Spatial-Temporal Change Characteristic Analysis and Environmental Risk Evaluation of Pesticide Application in Anhui Province

Abstract

The excessive application of pesticides causes the increase in pesticide residues, and pesticide pollution presents a point-line-plane 3D space pollution trend. This paper takes pesticide application intensity (PAI for short) as the object, adopts spatial autocorrelation theory to analyze the spatial-temporal change characteristics of PAI in Anhui Province from 2003 to 2020, and constructs the environmental risk index method to evaluate pesticide environmental risk, aiming to provide a reference for macro control pesticide dosage, zoning guidance, and agricultural sustainable development in Anhui Province. The following results were obtained: From 2003 to 2020, the PAI in Anhui Province showed a spatial-temporal change process of first increasing and then decreasing, as well as first clustering and then random distribution. Before 2013, the significant hot spots and cold spots of PAI in Anhui Province were located in southern and northern Anhui, respectively. In addition, the scope of both experienced a process of expanding and then contracting. In recent years, there have been no significant hot and cold areas of PAI in Anhui Province. The overall trend of pesticide environmental risk in the provincial cities in Anhui Province decreased. In 2020, pesticide environmental risk in various provincial cities is dominated by medium- and low-risk levels, which are still higher than in 2003. High-risk and extremely high-risk levels are mainly concentrated in Huainan and Suzhou in northern Anhui and Chizhou and Huangshan in southern Anhui.

1. Introduction

Pesticide is an indispensable basic material in agricultural production that can protect crops from pests and weeds and can increase yields [1,2]. However, there are problems of over-application of pesticides. Excessive application of pesticides can cause harm in two ways. On the one hand, excessive pesticide residues on agricultural products will affect food safety and threaten human health [3,4]. On the other hand, pesticides that erode soil will enter natural water bodies with rainwater runoff, thereby polluting the ecological environment [5]. Therefore, reasonable control of pesticide application intensity (PAI for short) is of great significance to food safety, human health, and the ecological environment.

Developed countries have begun to attach importance to pesticide application since the last century and have promoted the reduction of pesticide application. In the early 20th century, the pesticide dosage in Britain, the United States, Japan, and other countries dropped significantly. However, pesticide dosage in China continued to grow over the same period. After the introduction of the 2020 Action Plan for Zero Growth of Pesticide Dosage [6] in 2015, China’s PAI began to decline slightly, but it was still at a high level [7]. According to the relevant data from the United Nations Food and Agriculture Organization, China’s pesticide dosage accounts for 43% of the global total. However, China’s arable area only accounts for 8.64% of the global total. Therefore, China’s average PAI (13.1 kg/hm²) is much higher than the Asian average (3.7 kg/hm²) and the global average (2.6 kg/hm²) [8]. Pesticide pollution caused by excessive application of pesticides has become the largest organic pollution in China. Due to the persistence and enrichment of pesticide pollution, pesticide residues will gradually increase with the increase in the amount of use and the number of years of use, showing a point-line-plane 3D space pollution trend [9]. Obviously, pesticide application amount is directly related to pesticide environmental risk. That is to say, the top priority of the control of pesticide pollution is the reasonable control of pesticide application amount.

At present, domestic and foreign research on pesticide pollution mainly focuses on the residual concentration, regional characteristics, and environmental risks of a certain class of officially banned pesticides, such as organochlorine pesticides (OCPs) [10,11,12,13]. There are few studies on the environmental risk caused by pesticide application amount. In addition, the spatial distribution dependence and heterogeneity are relatively underexplored in studies on the regional characteristics of pesticide pollution. Spatial autocorrelation analysis is a statistical method to study the attribute correlation of adjacent regions, playing the role of characterizing dependencies and differences and revealing hot spots. Since the mid-20th century, this method has been widely used in research fields such as ecology [14], medicine [15], environment [16], fisheries [17], land use [18,19], and regional economy [20]. Therefore, the analysis of the spatial distribution characteristics of pesticide application amount through spatial autocorrelation can provide a useful reference for macro control and zoning guidance. As a typical agricultural province in China, Anhui’s agricultural GDP ratio exceeds the national average, and its arable area ranks eighth in the country. At the same time, Anhui PAI is relatively large. The PAI in Anhui in 2017 was as high as 16.94 kg/hm², which was higher than the national average level in the same period. Therefore, this paper takes pesticide application intensity (PAI for short) as the object, adopts the spatial autocorrelation theory to analyze the spatial-temporal change characteristics of PAI in Anhui Province, and constructs the pesticide environmental risk index method for the evaluation of pesticide environmental risk, aiming to provide a decision-making basis for macro control pesticide dosage, zoning guidance, and agricultural sustainable development in Anhui Province.

2. Research Area

Anhui Province is located in East China, between 114°54′ and 119°37′ East longitude and 29°41′ and 34°38′ North latitude. With 16 provincial cities, Anhui Province can be divided into northern Anhui, central Anhui, and southern Anhui according to different conditions such as climate, topography, and culture. Northern Anhui includes Huaibei, Bozhou, Suzhou, Bengbu, Fuyang, and Huainan. Central Anhui includes Hefei, Chuzhou, Lu’an, and Anqing. Southern Anhui includes Ma’anshan, Wuhu, Xuancheng, Tongling, Chizhou, and Huangshan. Northern Anhui is dominated by the plain terrain north of the Huaihe River, with dense population and high land utilization. Central Anhui is between the Huaihe River and the Yangtze River and consists of the Dabie Mountains in the west of Anhui, the Jianghuai platform hills, and the plains along the river. Southern Anhui is mainly mountainous terrain, with high vegetation coverage and small population density. Anhui Province belongs to the transition zone between warm temperate zone and subtropical zone. The north of the Huaihe River has a warm temperate semi-humid monsoon climate, while the south of the Huaihe River has a subtropical humid monsoon climate. Meanwhile, Anhui Province is also an important commercial grain production base in China. In the past 20 years, the average annual growth rate of grain crops in Anhui Province is about 6.35%. In 2021, the output of grain crops in Anhui Province will be as high as 40.876 million tons, ranking fourth in the country (Figure 1).

3. Materials and Methods

3.1. Basic Data

3.1.1. Data Source

The basic data such as arable area, planting area, and pesticide application amount used in this paper are mainly from the 2004–2021 Anhui Statistical Yearbook [21] compiled by the Anhui Provincial Bureau of Statistics and the 2004–2021 China Statistical Yearbook [22] compiled by the National Bureau of Statistics. In light of to the relevant data in the National Ecological Civilization Construction Demonstration Villages and Towns Indicators (Trial) [23] issued by the Ministry of Environmental Protection in 2014 and the data of developed countries such as the United States and Japan in 2018 [7], the environmental risk threshold of PAI was determined to be 6 kg/hm².

3.1.2. Data Processing

In 2011 and 2015, Anhui Province adjusted the administrative district planning of former Chaohu City, Shou County of former Lu’an City, and Zongyang County of former Anqing City. Chaohu City is divided under Hefei City, Ma’anshan City, and Wuhu City. Shou County of former Lu’an City was placed under Huainan City. Zongyang County of former Anqing City was placed under Tongling City. In order to ensure the reliability of the comparison of research results in different years, the basic data of the relevant regions before 2015 were re-integrated.

3.2. Evaluation Method of Environmental Risk

Environmental risk is divided into non-sudden risk and sudden risk [24]. For non-sudden risk, Rapant et al. proposed the heavy metal pollution risk index method in 2003 [25]. On the basis of the heavy metal pollution index method proposed by Rapant [25], Liu Qinpu [26] constructed an environmental risk assessment method for chemical fertilizer application and achieved a wide range of application effects in Jiangsu, Anhui, Henan, Shandong, and other provinces. On this basis, this paper constructs a pesticide environmental risk index method that reflects pesticide environmental risk by the multiple of PAI exceeding the threshold. The specific formula is as follows:

$$R=\left\{\begin{array}{l}0,\hspace{1em}\mathrm{if}F<FL\\ \frac{F}{FL}-1,\hspace{1em}\mathrm{if}F>FL\end{array}\right.$$

(1)

$$F=\frac{M}{A}$$

(2)

where R is the pesticide environmental risk index; F is the PAI (kg/hm²); FL is the PAI threshold (kg/hm²); M is the pesticide application amount in the year (kg); and A is the arable area (hm²).

According to formulas (1) and (2), when the environmental risk index is 0, the cultivated land has no pesticide environmental risk. When the environmental risk index is greater than 0, pesticide environmental risk appears in cultivated land, which increases with the increase in the value. On the basis of the research results of Wu Xubin et al. [27] regarding the classification standard of farmland pesticide pollution, as well as the multiple relationship between the risk screening value of organochlorine pesticides in soil (when the content of organochlorine pesticides in the soil is lower than this value, the risk to human health is negligible) and the risk control value (when the content of organochlorine pesticides in the soil is higher than this value, there is an unacceptable risk to human health) in the Soil Environmental Quality Risk control Standard for Soil Contamination of Development Land (GB36600-2018) [28] issued by the Ministry of Ecology and Environment, the classification criteria for pesticide environmental risk levels were determined, as shown in

Table 1. Classification standard of pesticide environmental risk.

Risk Level	No Risk	Low Risk	Medium Risk	High Risk	Extremely High Risk
Risk index	0	0~1	1~3	3~5	>5

.

3.3. Theory of Spatial Autocorrelation

Spatial autocorrelation can analyze the correlation of spatial attributes of different spatial units and reflect the aggregation degree of spatial attributes in the study area [29]. In accordance with the scope of analysis space, spatial autocorrelation can be divided into global spatial autocorrelation and local spatial autocorrelation. Global spatial autocorrelation reveals the degree of autocorrelation of spatial attributes across the study area. Local spatial autocorrelation reflects the degree of correlation between a spatial unit and its neighboring units on a certain spatial attribute.

3.3.1. Global Spatial Autocorrelation

Global spatial autocorrelation is often measured by Global Moran’s I coefficient [29], Global Geary’s C coefficient [30], and Global Getis’ G coefficient [31]. In this paper, Global Moran’s I coefficient was used to measure global spatial autocorrelation. The specific formula is as follows:

$$I=\frac{{\displaystyle \sum _{i=1}^n(x_i-\overline{x}){\displaystyle \sum _{j=1}^n{W}_{ij}(x_j-\overline{x})}}}{s^2{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{W}_{ij}}}}$$

(3)

$$s^2=\frac{1}{n}{\displaystyle \sum _{i=1}^n{(x_i-\overline{x})}^2}$$

(4)

where I represents the Global Moran’s I coefficient, which is distributed in the numerical range of [−1,1]; n is the number of spatial units; x_i and x_j represent the attribute values of the spatial units i and j, respectively; $\overline{x}$ represents the average value of the spatial attribute values; and W_ij is the spatial weight coefficient matrix that represents the adjacent relationship of each spatial unit.

When I > 0, the distribution of spatial attributes in the research area is clustered, which is positive spatial autocorrelation. The closer I is to 1, the stronger the agglomeration and positive correlation. When I < 0, the distribution of spatial attributes in the research area is in a discrete state, which is negative spatial autocorrelation. The closer I is to −1, the stronger the dispersion and negative correlation. When I is close to 0, there is no spatial autocorrelation for the spatial attributes, and it is randomly distributed in space.

The analysis of global spatial autocorrelation based on Global Moran’s I coefficient requires a significance test. The specific formula is as follows:

$$Z(I)=\frac{I-E(I)}{\sqrt{VAR(I)}}$$

(5)

where Z(I) represents the significant level of spatial autocorrelation, and E(I) and VAR(I) represent the mathematical expectation and variance of I, respectively.

3.3.2. Local Spatial Autocorrelation

Global spatial autocorrelation can judge the overall distribution of spatial attributes in space, but it is difficult to reflect the aggregation location and spatial differences of spatial attributes. To make up for this limitation, Anselin proposed LISA (Local Indicators of Spatial Association) [32] to reveal the correlation between spatial units and their neighboring units in spatial attributes, as well as to identify spatial agglomerations and spatial differences. In this paper, Local Moran’s I coefficient [32] was used to quantitatively judge the spatial attribute correlation between spatial units and neighboring units. The specific formula is as follows:

$$\begin{array}{ll}{I}_i& =\frac{n(x_i-\overline{x}){\displaystyle \sum _{j=1}^n{W}_{ij}(x_j-\overline{x})}}{{\displaystyle \sum _{i=1}^n{(x_i-\overline{x})}^2}}\\ & =z_i^{\prime }{\displaystyle \sum _{j=1}^n{W}_{ij}{z}_j^{\prime }}\end{array}$$

(6)

where I_i is the Local Moran’s I coefficient of the space unit i, and $z_i^{\prime }$ and $z_j^{\prime }$ are the attribute values normalized by the standard deviation. n, x_i, x_j, $\overline{x}$, and W_ij have the same meanings as above.

When I_i > 0, there is a strong positive spatial autocorrelation between the spatial unit i and the adjacent units in spatial attribute values, which is a local clustering of spatial attributes. When I_i < 0, there is a strong negative spatial autocorrelation between the spatial unit i and the adjacent units in spatial attribute values, and the spatial attributes are locally discrete. Similarly, the Local Moran’s I coefficient also needs to be tested for significance, and the method is the same as the Global Moran’s I coefficient.

With $z_i^{\prime }$ and ${\displaystyle \sum _{j=1}^n{W}_{ij}{z}_j^{\prime }}$ in formula (6) as the horizontal and vertical coordinates, respectively, Moran scatter diagram [33] is constructed, which can qualitatively distinguish the relationship between the i-th space unit and the adjacent space unit in PAI. According to the positive and negative values of the horizontal and vertical coordinates, the Moran scatter diagram is divided into four quadrants. Observations in the upper right quadrant HH (high–high) or the lower left quadrant LL (low–low) indicate, respectively, that a region and its adjacent regions have a high degree of agglomeration effect, so the properties of adjacent regions gradually tend to be consistent. Observations in the lower right quadrant HL (high–low) and the upper left quadrant LH (low–high) indicate, respectively, large differences in the properties of a region and adjacent regions.

4. Results and Discussion

4.1. Results of PAI

In light of the pesticide application amount and arable area in Anhui Province from 2004 to 2021 [21,22], the PAI of Anhui Province and the cities in the past 18 years was calculated by formula (2), as shown in Figure 2.

It can be seen from Figure 2 that during the period from 2003 to 2020, the PAI in Anhui Province showed a process of first rising, then a steady transition and finally falling. The PAI was as low as 13.20 kg/hm² and as high as 20.02 kg/hm². During 2003–2008, the growth rate was faster, with an average annual growth rate of as high as 8.16%. The period of 2008–2013 was a smooth transition period. During this period, the average PAI increased slightly, with an average annual growth rate of 0.57%. In 2013, the average PAI reached its maximum value. During the period of 2014–2020, the average PAI showed a decreasing state in general, with an average annual reduction rate of 4.01%. Since 2003, the PAI in southern Anhui has increased year by year. In 2008, the PAI reached the highest value of 32.69 kg/hm², with an average annual growth rate of 10.72%. During the period of 2008–2020, the PAI in southern Anhui showed a decreasing state as a whole, with an average annual decrease rate of 5.14%. Since 2003, PAI in central Anhui has increased year by year. In 2013, the PAI in central Anhui reached the highest value of 22.00 kg/hm², with an average annual growth rate of 8.25%. After 2013, the PAI in central Anhui decreased year by year, with an average annual decrease rate of 9.83%. The PAI in northern Anhui has been increasing year by year since 2003, reaching the highest value of 22.88 kg/hm² in 2015, with an average annual growth rate of 6.38%. After 2015, the PAI in northern Anhui decreased year by year, with an average annual decrease rate of 4.34%.

There are two reasons for the above-mentioned changes in PAI. First, PAI is highly correlated with multiple cropping rate. As shown in Figure 3, the highest cultivated land multiple cropping rate in southern and central Anhui occurred in 2009 and 2013, respectively, which was basically the same as the year in which the highest value of PAI appeared. Second, the National Ecological Civilization Construction Demonstration Villages and Towns Indicators (Trial) [23] issued by the Ministry of Environmental Protection in 2014 stipulated PAI as a binding indicator. In addition, the 2020 Action Plan for Zero Growth of Pesticide Dosage [6] issued by the Ministry of Agriculture in early 2015 indirectly urged local agricultural departments to take measures to improve the utilization rate of pesticides. Although the cultivated land multiple cropping rate in each year after 2015 in northern Anhui was slightly higher than that in 2015, the PAI decreased year by year.

As shown in Figure 2, the PAI in northern, central, and southern Anhui was compared. During 2003–2013, the PAI in southern Anhui was significantly higher than that in central and northern Anhui. In the study by Wang Xingqin [34], on the residues of organochlorine pesticides in farmland in Anhui Province in 2010, among the 19 sampling points, the soils of Qingyang County, Chizhou City, in southern Anhui had the highest residues of OCPs and ∑HCH, and the residues of OP’DDE + PP’DDD in the soil of Wuwei County, Wuhu City, southern Anhui were the highest, which indicated that the PAI in southern Anhui was higher than that in central and northern Anhui. After 2013, the PAI in northern Anhui was slightly higher than that in southern Anhui, and significantly higher than that in central Anhui. The main reason lies in the large proportion of cash crop sown area in southern Anhui. Before 2013, the average annual proportion of cash crop sown area in southern Anhui reached 42.78%, significantly higher than 33.18% in central Anhui and 23.14% in northern Anhui. Compared with food crops, the PAI of cash crops was higher [35]. Therefore, the PAI in southern Anhui was significantly higher than that in central and northern Anhui before 2013. Since 2014, the proportion of cash crop sown area in northern, central, and southern Anhui has decreased year by year, resulting in the decrease in PAI. Nevertheless, the PAI in Huainan and Suzhou in northern Anhui was continuously higher than 30 kg/hm², which was more than twice that of other cities in northern Anhui. In research conducted by Xie Lijin et al. [36] in 2019, it was found that the residues of organochlorine pesticides in the sediments of Huaihe River (Huainan section) were greater than those of the downstream of Huaihe River (Bengbu section), which also proved this. In addition, the crop sown area in Huainan and Suzhou accounted for about one-third of that in northern Anhui. Therefore, the PAI in northern Anhui did not decrease significantly and exceeded that in southern Anhui after 2013.

4.2. Results of Spatial Autocorrelation Analysis

In order to reveal the spatial variation relationship of PAI among cities in Anhui Province, this paper used the spatial autocorrelation theory to study the spatial correlation of PAI among various spatial units. The analysis of spatial autocorrelation focuses on the determination of spatial weights. Since each spatial unit in this study is an irregular surface area, the spatial proximity criterion was used to determine the spatial weight matrix. If the two spatial units i and j are adjacent, the weight coefficient W_ij is 1, otherwise it is 0.

Through GeoDA software, the Global Moran’s I coefficient of PAI in various cities in Anhui Province from 2013 to 2020 was calculated. The specific results are shown in Figure 4.

It can be seen from Figure 4 that the Global Moran’s I coefficient of PAI in the cities of Anhui Province is on the decline as a whole. During 2003–2013, Global Moran’s I coefficients were distributed between 0.19 and 0.41, and the corresponding p-values were all less than 0.05 (95% confidence). The PAI in various cities showed a significant positive spatial correlation, and the spatial distribution was clustered. During 2014–2020, Global Moran’s I coefficients were close to 0, and the corresponding p-values were all greater than 0.2. There was no spatial autocorrelation in PAI in various cities, and it was randomly distributed in space. This was because the PAI in southern Anhui during 2003–2013 was significantly higher than that in northern and central Anhui. During the period from 2014 to 2020, the PAI in southern Anhui and northern Anhui was not much different.

Although global spatial autocorrelation can reveal the spatial agglomeration of PAI, it cannot reflect the “hot spot” areas and spatial unit correlation of PAI. Therefore, it is necessary to further analyze the local spatial autocorrelation characteristics of PAI in Anhui Province. In this paper, the Moran scatter diagram in GeoDa software was used to reveal the heterogeneity among spatial units.

The abscissa of the Moran scatter diagram is the standard value of the PAI standard deviation of the spatial unit, and the ordinate is the average standard value of the PAI of the adjacent spatial units determined by the spatial weight matrix. Figure 5 shows the Moran scatter diagram of PAI in Anhui Province in 2003, 2008, 2013, and 2018.

It can be seen from Figure 5 that the cities with PAI belonging to the HH type in the cities of Anhui Province in 2003, 2008, and 2013 were Anqing, Tongling, Chizhou, and Huangshan. Among them, there were four HH-type cities in 2003, and five HH-type cities in 2008 and 2013. Furthermore, Wuhu in southern Anhui and Lu’an in central Anhui were added on the basis of 2003 in 2008 and 2013, respectively. In 2018, the only HH-type cities were Chizhou and Huangshan. Nevertheless, the Moran scatter diagram was not tested for significance. To this end, this paper uses the Z test (p = 0.05) to draw a LISA agglomeration map according to the standard value of PAI in various cities, as shown in Figure 6.

The LISA agglomeration map can visualize the spatial-temporal change pattern of PAI. According to Figure 6, the following findings are made:

(1) Under the 95% confidence level, in 2003, 2008, and 2013, southern Anhui significantly became the HH concentration area of PAI, while northern Anhui significantly became the LL concentration area of PAI. The main reason is that southern Anhui has more mountainous areas and less cultivated land. In order to increase agricultural income, southern Anhui adopted the method of increasing cultivated land multiple cropping rate and cash crop planting rate, which resulted in higher pesticide dosage than that in northern Anhui. Typically, Chizhou and Bozhou are located in southern and northern Anhui, respectively, with similar land areas. Taking 2008 as an example, the forestry area in Chizhou accounted for 64.23% of land area, which was larger than 17.39% in Bozhou. The arable area of Bozhou accounts for 55.81% of land area, which is larger than 9.7% of Chizhou. The geographical features of more mountains and less cultivated land led to the cultivated land multiple cropping rate and cash crop planting rate of Chizhou being 1.67 and 40.69%, respectively, which were higher than 1.48 and 19.50% of Bozhou.

(2) Both the HH agglomeration area and the LL agglomeration area have experienced the process of first expanding and then shrinking. By 2018, there were no significant PAI HH and LL agglomeration areas in Anhui Province. The main reason is that the cash crop planting rate in southern Anhui continued to decline. The Ministry of Environmental Protection and the Ministry of Agriculture issued relevant documents in 2013 and 2015 to urge and guide Anhui Province to continuously take technical measures to improve the utilization rate of pesticides and reduce the pesticides used.

(3) Huainan was a significant PAI HL area in all 4 years. Although Huainan is located in northern Anhui, its PAI was found to be significantly higher than the surrounding cities in northern Anhui, such as Fuyang, Bozhou, and Bengbu. The cultivated land multiple cropping rate and cash crop planting rate in Huainan were not higher than those in the surrounding northern Anhui cities, indicating that its pesticide utilization rate is relatively low. If the significance test is not considered, the same situation exists in Suzhou. Therefore, the above two cities should take necessary measures to reduce pesticide usage and improve pesticide utilization. Contrary to Huainan and Suzhou, Xuancheng in southern Anhui has more mountains and less cultivated land such as Huangshan and Chizhou, which was a significant PAI LH type area in 2003 and 2008. In contrast, its PAI was significantly lower than that of surrounding cities in southern Anhui such as Huangshan and Chizhou, with high pesticide utilization.

(4) Ma’anshan was a significant PAI LL type area in 2013 and a significant PAI HL type area in 2018. From 2013 to 2018, the PAI in Ma’anshan continued to decrease, but the reduction rate was significantly smaller than that in Anhui Province and southern Anhui. The average annual reduction rates of the three were 1.37%, 4.37%, and 7.13%, respectively.

4.3. Pesticide Environmental Risk

Although the analysis of spatial autocorrelation can reveal the spatial distribution characteristics of PAI in various cities in Anhui, it cannot reflect the environmental risk caused by PAI to cultivated land. According to the PAI of the cities in Anhui Province, the environmental risk threshold of PAI was determined to be 6 kg/ hm². The pesticide environmental risk index of the cities in Anhui Province in the past 18 years was calculated according to formula (1). According to Table 1, pesticide environmental risk level was determined. Some results are detailed in Figure 7.

It can be seen from Figure 7 that the pesticide environmental risk in Anhui Province as a whole presents a medium risk. In 2008 and 2013, the overall risk of central Anhui was medium, and the overall risk was low in other major years. The main reason is that Anqing and Lu’an were high or extremely high risk in 2008 or 2013. In 2003, northern Anhui presented a low risk overall, and in the other major years, the overall risk was medium. In addition to the fluctuations in Bozhou, the other five cities in northern Anhui also experienced similar situations. Among them, Huainan and Suzhou were high risk in 2008, 2013, 2018, and 2020. In 2008 and 2013, southern Anhui presented a high risk as a whole, and in the other major years, the overall risk was medium. The main reason is that Chizhou and Huangshan showed high or extremely high risk in each major year.

In accordance with the distribution of pesticide risk levels of the cities in Anhui, the pesticide environmental risk levels in the major years of the cities were gradually dominated by medium and low risks from 2008 to 2020, except for 2003. As shown in Figure 8, in 2020, there were six cities with pesticide environmental risk levels of medium risk and low risk in Anhui Province, accounting for 75% of the 16 cities. Compared with 2003, pesticide environmental risk of cities in Anhui Province was still higher. In 2020, there were still one and three cities with extremely high risk and high risk in Anhui Province, respectively. In 2003, there were zero and two cities in Anhui Province with extremely high risk and high risk, respectively, with one no-risk city. Although pesticide environmental risk in Anhui Province is gradually decreasing, it is necessary to continue to strengthen the implementation of various measures to reduce PAI in various cities, especially in high-risk and extremely high-risk cities.

5. Conclusions and Recommendations

5.1. Conclusions

(1) The PAI in Anhui Province from 2003 to 2020 showed a process of first rising and then falling. The PAI in southern Anhui was significantly higher than that in northern and central Anhui before 2013. Since 2014, the PAI in northern Anhui has been slightly higher than that in southern Anhui and significantly higher than that in central Anhui. The main reason is that PAI is significantly affected by cultivated land multiple cropping rate and cash crop planting rate, and relevant policies issued by the Ministry of Environmental Protection and the Ministry of Agriculture in 2014 and 2015 prompted local agricultural departments to take measures to reduce pesticide use and improve pesticide utilization.

(2) Through the analysis of spatial autocorrelation, the PAI in Anhui Province from 2003 to 2013 showed a significant spatial agglomeration distribution. The agglomeration distribution area showed a change process of first expanding and then shrinking. Since 2014, PAI in various cities in Anhui Province has been randomly distributed. The PAI hot spot, namely, the HH area, is mainly concentrated in southern Anhui. The PAI cold spot area, namely, the LL area, is mainly concentrated in northern Anhui. The main reason is that there are more mountains and less cultivated land in southern Anhui, as well as less mountains and more cultivated land in Northern Anhui. Therefore, the multiple cropping rate and cash crop planting rate in southern Anhui are significantly higher than those in northern Anhui.

(3) The overall pesticide environmental risk level in Anhui Province is medium. Among them, the overall risk level in Central Anhui is low, while the overall risk level in South Anhui and North Anhui is medium. Since 2008, the pesticide environmental risk levels of the cities in Anhui Province in major years have gradually been dominated by medium and low risks level. In 2020, the number of cities with pesticide environmental risk levels of medium and low accounted for 75% of the total number of cities in Anhui Province. However, pesticide environmental risk levels in some cities in northern and southern Anhui present high or extremely high risks. Compared with 2003, the pesticide environmental risk in the cities in Anhui Province is still relatively high.

5.2. Recommendations

Although the use of pesticides in Anhui Province has gradually decreased in recent years, there is still widespread pesticide environmental risk. Some cities have higher pesticide environmental risk levels. In order to effectively control the pesticides used and ensure the safety of agricultural production, the quality and safety of agricultural products and the safety of the ecological environment, the following suggestions are put forward:

(1) All cities and counties should establish the pesticide management system, encourage the development and registration of safe and efficient pesticide products, eliminate high-toxicity and high-risk pesticides, implement a designated management system for restricted use of pesticides, and formulate designated management plans for restricted use of pesticides.

(2) Environmental protection and agricultural departments should combine the work of agricultural pollution control to solve the problem of recycling and disposal of expired pesticides and pesticide wastes, reduce the harm of expired pesticides and pesticide wastes to the ecological environment, and promote sustainable agricultural development.

(3) It is necessary to continue to further promote the zero-growth action of pesticide use; promote the integrated development of crop pest control and green prevention; and strengthen the guidance of scientific and safe use of pesticides and the application, as well as the spread of harm reduction technology.

(4) Agricultural departments should regularly organize relevant technical personnel to go to villages and carry out pesticide application training. According to the characteristics of crops and the applicability of pesticides, technical personnel should guide farmers in a rational selection of pesticides, scientific dispensing, and precise application of pesticides, so as to improve the level of farmers’ pesticide application.

(5) There is a need to strengthen the supervision of pesticide use in cities such as Huainan, Suzhou, Chizhou, and Huangshan; to urge and guide the application of green control technologies; to promote the use of efficient, low-toxicity, low-residue pesticides and new types of equipment; to strengthen the scientific and rational use of pesticides; and to reduce pesticide waste.