. In addition, large biomass plants do not have specificity in species of heavy metals absorbed.

When compared to non-hyperaccumulator and biomass plants, hyperaccumulators have been intensively explored [80]. In 1976, Jaffre et al. [81] proposed the term ‘hyperaccumulator’ for the first time; since then, more than 500 hyperaccumulator species have been reported [82]. However, in practical cases, some intrinsic characteristics of hyperaccumulators are still problems and have brought significant limitations to phytoremediation technology. Hyperaccumulators are mostly endemic plants which are limited by geographical factors like local weather and soil environment [83]. In addition, the inability to grow throughout the year, low biomass and shallow root systems, and slow growth rate are all issues that can result in low efficiency in the extraction of heavy metals [84]. In fact, even though more than 400 hyperaccumulator plant species were found, most of them did not perform well in long-term and large-scale soil heavy metals balance regulation [85].

However, hyperaccumulators like Pteris vittate, amaranth grain, and ryegrass have relatively large biomass, rapid growth rate, and strong adaptability [86,87,88]. These hyperaccumulators have the potential to be used in agricultural fields, and more hyperaccumulator species like them should be explored.

In addition to hyperaccumulators, some non-hyperaccumulator and large biomass plants, which are moderate accumulators, are also potential options for soil remediation in fields. Ebbs et al. (1997) [89] mentioned that non-hyperaccumulator and large biomass plants have the advantages of fast growth rate, large biomass, and high total output of heavy metals when compared to most hyperaccumulators. Unfortunately, until now, there has been less research on large biomass non-hyperaccumulator plants. Therefore, exploring remediation plants with large biomass and a strong ability to accumulate heavy metals is important for plant remediation technologies towards long-term and large-scale applications.

Some large biomass ornamental plants that have been used successfully in soil heavy metals restoration are listed in

Table 7. Typical examples of high-biomass non-hyperaccumulator plants.

Ornamental Plant Species	Sources
Bamboo (e.g., moso bamboo and lei bamboo)	[91,92,93]
Poplar (e.g., Eastern cottonwood, Populus deltoides)	[94,95]
Willow (e.g., Salix miyabeana)	[96]
Eucalypti (e.g., Eucalyptus camaldulensis)	[97]
Morus (e.g., Morus alba)	[97,98]
Maple (e.g., Acer cappadocicum)	[99]
Camphor (e.g., Cinnamomum camphora)	[100]

. Large biomass and ornamental plants are more promising options for remediation plants than other plants because the heavy metals accumulated in these plants will not enter the human food chain and harm people [90].

Ultimately, for phytoremediation, the disposal of plants that accumulate heavy metals is also an important issue. Conventional methods mainly focus on diminishing or even eliminating the presence of heavy metals in plant materials, while advanced methods aim to reuse plant materials and obtain economic benefits [101]. Conventional methods mainly include heat treatment methods like incineration, gasification, hydrolysis, microbial methods like composting, and compression landfills. Advanced methods consist of biomining, synthesis of nanomaterials, etc., [102,103]. When compared with advanced methods, most conventional methods are easier to manipulate; however, high costs and secondary pollution caused by the discharge of effluent and waste gas are two problems. Advanced methods take advantage of high resource utilization efficiency and environmental benefits. On the other hand, they are not mature enough and have high costs [101].

• Rotation, intercropping, and relay intercropping of plants

Due to the large population in China, there is a trade-off between population demand and the limited productivity of arable lands and crops. In addition, phytoremediation of heavy metal pollution in farmland soil is a sustainable but long-term process, and most of the plants used for remediation cannot generate any benefits for local farmers. Therefore, it is better and practical to implement soil heavy metal remediation and crop cultivation works together. Intercropping, rotation, and relay intercropping main crops with remediation plants or other crops can not only significantly improve the utilization efficiency of natural resources and crop yield but also increase the efficiency of soil heavy metals remediation.

By exploring the Web of Science (WoS) database, 90 papers with the keywords ‘soil heavy metal’ and ‘intercropping’ and 75 papers with the keywords ‘soil heavy metal’ and ‘crop rotation’ were extracted for keyword co-occurrence analysis (Figure 8 and Figure 9). It can be seen that Cd is the heavy metal species that is most concerning. Maize and rice are two crops which are most commonly used in phytoremediation experiments.

4. Conclusions

In this study, the current status of heavy metal pollution in agricultural soils in China was explored. The input–output flux model was used to recalculate the contributions of main input and output sources to net fluxes of heavy metals and predict the future trend of heavy metal pollution in five representative areas: China, Hunan, Zhejiang, the Yangtze River Delta, and Hainan. The heavy metal species included were Cd, Hg, Cr, Pb, and As. In addition, the strategies which can either reduce input or increase output of heavy metals were discussed. Overall, key points and prospects are summarized as follows:

(1) Based on results generated by the input and output flux model, except Hainan, atmospheric decomposition was one of the most important input sources of all five heavy metals in all representative areas. In Hainan, the contributions of atmospheric decomposition to As and Hg were approximately zero. The relative contributions of fertilizers, irrigation water, and straw returning varied greatly among different areas. In the Yangtze River Delta, irrigation water was the primary input source of Cd, As, and Hg; in Hainan, irrigation water was the dominant input source of As and Hg. On the other hand, in Hunan and China, the relative contributions of irrigation water to all five heavy metals were negligible. In Hunan, under the straw returning case, straw returning was the primary input source of Cr and Cd, and its contributions to the input of As, Hg, and Pb were significant (14–40%). In Hainan, fertilizers (including manures) were the most important input source of Cd in both straw returning and straw removal cases. Nevertheless, the contributions of fertilizers to inputs of all five heavy metals in the Yangtze River Delta were insignificant. For output fluxes, leaching was the primary output source of all heavy metals in the Yangtze River Delta; in Hunan, plant removal was the dominant output source. For the other three sites, leaching and surface runoff were neglected during calculations.

(2) The extent of heavy metal pollution varied among different areas. According to the outcomes of future trend predictions, Cd pollution is the most serious in all five representative areas, followed by Pb and Cr. Cd pollution in Hunan is now serious and needs to be controlled efficiently. The results also showed that returning polluted straw could shorten the time required to raise soil heavy metal concentrations from average values of topsoil to the maximum permissible limits significantly. Therefore, not returning contaminated straw to the soil can be an effective strategy to reduce the accumulation of heavy metals in agricultural soils.

(3) For the strategies utilized in heavy metal input and output fluxes regulation, due to the complex composition of sources, there are no single methods which can reduce the input of heavy metals from atmospheric decomposition. For fertilizers, irrigation water, and straw returning, China began to formulate a series of policies to manage the quality and application of them, including the Agricultural Irrigation Water Quota for different provinces, Standards for Irrigation Water Quality (GB 5084-2021) [61], Determination of Arsenic, Cadmium, Chromium, Lead, and Mercury Contents for Fertilizers (GB/T 23349-2020) [64], Limit Standard for Fertilizer Quota System for Major Crops (DB330185/T 005-2020) [70], and Environmental Quality Standards for Agricultural Land Soil (GB15618-2018) [76].

(4) For plant uptake, the development of phytoremediation technologies is remarkable; however, the application cases of high-biomass non-hyperaccumulators are still insufficient. Hyperaccumulators have drawbacks such as the inability to grow throughout the year, low aboveground biomass, and slow growth rate. Therefore, it is recommended to explore and screen more high-biomass non-hyperaccumulators, which can be used with other technologies to increase the output of heavy metals from agricultural soil.

(5) In China, the areas of arable land are limited while the population continues to increase. Therefore, intercropping, relay intercropping, and rotating crops with hyperaccumulators and/or high-biomass plants are feasible ways of applying phytoremediation technologies on large-scale, long-term heavy metal balance regulation. Therefore, papers on related topics were extracted for keyword co-occurrence network analysis. For both intercropping and crop rotation cases, Cd was still the most concerning heavy metal species. In the meantime, Pteris vittata L., Thlaspi caerulescens, and Brassica juncea were considered the three most commonly used hyperaccumulators in China. Maize and rice were two crops used frequently in field phytoremediation experiments.