Characterization and Source Analysis of Heavy Metal(loid)s Pollution in Soil of an Industrial Park in Kunming, China

Abstract

This study investigated the characteristics and sources of heavy metal(loid) pollution in the soil of a key industrial park in Kunming, China. In total, 60 soil samples (40 from agricultural land and 20 from construction land) were collected from and around the park. The soil pH and contents of Arsenic (As), lead (Pb), copper (Cu), zinc (Zn), cadmium (Cd), mercury (Hg), nickel (Ni), and chromium (Cr) were measured. The contents of the eight heavy metal(loid)s were analyzed using the background values of heavy metal(loid)s in the Kunming soil. The pollution load, geoaccumulation, and Nemero Comprehensive Pollution Indices were used for environmental risk evaluation. Cluster and principal component analyses were used to resolve heavy metal(loid) sources. Cd was enriched in construction and agricultural soils. As, Hg, Cd, Pb, Cu, and Zn exhibited large spatial differentiation and were significantly affected by the external environment. A regional pollution load index of 3.02 indicated overall heavy pollution. The pollution load index for each heavy metal(loid) indicated light pollution. The geoaccumulation index indicated relatively severe As, Cd, Cu, Pb, and Zn pollution. The Nemero Composite Pollution Index value showed that the study area was heavily polluted, with construction land being mildly polluted by Cd, and agricultural land being moderately polluted. The results of the spatial distribution show that there were high levels of contamination in the center. Correlation and principal component analyses showed that the pollution sources of the eight heavy metal(loid)s varied. Hg, Cd, and Pb originate primarily from industrial and agricultural pollution. Traffic sources significantly impacted Cu, Pb, Cd, and Cr. Natural sources are the main sources of Cr, Ni, and Cd. Ni is also affected by industrial sources, whereas Zn and Cu are affected by agricultural pollution. The influences of As, Cd, and Pb on the surface soil in the study area were more serious. Cd is more widely polluted and should be a priority in controlling soil heavy metal(loid)s.

1. Introduction

Heavy metal(loid)s are characterized by their accumulation, persistence, and toxicity. Heavy metal(loid)s are easily absorbed by humans and animals through the food chain, jeopardizing the ecology and human health [1]. The sources of heavy metal(loid)s in soil are relatively complex and mainly comprise anthropogenic and natural sources [2]. Anthropogenic sources include mining, smelting, wastewater, exhaust, fertilizers, and pesticides. Natural sources include rock weathering during soil and volcanic eruptions [3]. Qualitative identification of soil heavy metal(loid) pollution and source analysis are important for controlling soil heavy metal(loid) pollution. Timely determination of soil heavy metal(loid) levels and their source pathways is particularly important for preventing and controlling soil heavy metal(loid) pollution.

Identification of pollution sources is a scientific prerequisite to carry out pollution prevention and control work, and previous researchers have used stable isotope techniques (Pb) [4,5], new geochemical tracers (Cd) [6], and multivariate statistical analyses (PCA and CA) [7] to qualitatively analyze the sources of heavy metals, whereas the chemical mass balance (CMB) model [8], the absolute principal component score-multivariate linear regression (APCS-MLR) model [9], UNMIX model [10], and Positive Matrix Factorization (PMF) [11] were used for quantitative analysis, and the accuracy of their results can be significantly affected by the number of samples. T-NSE [12], which has higher accuracy and flexibility, uses neural networks to derive t-SNE hyperparameter combinations. Semi-supervised learning, which has the advantages of both supervised and unsupervised learning, can be used mainly for datasets with mainly unlabeled data, and few labeled data and automatic clustering of unlabeled data [13] improves the machine learning prediction performance to a great extent.

With industrialization and urbanization, soil heavy metal(loid) pollution from industrial parks has become increasingly significant. Different concentrations of Cd, Zn, Pb, and Hg pollution have been reported in various industrial parks [14]. The highly contaminated samples were mainly distributed in enterprise-intensive areas [15].

The important mineral resource regions in China include Sichuan, Yunnan, and Guizhou. The development of these regional economies has driven heavy metal(loid) pollution by Cd, Zn, Pb, As, and other heavy metal(loid)s [16,17]. Heavy metal(loid) pollution of soils is more serious in the Yunnan Province, especially in industrial parks where polluting enterprises are concentrated, owing to the early non-standardized treatment of pollutants, lack of a functional layout, and other reasons. This could result from the extensive accumulation of heavy metal(loid)s in the surrounding soil. However, little is known about heavy metal(loid) pollution in the soil of industrial parks. Furthermore, the unknown characteristics of soil heavy metal(loid) pollution, unclear spatial distribution patterns, and unknown sources make it difficult to provide a basis for controlling soil pollution sources and remediating contaminated land in the area.

This study was conducted in an industrial park in Kunming, Yunnan Province, China. Heavy metal(loid)s in the surface soil (0 to an approximate depth of 20 cm) were studied; evaluation of soil contamination in different land use types using the pollution load index, geoaccumulation index, and Nemero composite index methods; and analysis of heavy metal(loid) sources, which provide theoretical support for the prevention and control of heavy metal(loid) pollution of soil in Kunming City, will inform efforts to enhance regional environmental quality.

2. Materials and Methods

2.1. Overview of the Study Area

The study area is a key industrial park in the northernmost area of Kunming City, Yunnan Province, covering a total area of 1858.79 km². The study area is divided into two groups. Block 1 is mainly used in chemical industries involving heavy metal(loid) Cu and phosphorus (P). This area focuses on the comprehensive utilization of resources, the processing of ferrous, rare, and precious metals, and the extension and development of new building materials. Block 2 contains advanced manufacturing industries that focus mainly on foodstuffs, light industry, and machinery.

At the end of 2018, there were ferrous metal enterprises, such as the production of ferroalloys, manganese (Mn), Cr, silicon (Si), furnace charge, and steelmaking deoxidizers; nonferrous and rare metal enterprises, such as Cu smelting and sulfuric acid production; P-related chemical enterprises, such as production of feed additives, hazardous chemicals, fertilizer, and other products; machinery manufacturing industries; food and drug processing industries, including pharmaceutical manufacturing, food processing, and others; building materials enterprises, including production of cement, gypsum, mechanized sand, and other products; battery car manufacture; and processing of plastic and other products [18].

2.2. Sample Collection and Analyses

Based on the sewage discharge of enterprises in the industrial park, the distribution of sensitive points within a certain range in the surrounding area, the original use function of the land parcels, the characteristics of contamination, the soil monitoring network in the industrial park, and the surrounding area were constructed. One hundred samples of differently utilized land surface soil (0–20 cm) were collected; 60 samples were from construction land and 40 from agricultural land. To accurately determine whether the soil was polluted, the pollutant type, and the degree of pollution, sampling points were designed by combining professional judgment and systematic random distribution (Figure 1).

Soil samples were collected in strict accordance with the China Soil Environmental Quality Risk Control Standards for Soil Pollution of Agricultural Land (GB15618-2018) [19], Soil Environmental Quality Risk Control Standards for Soil Pollution of Construction Land (GB36600-2018) [20], Technical Specification for Soil Environmental Monitoring (HJ/T166-2004) [21], Determination of eight Effective State Elements in Soil (HJ804-2016) [22], and other relevant technical specifications. The samples were prepared using non-metallic tools and utensils to remove impurities such as plant roots and gravel (at least 2 kg per portion), packed in cloth bags, dried naturally, crushed with a wooden stick, passed through a 2 mm aperture sieve, mixed and weighed, and then sent to the laboratory for analysis. Sample testing was performed at the Testing Center of the China Inspection Body and Laboratory Mandatory Approved by the qualified Kunming Geological Survey Institute of the General Administration of Metallurgical Geology of China. The sample testing process adopted national-level standard substances (GBW series) for quality control. Each batch of the analytical process involved quality control measurements of parallel double samples. Eight elements (As, Pb, Cu, Zn, Cd, Hg, Ni, and Cr) were detected. The Cr content was determined using flame atomic absorption spectroscopy. The As and Hg contents were determined using atomic fluorescence spectrometry. Cd, Pb, Ni, and Cu contents were determined by inductively coupled plasma mass spectrometry. Zn content was determined using inductively coupled plasma emission spectrometry. Finally, pH was determined using a potentiometric method.

2.3. Pollution Load Index (PLI)

The PLI evaluates the integrated pollution levels of multiple heavy metal(loid)s in the soil, visualizes the contribution of pollutants to the polluted area, and determines the spatial and temporal trends of heavy metal(loid)s a site [23]. The PLI can be used to determine whether pollutants are under the integrated control of natural and anthropogenic sources [24,25]. The formulae used are as follows:

$$\mathrm{PLI}=\sqrt[\mathrm{n}]{{\mathrm{C}\mathrm{F}}_1\times {\mathrm{C}\mathrm{F}}_2\times \dots \times {\mathrm{C}\mathrm{F}}_{\mathrm{n}}}$$

(1)

CF_i = C_i/C0_i

(2)

where PLI is the pollution load index of a point in the study area, n is the number of polluted heavy metal(loid) elements, CF_i is the pollution coefficient of heavy metal(loid) i in the soil, C_i is the measured content of heavy metal(loid) i in the soil (mg/kg), and C0_i is the evaluation standard of heavy metal(loid) i. The background value of the heavy metal(loid)s in the soil of Kunming was chosen as the standard evaluation value (mg/kg).

The PLI for a given region is calculated as follows.

$${\mathrm{P}}_{\mathrm{zone}}=\sqrt[\mathrm{m}]{{\mathrm{P}\mathrm{L}\mathrm{I}}_1\times {\mathrm{P}\mathrm{L}\mathrm{I}}_2\times \dots \times {\mathrm{P}\mathrm{L}\mathrm{I}}_{\mathrm{m}}}$$

(3)

where P_zone is the regional PLI, n is the number of metal elements, and m is the number of sampling points. The evaluation levels are presented in

Table 1. Evaluation and classification of soil heavy metal(loid) pollution.

Geological Cumulative Index	Degree of Contamination	Nemero Combined Pollution Index	Degree of Contamination	Pollution Load Index	Degree of Contamination
Igeo ≤ 0	Uncontaminated	PN ≤ 0.7	Uncontaminated	PLI < 1	Uncontaminated
0 < Igeo ≤ 1	Low pollution	0.7 < PN ≤ 1	Warning Level of Caution	1 ≤ PLI < 2	Low pollution
1 < Igeo ≤ 2	Medium pollution	1 < PN ≤ 2	Low pollution	2 ≤ PLI < 3	Moderate pollution
2 < Igeo ≤ 3	Medium-strong pollution	2 < PN ≤ 3	Moderate pollution	PLI > 3	High pollution
3 < Igeo ≤ 4	Strong pollution	PN > 3	High pollution
4 < Igeo ≤ 5	Strong-extreme pollution
Igeo > 5	extreme pollution

.

2.4. Geoaccumulation Index (I_geo)

This index integrates the effects of anthropogenic factors and soil matrices on the distribution of heavy metal(loid) contents. I_geo is widely used to study soil heavy metal(loid) pollution caused by anthropogenic activities [26]. The formula is as follows.

I_geo = log₂[C_n/1.5 × B_n]

(4)

where I_geo is the geoaccumulation index, C_n is the measured value of elements in soil (mg/kg), B_n is the background value of heavy metal(loid)s in the soil (mg/kg), and 1.5 reflects the fluctuation of the background value in different areas due to differences in soil matrices. The evaluation levels are listed in Table 1.

2.5. Nemero Combined Pollution Index

Based on a single factor index, this index synthesizes the pollution level of each heavy metal(loid) and highlights the environmental damage caused by heavy metal(loid)s [27,28]. The formula is as follows.

$$\mathrm{P}{\mathrm{N}}_{ }=\sqrt{{\displaystyle \frac{{\mathrm{p}}_{\mathrm{i}\mathrm{a}\mathrm{v}\mathrm{e}}^2+{\mathrm{p}}_{\mathrm{i}\mathrm{m}\mathrm{a}\mathrm{x}}^2}{2}}}$$

(5)

P_i = C_i/S_i

(6)

where P_N is the soil heavy metal(loid) Nemero Comprehensive Pollution Index, P_iave is the average value of n heavy metal(loid) single factor pollution index; P_imax is the maximum value of n heavy metal(loid) single factor pollution index; C_i is the measured value of soil heavy metal(loid) i content (mg/kg), and S_i is the evaluation standard for heavy metal(loid) i pollution in soil. Soil Environmental Quality Construction Land Soil Pollution Risk Control Standards (GB15618-2018) and Soil Environmental Quality Agricultural Land Soil Pollution Risk Control Standards (GB36600-2018) [19,20] were selected as the heavy metal(loid) control standard values. The evaluation levels are listed in Table 1.

3. Results

3.1. Statistical Characterization of the Meta-Content of the Selected Heavy Metal(loid)s in Soil Samples

The mean soil pH was 7.8 and 9.02 for construction and agricultural lands, respectively (

Table 2. Heavy metal(loid) contents and soil statistical characteristics.

Land Use Patterns		pH	As	Hg	Cd	Pb	Ni	Cu	Zn	Cr
construction land	Max.	9.11	1912.00	18.40	1208.00	6484.00	93.40	3429.00	7455.00	1.40
	Min.	5.23	14.70	0.06	0.61	43.30	19.40	48.20	63.70	0.50
	Avg.	7.80	423.72	3.23	92.15	1153.11	60.21	692.78	1136.99	0.87
	SD	0.90	555.37	4.86	259.33	1561.01	18.48	909.51	1664.64	0.27
	CV	0.12	1.31	1.51	2.81	1.35	0.31	1.31	1.46	0.31
agricultural land	Max.	9.46	38.40	0.50	7.85	298.00	74.40	249.00	388.00	129.00
	Min.	6.05	11.20	0.05	0.69	35.60	22.40	43.20	111.00	51.10
	Avg.	8.02	23.08	0.16	2.29	76.61	45.40	141.10	206.38	80.57
	SD	0.54	7.11	0.10	1.45	43.01	13.96	64.23	65.96	15.08
	CV	0.07	0.31	0.63	0.63	0.56	0.31	0.46	0.32	0.19
Upper Crustal Mean (UCC)			4.80	0.05	0.09	17.00	47.00	28.00	67.00	92.00
Enrichment Factor (EF)	construction land		88.28	64.60	1023.89	67.83	1.28	24.72	16.97	0.01
	agricultural land		4.81	3.20	25.44	4.51	0.97	5.04	3.08	0.86
screening value	construction land		60	38	65	800	900	18,000	-	5.7
	agricultural land	5.5 ≤ pH ≤ 6.5	30	0.5	0.3	90	70	50	200	150
		6.5 < pH ≤ 7.5	25	0.6	0.3	120	100	100	250	200
		pH > 7.5	20	1	0.6	170	190	100	300	250

). The average heavy metal(loid) content in the surface soil was in the order of Pb > Zn > Cu > As > Cd > Ni > Hg > Cr for construction land and Zn > Cu > Cr > Pb > Ni > As > Cd > Hg for agricultural land.

The coefficients of variation (Cv) [29] of As, Hg, Cd, Pb, Cu, and Zn in construction land were all >1, which is an anomalously strong variation, indicating that they are greatly influenced by human activities. Cd content was significantly higher than the background value in Kunming [30]. The Cv values of the contents of Hg, Cd, and Pb contents in the agricultural soils were >0.5, which also indicated strong variation. Relative to the background values in Kunming and the contents of the upper crust [31], the study area was significantly enriched in Cd (1023.89 times). All elements exceeded the background values in the agricultural land in Kunming, except for Cr (the data were logarithmically (In) treated to observe subtle changes in heavy metal(loid) content data in a more subtle manner considering that the data differed too much) (Figure 2).

A comparison of the screening values (Table 2) revealed certain points in the soil samples where different heavy metal(loid) concentrations were exceeded. In the 20 sample points of the construction land soil, the contents of As, Cd, and Pb exceeded the standards in 14 (70%), five (25%), and eight (40%) samples, respectively. The concentrations of Cd, As, and Cu in the 40 soil sample points on the agricultural land exceeded the standards in 40 (100%), 22 (55%), and 25 (62.5%) samples, respectively. In these soil samples, Zn and Pb exceeded the relevant standards in only 15% and 2.5%, respectively. None of the remaining heavy metal(loid)s exceeded the relevant standards.

3.2. Characterization of Spatial Distribution and Pollution Evaluation

3.2.1. PLI

The PLI was used to evaluate and analyze the pollution characteristics of soil heavy metal(loid)s in the study area. The zone of soil heavy metal(loid)s in the study area was 3.02, indicating heavy soil pollution. The PLI values for each sampling point (

Table 3. Statistical characteristics of pollution levels in soil samples of different land use types.

Pollution Load Index	Agricultural Land	Construction Land	Total	Proportions (%)
Uncontaminated	1	0	1	1.7
Low pollution	28	4	32	53.3
Moderate pollution	9	0	9	15
High pollution	2	16	18	30
Total (number of sampling points)	40	20	60	100
Proportion of pollution (%)	97.5	100	98.3

Note: The number of statistics is the number of sampling points.

) showed that the proportions of non-polluted, lightly polluted, moderately polluted, and heavily polluted samples were 1.7, 53.3, 15, and 30%, respectively, with light pollution being predominant. The statistical results of the soil contamination levels of different land use types in the study area showed that the percentage of contaminated sample points on construction land (100%) was higher than that on agricultural land (97.5%).

3.2.2. Geoaccumulation and Nemero Composite Pollution Indices

The I_geo results revealed slight contamination with all heavy metal(loid)s, except for Cr (Figure 3). The contaminated samples, relative to the total number of samples, were in the following order Cd > As > Zn > Cu = Pb > Hg > Ni > Cr.

The degree of soil heavy metal(loid) pollution is shown in

Table 4. Degree of soil heavy metal(loid) pollution.

Item	Pi								PN
Heavy Metal(loid)	As	Hg	Cd	Pb	Ni	Cu	Zn	Cr
Construction land	7.06	0.08	1.42	1.44	0.07	0.04	-	0.15	35.51
Agricultural land	0.31	0.09	2.22	0.17	0.1	0.62	0.2	0.06	7.38

Note: ‘-’ means that there is no such reference value in the Soil Pollution Risk Control Standard for Construction Land (GB36600-2018).

. All samples showed heavy pollution (P_N > 3). However, the Nemero Composite Pollution Index of the construction land was higher than that of agricultural land. Construction land was contaminated with As and Pb to a greater extent than agricultural land. Agricultural land was heavily contaminated with As, mildly contaminated with Cd and Pb, and moderately contaminated with Cd.

3.3. Source Analysis of Heavy Metal(loid)s

3.3.1. Spatial Distribution Characteristics

Using the inverse distance-weighted interpolation method in the statistical module of ArcGIS 10.8 software (10.8.12790), the contents of 60 surface soil sample points were interpolated and analyzed to obtain the spatial distribution of the contents of the eight heavy metal(loid)s (Figure 4) and different distribution characteristics in the region. The soil As content was high in the center of the area and symmetrically distributed in the northwest and southeast. The Ni content is high in the central and southeastern regions. The content distributions of Hg and Pb were very similar, with two highly polluted areas in the center of the area. Areas were highly polluted with Cd and Zn. The Cu content was distributed as an island with a high center and low circumference, and its content gradually decreased along circumference. The distribution pattern of Cr was completely different from that of other heavy metal(loid)s, with the presence of two highly polluted areas in the southeast. In general, the distribution characteristics of the other seven elements are similar. The high pollution content in the center of the area was primarily due to it being an industrial area dominated by heavy metal(loid) smelting, P-related chemical industries, and other relevant industries. It was initially inferred that some heavy metal(loid)s in the study area originated from industrial parks.

3.3.2. Heavy Metal(loid) Correlation Analysis

To investigate the geochemical characterization of heavy metal(loid)s in the soils of construction and agricultural lands in the study area, Pearson correlation was performed for analysis of the contents of heavy metal(loid)s (

Table 5. Pearson correlation coefficients of different elements in construction land and agricultural land.

Item		As	Hg	Cd	Pb	Ni	Cu	Cr	Zn
Construction land (A)	As	1
	Hg	0.950 **	1
	Cd	0.954 **	0.906 **	1
	Pb	0.925 **	0.816 **	0.918 **	1
	Ni	0.541 *	0.456 *	0.640 **	0.616 **	1
	Cu	0.795 **	0.769 **	0.764 **	0.777 **	0.631 **	1
	Cr	0.266	0.157	0.380	0.281	0.278	0.154	1
	Zn	0.874 **	0.756 **	0.920 **	0.898 **	0.608 **	0.701 **	0.408	1
Agricultural land (B)	As	1
	Hg	0.583 **	1
	Cd	0.680 **	0.817 **	1
	Pb	0.515 **	0.620 **	0.517 **	1
	Ni	0.199	0.294	0.346 *	0.256	1
	Cu	−0.017	0.361 *	0.411 **	0.061	0.586 **	1
	Cr	0.250	0.168	0.197	0.163	0.721 **	0.286	1
	Zn	0.203	0.397 *	0.325 *	0.404 **	0.093	0.293	−0.012	1

Note: * indicates that the correlation is significant at the p < 0.05 level (two-sided); ** indicates that the correlation is significant at the p < 0.01 level (two-sided).

). In the construction land (A), at the p < 0.01 and p < 0.05 levels, all the elements, except for Cr, showed highly significant positive correlations between the two elements. Excess As, Cd, and Pb showed highly significant positive correlations with other elements. In the Pearson correlation coefficient table of heavy metal(loid)s in agricultural land soil (B), at p < 0.01 level, Cr was the only element showing highly significant positive correlation with Ni. At p < 0.05, Ni was the only element that showed a highly significant positive correlation with Cd, indicating that a portion of the heavy metal(loid)s in the soil (Cr, Ni, and Cd) had other sources. Cd, which exceeded the standard, showed a highly significant positive correlation with the other elements, with the exception of Cr.

3.3.3. Principal Component Analysis of Heavy Metal(loid) Elements

To further analyze the heavy metal(loid) pollution characteristics of construction and agricultural lands, the variables were subjected to principal component analysis. This analysis is one of the most important means for determining the sources of heavy metal(loid)s [32] (

Table 6. Principal component analysis results of soil heavy metal(loid) elements in construction land and agricultural land.

Item	Construction Land (A)			Agricultural Land (B)
	Principal Component Factors (Maximum Variance Rotated Solution)
	PC1	PC2	PC3	PC1	PC2	PC3
As	0.956	0.223	0.124	0.877	0.128	−0.166
Hg	0.951	0.137	0.002	0.805	0.130	0380
Cd	0.893	0.324	0.254	0.798	0.214	0.323
Pb	0.869	0.352	0.159	0.781	0.038	0.147
Ni	0.319	0.920	0.156	0.155	0.898	0.222
Cu	0.739	0.505	−0.050	−0.020	0.533	0.756
Cr	0.125	0.104	0.974	0.157	0.884	−0.128
Zn	0.813	0.332	0.325	0.338	−0.166	0.729
eigenvalue	4.696	1.519	1.185	2.827	1.980	1.467
variance (%)	58.596	18.985	14.813	35.340	24.752	18.336

A). The three principal component factors obtained by rotational solving contributed approximately 92.39%, and the first principal component factor (PC1) contributed 58.60% of the variance. Six elements (As, Hg, Cd, Pb, Cu, and Zn) exhibited high loadings, indicating similar origins. The second principal component factor (PC2) contributed 18.99%, with high loadings of Ni and Cu. The contribution of the third principal component (PC3) was approximately 14.81%, and Cr had a high loading. Figure 5A depicts that the elemental components with higher scores in PC1 and PC2 were identical, indicating similar origins. Pearson’s correlation analysis indicates a highly significant correlation between these elements.

The three principal component factors of the rotation solving soil samples from the agricultural land (Table 6B) contributed approximately 78.43%, whereas PC1 contributed approximately 35.34%, with As, Hg, Cd, and Pb having the highest scores. Ni and Cr had the highest PC2 scores (with approximately 24.75%). PC3 had the highest scores for Cu and Zn, explaining approximately 18.34% of the overall variation. The results of the Pearson correlation analysis are shown in Figure 5B. Therefore, heavy metal(loid)s have similar sources in construction and agricultural soils.

4. Discussion

4.1. Heavy Metal(loid) Pollution and Risk Assessment

Three evaluation methods were selected to accurately assess the status of soil heavy metal(loid) pollution in the industrial parks, understand the contribution of heavy metal(loid) pollutants to the soil environment in the study area, and perform risk evaluation. The results differed owing to the different evaluation methods used. The three evaluation methods revealed that As, Cd, and Pb pollution in the study area was concerning. In particular, Cd contamination sites were more extensive, consistent with a previous study [33]. Varying degrees of As (which has a high risk of carcinogenicity), Pb, and, most seriously, Cd contamination were evident in all areas of Yunnan [34,35,36,37].

4.2. Source Analysis of Heavy Metal(loid) Pollution

Accurate identification of pollution sources is vital for the prevention and control of soil heavy metal(loid) pollution [38,39]. Pearson and principal component analyses were used to broadly classify the sources of the eight heavy metal(loid)s contaminating construction land into three categories. During the first stage (As, Cd, Hg, Pb, and Zn), the mean values of these elements significantly exceeded the soil background values in Kunming (Figure 2). The Cv values of As, Hg, Cd, Pb, and Zn were high, indicating that they were subjected to large anthropogenic factors. Industrial pollution significantly affects As and Hg are greatly affected by industrial pollution [40,41]. There are more vehicles in industrial parks, and these enterprises include automobile recycling and dismantling. Cd and Pb may originate from automobile exhaust emissions and tire wear, whereas Pb and Zn may originate from engine wear, braking, or other sources [42]. Metal smelting is the primary cause of Cd and Pb pollution [43,44].

The second category (Ni and Cu) includes elements that occur in metal smelting wastes [45,46]. The third category comprises only Cr and is strongly influenced by motor vehicle exhaust [47], although Cr is released in varying degrees from various parts of the cement production process [48].

The eight heavy metal(loid)s in the agricultural land were divided into three categories. In the first stage (As, Hg, Cd, and Pb), the average values of the four elements were higher than the background values of the soil in Kunming. The spatial variability of Hg, Cd, and Pb was more obvious. Irrational use of fertilizers and pesticides has led to increased concentrations of As, Cd, and Pb in soil [49,50,51]. Hg contamination is caused by industrial and agricultural pollutants (the latter involves the long-term use of chemical pesticides) [52,53]. Cd can reflect a high Cd geological background and transportation source [54,55].

The second category (Ni and Cr) is less affected by anthropogenic factors. Ni and Cr are closely related to soil matrix formation [56,57]. The third category comprises Cu and Zn, which originate from chicken and pig manure applied during cultivation. Fertilizers containing P or other chemicals can also produce Zn when applied [58].

4.3. Prevention and Control Measures for Heavy Metal(loid) Pollution

This study identified As, Cd, and Pb as the main heavy metal(loid) pollutants found in industrial parks. Their main sources are industries, such as preparing raw materials for the smelting of Cu and Fe and the smelting process, in which the energy mainly comes from coal combustion [59,60,61]. Second, Pb and Cd originate from transportation and natural sources, respectively. Additionally, exhaust emissions, discharge or leakage of heavy metal(loid)-containing wastewater, storage or improper disposal of waste residues, and steel processing and stacking may result in Cd infiltration into site soil [62,63]. Industrial activities, such as hazardous chemical production, pharmaceutical manufacturing, and fertilizer production, can also produce As and Pb pollution [64,65]. Pb pollution also results from automobile exhaust, oil leakage, and wear and tear of concrete pavements, rubber tires, brake pads, etc. [66,67].

At the national level, regular updates on online environmental pollution monitoring information and data sharing should be jointly promoted. Local governments should properly manage and dispose of waste in accordance with local environmental protection laws, actively promote the upgrading of processes and industries, determine diversified evaluation standards and systems, strictly implement the requirements of air pollutant emission limits, formulate scientifically sound regional soil heavy metal(loid) pollution management, soil utilization and remediation, and balanced governance programs for other ecological elements [68,69].

Comprehensive management of industrial parks is performed at the park level, with management projects planned in non-polluted areas to control the generation of new pollutants and a permanent protection system established to ensure park safety. In places affected by heavy metal(loid) pollution, strict control, research, and development of new technologies for heavy metal(loid) pollution remediation should be implemented, such as the study of soil heavy metal(loid) pollution management systems and mechanisms and improved soil heavy metal(loid) pollution management capacity and level. The aim is to achieve sustainable development of the soil environment. Furthermore, enterprises in the affected industrial parks should be supervised to strictly enforce national pollutant emission standards and eliminate the “three industrial wastes.” Relevant companies must formulate technical specifications for the prevention and control of heavy metal(loid)s to enable their supervision and management of heavy metal(loid) emissions. Industrial parks need to establish centralized collection, treatment, and disposal facilities for solid and hazardous wastes and improve measures to prevent seepage, loss, and dispersal of heavy metal(loid) contaminated soil. For the copper smelting industry, strict implementation of measures to curb the emissions of particulate matter and key heavy metal(loid) pollutants is required. The development of clean energy should be encouraged through the development of environmentally friendly raw materials, and the filtration of automobile exhaust pollutants, purification, and prevention capabilities should be strengthened.

As, Cd, Pb, and other heavy metal(loid)s cannot be naturally degraded and are easily enriched in all chains. Acid-based degradation and other methods must also be considered when discharging exhaust gases and treating wastewater [70,71].

5. Conclusions

The present findings revealed significant enrichment of As, Hg, Cd, Pb, Ni, Cu, and Zn in construction land soil in Kunming City. The surrounding agricultural land was significantly enriched in As, Hg, Cd, Pb, Cu, and Zn. Cd was the most enriched element in construction and agricultural lands. In the construction land, seven of the heavy metal(loid)s were higher than the background value of Kunming City, except for Cr, and the contents of As, Cd, and Pb in the construction and agricultural land soil samples were all higher than the soil risk screening values of construction and agricultural lands in China.

The PLI results indicated that the study area was heavily polluted. The I_geo values indicated slight pollution by all eight heavy metal(loid)s, with relatively more serious pollution by As, Cd, Cu, Pb, and Zn. The Nemero Integrated Pollution Index showed that the P_N for construction and agricultural land exceeded 3. Construction land was more polluted by As and Pb than agricultural land. The As pollution of the construction land was heavy, the Cd and Pb pollution of the construction land was mild, and the Cd pollution of the agricultural land was moderate.

The results of the spatial distribution show that, except for Cr, the distribution characteristics of the other seven elements were relatively similar, and there were noticeably high levels of pollution in the center. Correlation and principal component analyses showed that the sources of pollution were different in general. Arsenic, mercury, cadmium, and lead were primarily derived from industrial and agricultural sources. Traffic sources had a greater impact on Zn, Cd, Pb, and Cr, whereas natural sources were the primary sources of Cr, Ni, Cd, and Ni were also affected by industrial sources. Zn and Cu concentrations were also affected by agricultural sources.

In summary, the surface soil of the study area is severely polluted by As, Cd, and Pb. Cd is a widespread pollutant and should be a priority for the management of soil heavy metal (loid) pollution.