Analysis of the Spatial Pattern and Influencing Factors of Green Food Enterprises in Yunnan Province ()
1. Introduction
Green food refers to food products that are processed according to specific standards and production methods and that have been certified through specialized procedures. The development status of green food enterprises serves as an important indicator for assessing regional development levels, reflecting both the progress of agricultural modernization and the advancement of ecological civilization within a region [1]-[3]. Therefore, studying the quantity and spatial distribution of green food enterprises is of significant importance for regional social development.
The spatial distribution characteristics of regional green food enterprises can effectively reflect the development status of the local green food industry [4]-[6]. The systematic research on green food enterprises can be traced back to 1999, when scholars such as Liu Lianfu and Huang Li focused on discussing the role of these enterprises within the market economy at that time [7] [8]. Scholars such as Zeng Geqi and Zhang Yong conducted provincial-level, regionally segmented research on the development status of green food enterprises [9]-[11]. Subsequently, related studies gradually shifted toward employing spatial analysis methods to analyze the spatial distribution patterns and characteristics of green food enterprises. Cai Yingying [12] systematically analyzed the spatial coupling between the industrial structure and industrial agglomeration of green food enterprises in Hunan Province using methods such as the entropy method, grey relational model, and spatial autocorrelation model. Subsequently, Gu Bojun [13] investigated the spatial distribution of green food-certified enterprises across China from 2011 to 2020 by applying spatial correlation descriptive methods combined with long-term time-series data, thereby elevating the research scale to the national level.
Building upon the research findings of the aforementioned scholars, the authors take Yunnan Province as the study area. Utilizing data from the 2024 Green Food Database of the China Green Food Development Center and applying the kernel density estimation model, the authors analyze the spatial agglomeration of four categories of green food-certified enterprises in Yunnan Province—vegetables, grains, fruits, and tea. Furthermore, a geographical detector is employed to explore the influencing factors behind their spatial agglomeration patterns. The study aims to provide a decision-making basis for future socio-economic development and ecological civilization construction in Yunnan Province.
2. Data Sources and Research Methods
2.1. Overview of the Study Area
Yunnan Province is located in the southwestern frontier of China and comprises 16 prefectural-level administrative divisions (including prefecture-level cities and autonomous prefectures). Characterized by a predominantly plateau and mountainous topography, the province features extensive mountainous terrain. The landform generally descends stepwise from the higher northwest to the lower southeast. While Yunnan Province is abundant in hydrothermal resources, the distribution of these resources is uneven both temporally and spatially.
2.2. Data Sources
2.2.1. Sources and Processing of Vector Map Data
The vector administrative boundary data used in this study were obtained from the 1:250,000 National Fundamental Geographic Information Database, accessible through the National Catalogue Service for Geographic Information. The base data were produced in 2017. Subsequent administrative division adjustments within Yunnan Province have been incorporated by modifying the dataset accordingly. Finally, the data were projected and converted to the WGS84 coordinate system. Meanwhile, the geographic coordinates of the enterprises involved in the study were obtained from the Baidu Map Coordinate Picking System and were spatially adjusted to the WGS84 coordinate system using ArcGIS10.8.
2.2.2. Sources and Processing of Raster Map Data
The topographic raster data used in this study were sourced from the SRTM DEM 90M resolution original elevation dataset provided by the Geospatial Data Cloud. These data were then spatially rectified to the WGS84 coordinate system using ArcGIS10.8. Subsequently, they were masked and extracted within the boundary of the study area based on research requirements.
2.2.3. Sources and Processing of Statistical Data
The names, types, and quantities of green-certified enterprises within the study area involved in this research were obtained from the China Green Food Development Center. The studied enterprises were categorized into four major types—vegetables, grains, fruits, and tea—based on the product category they primarily produce. In cases where an enterprise was involved in multiple product categories, it was classified according to the category with the highest output volume. The precipitation data for various prefecture-level cities (autonomous prefectures) in Yunnan Province for 2024 were sourced from the Yunnan Province Water Resources Bulletin 2024, published by the Yunnan Provincial Department of Water Resources. Data on the annual Gross Domestic Product (GDP), registered household population, and total retail sales of consumer goods for prefecture-level cities (autonomous prefectures) in Yunnan for 2024 were obtained from the Yunnan Statistical Yearbook 2024, published by the Yunnan Provincial Bureau of Statistics. The annual accumulated temperature data for Yunnan Province were obtained from meteorological station records published by the National Oceanic and Atmospheric Administration (NOAA) of the United States.
All the aforementioned statistical data were imported into ArcGIS10.8, underwent spatial interpolation, and were subsequently rectified to the WGS84 coordinate system.
2.3. Research Methods
2.3.1. Kernel Density Estimation Model
The kernel density estimation is a method for analyzing spatial density distribution, which provides a smooth and intuitive representation of the spatial differentiation patterns of geographical phenomena and objects within geographic space. Its fundamental principle is that geographical events or objects are assumed to be possible at any location within the geographical space, yet their probabilities of occurrence vary across different positions. Areas with a high concentration of points correspond to a higher probability of event occurrence, whereas sparsely distributed points indicate a lower probability [14]-[16]. The kernel density estimation is defined as follows: Let the set of n points, X1, X2, X3, ..., Xn, be a sample drawn from a population with a distribution density function F. The kernel density estimate of F at a given point X is denoted as F(X). The expression is given by:
In the formula, F(X) represents the kernel density value at the spatial location of point X; h is the distance decay threshold; n denotes the number of feature points whose distance from point X is less than or equal to h; and k represents the spatial weight function. The geometric interpretation of this formula is that the kernel density value reaches its maximum at each core element ci and gradually decreases as the distance from ci increases, eventually falling to zero beyond a specific distance [16].
2.3.2. Geographical Detector
Geographical Detector is a spatial statistical method that serves as a tool for investigating, analyzing, and quantifying the spatial heterogeneity of geographical elements and phenomena, and for explaining the changes in geographical spatial entities along with their underlying driving factors. Its core theoretical premise is that if an independent variable exerts a significant influence on a dependent variable, then the two variables should exhibit similarity in their spatial distributions [17]. The geographical detector has been widely applied to the detection of spatial heterogeneity across various geographical phenomena [18]-[20]. Based on the objectives of this study, the factor detector within the Geographical Detector is employed to quantify and explain the extent to which objective geographical factors influence the spatial distribution of green food enterprises. Its mathematical expression is as follows:
In the formula, q represents the extent to which the selected independent variable factors influence the spatial distribution of green food-certified enterprises, with its value ranging from [0, 1]. The value of q closer to 1 indicates a stronger correlation between the independent variable factor and the spatial distribution of green food-certified enterprises, signifying a greater degree of influence. Conversely, the value of q closer to 0 reflects a weaker correlation and lesser influence. Here, L denotes the total number of selected independent variable influencing factors; Nh and N represent respectively the number of unit partitions in stratum h and the total number of unit partitions in the entire sample;
and σ2 are the variances of the Y values in stratum h and the entire sample, respectively; SSW and SST refer to the within-stratum sum of squares and the total sum of squares, respectively.
3. Results and Analysis
Using the Kernel Density Estimation model, we calculated the kernel density values for four categories of green-certified enterprises. Based on these values and utilizing ArcGIS10.8, we mapped the spatial distribution ranges of the kernel density for these four categories of green-certified enterprises. Furthermore, by integrating the geographic coordinates of the relevant enterprises, the spatial distribution of the four categories of green food-certified enterprises across Yunnan Province was visualized.
3.1. Overall Spatial Characteristics of Green Food-Certified Enterprises
As shown in Figure 1, green food-certified enterprises in Yunnan Province are primarily concentrated in regions such as central, eastern, and northeastern Yunnan—including Kunming, Yuxi, Zhaotong, Chuxiong, Honghe, and Qujing—demonstrating a clear trend of regional agglomeration. Along the corridor from
Figure 1. Overall spatial distribution of green food-certified enterprises in Yunnan province.
Dali to Dehong, green food enterprises exhibit a linear concentration pattern. Although the number of enterprises in this belt is relatively fewer compared to the main concentrated areas, it still demonstrates a trend of agglomerated distribution. In contrast, the number of green food-certified enterprises in northwestern Yunnan—including Diqing, Nujiang, and Lijiang—as well as in southwestern regions such as Lincang, Pu’er, and Xishuangbanna remains relatively low, and their spatial distribution is scattered, without forming a distinct agglomeration pattern.
3.2. Spatial Agglomeration Characteristics by Category
As shown in Figures 2-5, all categories of green food-certified enterprises exhibit strong spatial agglomeration within the study area, which is manifested geographically by high kernel density values in specific regions. Different categories of green food-certified enterprises exhibit distinct spatial clustering patterns across geographical locations. Green vegetable-certified enterprises show high kernel density values primarily concentrated in the Yuxi, Qujing, and Chuxiong regions, while their agglomeration is less pronounced in other areas. Green grain-certified enterprises exhibit notable spatial clustering in Chuxiong and Dehong, with relatively weaker agglomeration observed in a few other regions. Green fruit-certified enterprises show relatively pronounced agglomeration in Zhaotong (northeastern Yunnan), Chuxiong and Kunming (central Yunnan), as well as Yuxi and Honghe (southern Yunnan). Green tea-certified enterprises tend to cluster spatially in
![]()
Figure 2. Green vegetable-certified enterprises in Yunnan province.
Figure 3. Green grain-certified enterprises in Yunnan province.
Figure 4. Green fruit-certified enterprises in Yunnan province.
Figure 5. Green tea-certified enterprises in Yunnan province.
Baoshan, Lincang, Xishuangbanna, and Pu’er.
3.3. Spatial Correlation Analysis
Both natural geographical factors and socioeconomic factors can influence the location selection of food enterprises to a certain extent [21]-[25]. The distinctive topography, climatic conditions, and complex socioeconomic context of Yunnan Province may potentially affect the geospatial distribution of green food-certified enterprises in the region. Therefore, building on previous research [26]-[30] and taking into account the practical scope of this study as well as data availability, the following variables were selected as potential natural and social factors representing influences on the spatial distribution of green food-certified enterprises in Yunnan Province: topographic slope (X1), annual accumulated temperature for 2024 (X2), average annual precipitation for 2024 (X3), annual Gross Domestic Product (GDP) of prefecture-level cities (autonomous prefectures) for 2024 (X4), registered household population of prefecture-level cities (autonomous prefectures) for 2024 (X5), and total retail sales of consumer goods in prefecture-level cities (autonomous prefectures) for 2024 (X6).The dependent variables selected were: green vegetable-certified enterprises in Yunnan Province (Y1), green grain-certified enterprises (Y2), green fruit-certified enterprises (Y3), and green tea-certified enterprises (Y4).First, the relevant statistical data were imported into ArcGIS10.8 and then linked to the spatially rectified location points of prefecture-level cities (autonomous prefectures) in Yunnan Province. Subsequently, spatial interpolation was applied to the data associated with each location point, constrained within the extent of the study area. This process generated raster data, which were then further subjected to spatial correction. Finally, sampling points were generated at the pixel level for the derived raster data and the spatial aggregation data of the four categories of green food-certified enterprises, and sampling was performed accordingly. The sampling results were exported as a tabular file via ArcGIS10.8, followed by the removal of outliers. The results from the previous step were imported into the geographical detector, with the six selected influencing factors as independent variables and the spatial agglomeration data of the four categories of green food-certified enterprises as dependent variables. The model was then executed to obtain the results.
Based on the q statistic derived from the factor detector of the geographical detector, the degree of influence exerted by each independent variable on the spatial agglomeration patterns of various green food-certified enterprises was reflected.
Table 1. Geographical detector results of factors influencing the spatial distribution of green food-certified enterprises in Yunnan province.
|
X1 |
X2 |
X3 |
X4 |
X5 |
X6 |
Y1 |
0.89 |
0.91 |
0.95 |
0.73 |
0.61 |
0.83 |
Y2 |
0.84 |
0.93 |
0.97 |
0.78 |
0.67 |
0.82 |
Y3 |
0.85 |
0.96 |
0.98 |
0.83 |
0.72 |
0.87 |
Y4 |
0.86 |
0.92 |
0.94 |
0.74 |
0.65 |
0.81 |
MEAN |
0.86 |
0.93 |
0.96 |
0.77 |
0.6625 |
0.8325 |
As shown in Table 1, overall, all six influencing factors exhibit a significant effect on the spatial distribution of green food-certified enterprises, with the mean q value of each exceeding 0.6. Moreover, the average q-value of natural geographical influencing factors is higher than that of socio-economic independent factors, indicating that natural geographical conditions exert a greater influence on the spatial distribution of green food-certified enterprises compared to socio-economic conditions.
Within the natural geographical influencing factors, the magnitude of influence among the factors is ranked in descending order as: precipitation (X3) > temperature (X2) > topography (X1). This indicates that, among the various natural elements, precipitation exerts the most significant influence on the spatial distribution of green food-certified enterprises.
Among the socio-economic factors, the total retail sales of consumer goods in prefecture-level cities (autonomous prefectures) (X6) exert a greater influence on the spatial distribution of green food-certified enterprises compared to demographic factors (X5) and economic output factors (X4).
4. Conclusions and Suggestions
4.1. Conclusions
In 2024, green food-certified enterprises in Yunnan demonstrated a pronounced clustered distribution pattern within the geographical space. Among them, central Yunnan—including Chuxiong, Kunming, Yuxi, and Qujing—along with Zhaotong in northeastern Yunnan and Honghe in southeastern Yunnan, form the primary clusters of green food-certified enterprises. Areas such as Dali and Dehong in western Yunnan, along with Lincang and Xishuangbanna in southwestern Yunnan, constitute secondary clusters of green food-certified enterprises. Other regions are characterized by a relatively sparse presence of such enterprises.
Different types of green food-certified enterprises exhibit distinct spatial clustering patterns across geographical locations. Green vegetable-certified enterprises are predominantly clustered in the Yuxi and Qujing regions; green grain-certified enterprises are primarily concentrated in Chuxiong and Dehong; green fruit-certified enterprises show agglomeration in Zhaotong, Chuxiong, Yuxi, and Honghe. Green tea-certified enterprises are primarily clustered along the Chuxiong-Lincang-Dehong corridor, with limited agglomeration also observed in Pu’er and Xishuangbanna, albeit on a smaller scale.
The spatial agglomeration pattern of green food‑certified enterprises in Yunnan Province is shaped by multiple factors. Overall, natural geographical factors exert a stronger influence on their geospatial distribution relative to socioeconomic factors. Among the specific variables examined, precipitation shows a higher correlation with the distribution pattern compared to other influencing factors, suggesting that it may serve as a fundamental driver in shaping the spatial layout of these enterprises in the province. Meanwhile, socioeconomic factors, particularly market scale, also contribute to influencing the geospatial distribution pattern of green food-certified enterprises in Yunnan to some extent.
Studying the geospatial distribution of green food‑certified enterprises holds significant implications. First, based on the findings of this research, government departments at various levels can utilize the basic spatial distribution pattern of such enterprises to establish a provincial green food industry chain. This would allow for financial support and policy assistance to areas with a low agglomeration level of green food‑certified enterprises, while encouraging enterprises in highly concentrated zones to relocate to less concentrated regions. As a result, the spatial layout of green food‑certified enterprises across the province can be optimized. Second, government departments can formulate annual regional economic development goals and policies based on the results of this study. Finally, relevant green food‑certified enterprises can use the research findings, in conjunction with their own practical circumstances, to identify suitable development space.
4.2. Suggestions
Although the above analysis has provided insights into the geospatial agglomeration of green food-certified enterprises in Yunnan Province and its influencing factors, this study still has certain limitations. First, the data selected for this research were limited to the year 2024, which precludes an analysis of the temporal dynamics and evolution of green food-certified enterprises in the province. Second, this study does not encompass all potential factors that could influence the distribution of such enterprises in Yunnan. Variables such as regional soil type distribution, transportation infrastructure, and government economic and development policies may also affect the spatial layout of green food-certified enterprises, but they were not included due to significant challenges in data acquisition. These limitations are expected to be addressed in subsequent related research.