Assessment of Potential Heavy Metal Contamination in the Agricultural 1 Soils based on various improved evaluation methods in Beijing ， China 2

15 The evaluation of the soil contaminated by heavy metals can help to judge whether the soil meets the 16 standard and whether the pollution will threaten human health and the ecological environment. In this study, the 17 farmland soil from eight districts in Beijing was used as the research object, and the concentration of heavy 18 metal elements, Pb, As and Cd in the soils and agricultural products were analyzed. The analysis results showed 19 that: (1) The evaluation based on the improved Hakanson method suggested that the crops exhibit a significantly 20 higher ability to absorb Cd than to absorb Pb and As. Pb, As and Cd are all at normal level of ecological risk; 21 among them, Cd is mainly in a moderate ecological risk, without strong ecological risk. (2) Based on the 22 Improved analytic hierarchy process(AHP) of evaluation, 0.2317 is the average value of the integrated index of 23 heavy metal pollution of soil in the study area, which is a mild level of pollution. (3) Through the calculation of 24 various parameters in the Influence index of comprehensive quality(IICQ) of soil and agricultural products, it 25 was found that 0 ＜ IICQS ＜ 1, suggesting that the environmental quality of soil is at a clean level. In summary, 26 the pollution of heavy metals Pb, As and Cd in the farmland soils and crops in the eight districts of Beijing, 27 including Fangshan, Daxing, Shunyi, and Shijingshan is at a low level, and no significant impact has been 28 brought to the surrounding environment. 29


31
In recent years, the rapid economic and industrial growth in China has triggered some prominent 32 environmental problems, especially the serious industrial heavy metal pollution to a bulk amount of arable land. 33 Heavy metal pollution of arable land in China has seriously harmed people's health and agricultural economic 34 development (Fu et al. 2020). Most soils contaminated by heavy metals contain various toxic metal elements, 35 which will not only reduce crop yields, but also threaten food safety after penetrating the crops (Niu et al. 2013). 36 Surveys showed that more than 70% of the arable soil in China has been contaminated by heavy metals, 37 accounting for 1/6 of China's total area of cultivated soil. Therefore, it is not hard to see that the agricultural land 38 in China is suffering severe heavy metal pollution . 39 Meanwhile, One of the main sources of heavy metal contamination in crops is heavy metals in the soil.

100
The Analytic Hierarchy Process (AHP) was proposed by the famous American operations researcher Satty T. 101 L in the late 1970s (Saaty 1990). It is an unstructured multi-criteria decision-making method that combines 102 qualitative analysis and quantitative analysis to bring people's thinking. The process is hierarchical and 103 quantified, which is especially suitable for situations where the target structure is complex. The improved 104 analytic hierarchy process comprehensively considers the nature of the arable land, and adds the limit value of heavy metals in crops ("Limits of Contaminants in Food" GB2762-2017). At the same time, the toxicity response 106 coefficient of heavy metals (Hakanson, 1980, Liu et al. 2019) and the influence of heavy metals on crops planted 107 on cultivated soil are used as the basis for the application of the analytic hierarchy process. 108 The first layer of the improved analytic hierarchy process structure is the target layer, that is, the heavy 109 metal pollution evaluation of the soil in a certain area, defined as layer A; the second layer is the standard layer, 110 that is, the selected three heavy metals of Pb, As, and Cd are in the crops Limit values (respectively 0.2 • 111 −1 , 0.2 • −1 and 0.5 • −1 ) and heavy metal toxicity response coefficient (respectively 1, 20, 10) 112 two standards, defined as B level, of which the two standards are defined as B1, B2; The third layer is the 113 specific evaluation factors, namely three heavy metal elements of lead, cadmium, and arsenic, which are defined 114 as layer C; this layer analysis is a three-layer structure type (Chen et (Table 1). 128

Soil and Agricultural Products Influence index of comprehensive quality
Through the pollutant category-absorption coefficient correction method, the index of the Hakanson method 129 was improved, and the limits of grading standard and are shown in Table 2. cabbage, cabbage and alfalfa), and reported that Cd and Pb accumulate more in the root system. The conclusions 151 of this paper are consistent with the above results, indicating that Cd in the soil is more easily absorbed by crops. 152

153
In the evaluation results obtained by the improved correction method, the average value of is 154 (30.67)> (10.59)> (1.02). Pb, Cr and As are all at normal level of ecological risk; the proportion of low, 155 moderate and strong ecological risk of Cd is (3.1%) and (96.9%) respectively, suggesting that Cd is mainly in 156 the moderate ecological risk without strong ecological risk. The contribution of , and to RI is 157 72.54%, 25.06% and 2.4% respectively. It can be seen that have the largest contribution. 158 The study showed that the overall background value of As in Beijing is small, and only point source 159 pollution characteristics have been found, thus the pollution risk of As is at a low level. Although the surface 160 pollution characteristics of Cd and Pb have been observed, Pb exhibits excellent performance in soils in Beijing, 161 with a normal level of ecological pollution risk; the environmental quality of Cd element is generally good, and 162 light pollution exists in some areas. 163 can be analyzed to infer whether the regional sources of heavy metal pollution are the same. If there is a 230 significant correlation between the content of several heavy metals, it indicates high homology, otherwise, the 231 sources may be complicated and interfered by various factors. 232

233
Spearman correlation analysis was conducted on three kinds of heavy metals in farmland soils in Beijing, 234 and the results are shown in Table 10. 235 It can be seen that Pb, As, and Cd in farmland soils in Beijing are significantly correlated with each other, 236 with the level of P<0.05. Considering that there may be differences in the accumulation of heavy metal elements 237 in the soils in different areas, the correlation analysis was further performed on the heavy metals in the soils in 238 different districts. 239 Table 8 shows that the significant correlation between the three kinds of heavy metals is roughly the same 240 in different districts of Beijing, but the correlation intensity is different. There is a significant correlation between 241 Pb and Cd in all districts, a very strong significant correlation has been found in Fangshan, Daxing, Mentougou, 242 and Miyun, and a significant correlation is presented in Shunyi, Shijingshan, and Changping; significant 243 correlation between Pb and As is shown in some districts, and extremely strong significant correlation has been 244 observed in Mentougou, Changping, and Miyun; Cd and As are weakly correlated or not correlated in most 245

districts. 246
In general, the correlation between Cr and As is relatively not strong in 7 districts of Beijing, but the strong correlation has been found in Mentougou. There is a correlation between Pb and Cd, and between Pb and As. 248 The correlation between heavy metals is relatively strong in the soils in Daxing, Shijingshan, and Changping, 249 while relatively weak correlation exists in the soils in Fangshan, Shunyi, and Huairou. 250

251
In order to better understand the relationship between the three kinds of heavy metal elements in the soils in 252 Beijing and explore the correlation and homology of the data, SPSS statistics 26.0 software was used to perform 253 principal component analysis on the content of seven heavy metals in 64 soil samples. It can be seen from 4.1 254 that there is a linear correlation between the heavy metals. KMO and Bartlett tests were conducted on the data. 255 The coefficient of KMO test is 0.604>0.600, and the Bartlett test result is P=0.000<0.001. Therefore, the data 256 structure can be extracted for principal components. The varimax method was employed to rotate the data, so 257 that the component factors can be better explained. The results of principal component analysis are shown in 258 Table 12. In this paper, the components with eigenvalues greater than 1 are selected as principal components, and 259 two principal components are extracted, which can explain 86.2% of the total variance. 260 Figure 9 and Table 9 show that component 1 mainly reflects the composition information of Pb and Cd, and 261 these two heavy metals also have a strong correlation. Cd and Pb almost all exceed the background value at the 262 point location, thus it is speculated that the sources of these two elements in the research area are mainly the 263 human activities such as agricultural fertilizers, sewage irrigation, and industrial production. 264 Component 2 mainly reflects the information of As. As has a weak correlation with other metal elements. 265 As is a diagenetic element, and its accumulation in soil comes from both natural and artificial sources. The 266 artificial sources involve the use of agrochemical products, mining, and industry. parameters, and the results showed that 0＜IICQS＜1 in 8 districts of Beijing, suggesting the environmental 281 quality of soil is at a clean level. The IICQ method comprehensively considers the interaction between heavy 282 metals in soils and agricultural products and the environmental quality of farmland soils. It is more applicable to 283 the evaluation of the combined and individual influence of heavy metals in soil, and this method takes the 284 valence state effect of heavy metals into full consideration, so that it can effectively analyze the effective state of 285 heavy metals. In addition, through source analysis of farmland soils in Beijing's eight districts, it was found that 286 the pollution mainly comes from artificial sources such as agricultural fertilizers and irrigation. Therefore, it is 287 necessary to strengthen the management of agricultural fertilizers and sewage irrigation. 288 In summary, the quality of urban agricultural soil is vital to human health. In order to effectively reduce the risk of heavy metal pollution in urban agricultural area and further develop reliable protection measures, risk 290 evaluation is required. In the evaluation, different evaluation methods should be adopted according to different 291 soil pollutants, different valence states of heavy metals and different pollution sources, so that the potential 292 environmental risk of farmland soil in different areas can be accurately, comprehensively and objectively 293 evaluated and predicted. 294