Production Agglomeration and Spatiotemporal Evolution of China’s Fruit Industry over the Last 40 Years

Abstract

This study analyzes the dynamics of China’s fruit industry using a range of analytical tools, including the location Gini coefficient, industry concentration ratio, spatial autocorrelation index, specialization index, and the industry gravity model. It explores the industry’s evolving characteristics and trends since the economic reforms, culminating in a trajectory map that highlights shifts in the industry’s gravitational center. This study also offers a qualitative analysis of the factors influencing the agglomeration and relocation of fruit production centers. The findings show a steady increase in both total output and yields per unit area within China’s fruit industry over time. Although the overall degree of agglomeration has decreased, regional agglomeration effects remain significant. Furthermore, the data reveal significant spatial autocorrelation in fruit production, indicating a long-term westward shift in core production areas. Different geographic areas exhibit varying levels of gradational shifts, with marked differences in production concentration patterns across different fruit types. This study provides a comprehensive framework for understanding production agglomeration, integrating interdisciplinary methods from statistics and geography.

1. Introduction

As the global population increases and consumption expands [1], fruits play an increasingly vital role in human life [2]. The development of the fruit industry primarily aims to enhance global food security while also having the potential to stimulate rural growth and generate significant economic benefits for farmers [3]. China, as a major global producer and consumer of fruit, is characterized by its diverse varieties and substantial production volume. Data from the National Bureau of Statistics of China, https://data.stats.gov.cn/ (accessed on 24 February 2025), show that both the area dedicated to fruit cultivation and the output of fruit production have consistently increased. In 2022, the per capita consumption of fresh and dried fruits and vegetables in China reached 54.7 kg. The area covered by orchards was approximately 13.0 million hectares, representing a year-on-year growth rate of 1.6%. Furthermore, in 2023, China’s fruit production reached approximately 327.4 million tons, reflecting an annual increase of 4.6%.

In recent decades, as China’s consumption structure has evolved and consumption patterns have changed, structural contradictions between the supply and demand of fruit have gradually emerged. Fruit consumption preferences have notably shifted, transitioning from an emphasis on affordability to prioritizing health benefits and environmental sustainability [4,5]. Additionally, the focus has shifted from the merely functional consumption of supplemental nutrition to quality consumption [6], emphasizing the origin, variety, and branding of fruits. Meanwhile, the low-end fruit market faces overcapacity and regional imbalances between production and consumption [7], leading to recurring stagnation in major fruit-producing regions. Relevant studies have shown that these contradictions primarily originate from the irrational layout of fruit plantations, a scarcity of suitable planting areas for certain fruit varieties, an underdeveloped industrial sector, and policy-related factors [7,8]. In addition, fruits are highly dependent on natural conditions, have a limited storage lifespan, and are perishable [9]. Unlike other agricultural commodities, fruits differ markedly in their biological traits, storage and preservation requirements, transportation requirements, and production agglomeration patterns. Given the existing constraints, there is an urgent need to develop a multifaceted food supply chain and exploit diverse food sources through multiple channels to ensure the sustainable development of the fruit industry. Therefore, the spatial and temporal characteristics and evolutionary trends of production agglomeration in China’s fruit industry should be clarified to further optimize the layout of fruit production and explore countermeasures for the industry’s future development. These steps can promote the safety and stability of the fruit industry. This process is deeply intertwined with the Chinese path to modernization, providing a scientific basis for improving the fruit industry and cropping framework within the agricultural sector. Furthermore, these factors not only significantly influence the global equilibrium between fruit supply and demand but also promote high-quality agricultural development [10].

The concept of industrial agglomeration, initially introduced by Marshall, is typically understood as the extent of spatial closeness between related activities within the same sector [11,12]. The concept of ”production agglomeration” in the context of the fruit industry pertains to a high concentration of fruit cultivation, processing, logistics, and sales within a specific geographical area. This concentration gives rise to economies of scale, a specialized division of labor, and technological spillover, commonly referred to as the clustering effect. As research progresses, the definition of production agglomeration is expanding beyond its initial scope to encompass spatial expansion, the deepening of the industrial chain, the integration of production factors, and dynamic evolution. Production agglomeration is classified based on driving factors and can be divided into natural and policy-driven dimensions. Natural production agglomeration is driven by long-term environmental conditions (e.g., climate, soil, and water), which constrain it within geographically suitable agricultural zones and render it relatively stable over time. Conversely, policy-driven agglomeration results from government intervention, including tax incentives, subsidies, and infrastructure investment. Such interventions potentially mitigate natural restrictions, for example, by establishing facility-based agricultural parks. However, policy-driven agglomeration is relatively unstable, prone to fluctuations or dissolution in response to policy changes, and primarily designed to promote balanced regional development. Moreover, it heavily relies on technological advancements.

The study of agricultural industrial agglomeration has emerged as a major focus in industrial agglomeration research. This phenomenon has been extensively explored by numerous researchers, using diverse arrays of theoretical models and analytical methodologies. For example, Madhukar et al. (2020) utilized spatial visualization techniques and statistical approaches to analyze the distribution patterns of three major crops in India: wheat, rice, and maize [13]. Wang et al. (2018) examined the spatiotemporal changes in vegetable production across China by utilizing adjusted Gini coefficients and Moran’s I index [14]. Additionally, the Spatial Durbin Model was applied to explore the factors driving spatiotemporal evolution. Xiao et al. (2018) investigated the trends and driving factors influencing the spatiotemporal evolution of tea production in China, focusing on the periods before and after the country’s accession to the World Trade Organization [15]. This analysis was conducted using the production concentration index model and theory of the industrial center of gravity. Furthermore, certain researchers have examined the spatial dynamics of grain output growth in China post-2003, along with its implications at the county level. They applied the production center of gravity model and exploratory spatial data analysis [16].

Significant convergence was observed in the metrics and methodologies applied in the agricultural industry agglomeration field. The majority of scholarly works utilize a range of indicators to evaluate the extent of agglomeration within industries, including metrics like the rate of industrial concentration, the Herfindahl Index [17,18], the Gini coefficient [19], Moran’s I coefficient [20], and location entropy [21]. Such measures are frequently employed to determine industrial concentrations. The choice of research subjects, datasets, and analytical techniques varies significantly between scholars, thereby providing a robust foundation for this study. Furthermore, when examining the spatiotemporal dynamics of the fruit industry, certain researchers have centered their research on specific types of fruits [10,22,23]. Several studies have examined the spatiotemporal trends and patterns of agricultural production in various regions. For instance, Svobodová et al. [24] conducted an in-depth examination of vineyards in the Czech Republic and advocated holistic ways to develop integrated planting and organic viticulture. Kerry et al. (2017) delved into the specifics of cranberry cultivation in New Jersey agricultural zones in the U.S., employing techniques such as local clustering analysis and geographically weighted regression to understand production dynamics [25]. Zhang et al. (2020) embarked on a detailed exploration of the evolving trends in apple production across various regions in China, utilizing the apple production concentration index and spatial econometric models to analyze these patterns [26]. Lin et al. (2022) studied the main factors influencing citrus production distribution in Sichuan Province, China, and developed a spatial simulation method for citrus production under different scenarios in 2025 [27]. However, the temporal spans selected by prior analyses vary considerably, predominantly focusing on individual fruit types. Consequently, this approach results in an incomplete representation of the Chinese fruit sector’s overall characteristics. Furthermore, there is a noticeable gap in the literature concerning fruit industry agglomeration within specific geographical areas. Accordingly, this study undertakes a comprehensive and rigorous analysis of the following research questions:

(1) What are the characteristics of production agglomeration in China’s fruit industry since 1978, and what are the trends in the spatiotemporal evolution of its spatial layout? (2) What are the degrees of production agglomeration and spatial distribution characteristics of China’s five principal fruits: oranges, pears, grapes, bananas, and apples? (3) Considering the factors influencing the dynamic changes in fruit production agglomeration in China, what measures can be taken to ensure the future sustainable development of its fruit industry? What lessons can be drawn from this study for developing the fruit industry globally and in other major fruit-producing countries?

Building on the foundational theories of industrial layout and agglomeration, this study employs a comprehensive suite of five measurement methods to analyze data related to fruit planting and production in China from 1978 to 2021. Utilizing advanced tools such as the Geographic Information System (GIS) and Python 3.9, we delve into the spatiotemporal characteristics and trends of China’s fruit industry. A multifaceted analytical approach is adopted, employing various indicators to provide insights into the spatial and temporal development of the industry, including the locational Gini coefficient, the spatial autocorrelation index, and the industry center of gravity model. This study aims to investigate the various factors contributing to the expansion of China’s fruit industry, offering both theoretical and practical implications.

2. Materials and Methods

2.1. Data Source and Study Area

This research focuses on China’s fruit sector, analyzing the industry along with the cultivation and production dynamics of five principal fruits—oranges, pears, apples, grapes, and bananas—from 1978 to 2021. It aims to discern agglomeration patterns within the Chinese fruit industry. Given the unavailability of certain datasets, the analysis of the area under cultivation was confined to the period between 1978 and 2020. To ensure temporal consistency, the data for certain analyses were truncated to 2020. However, the 2021 data were retained for consecutive-year trend analysis to enhance the robustness of the findings. This study draws upon data from authoritative sources, including the China Statistical Yearbook, Provincial Statistical Yearbooks, the China Rural Statistical Yearbook, and a comprehensive publication titled “Sixty Years of Agricultural Statistics in New China”. For sporadic missing values from 1978 to 2020, this study employs linear interpolation and the neighborhood mean method to impute the missing data. With respect to the selection of the sample scope, two points should be noted: First, at present, China has 34 provincial-level administrative regions, including 23 provinces, 5 autonomous regions, 4 municipalities directly under the central government, and 2 special administrative regions. However, this study does not include the Hong Kong, Macao, and Taiwan regions of China due to the absence of pertinent data and the difficulty in accessing it. Second, this study categorizes the regions into four major areas–Eastern, Central, Western, and Northeastern China–based on the standard economic regional division of China. The specific details are as follows: the Eastern region includes Beijing, Tianjin, Hebei, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and Hainan; the Central region includes Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan; the Western region comprises Inner Mongolia, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang; and the Northeast region includes Liaoning, Jilin, and Heilongjiang. In addition, to ensure data comparability, the relevant data for Hainan Province and Chongqing Municipality are categorized under Guangdong Province and Sichuan Province, respectively. This classification is necessary because Hainan Province was separated from Guangdong and established as an independent province in 1988, while Chongqing Municipality was separated from Sichuan Province and designated as a direct-controlled municipality in 1997. Therefore, the research areas involved in this paper include a total of the aforementioned 29 regions.

2.2. Research Methods

Considering the appropriateness of these methodologies and the accessibility of pertinent data, this study adopts five distinct approaches to evaluate the degree and characteristics of production agglomeration, spatial heterogeneity, and interdependence within China’s fruit industry. These methods include the Gini coefficient of location, the industry concentration rate, the spatial autocorrelation index, the specialization index, and the industry center of gravity model.

Specifically, the Gini coefficient of location is used to measure the spatial disparities and overall agglomeration of the fruit industry. The degree of spatial concentration and inter-regional differences was quantified based on the industry concentration rate and specialization index. Given certain mutual influences and connections between study regions, global and local spatial autocorrelation coefficients were employed to analyze the spatial correlation characteristics between regions. We employed these three methods because the Gini coefficient identifies global imbalance, industrial concentration measures changes in regional industrial distribution over time, and spatial autocorrelation (Moran’s I) identifies synergistic agglomeration effects in neighboring regions. Combining these three indices helps overcome the limitations of using a single index. Specifically, the Gini coefficient cannot distinguish between “unipolar agglomeration” and “polycentric agglomeration”, while spatial autocorrelation provides complementary insights into these patterns. Furthermore, the industrial center of gravity model was used to intuitively depict the shifting trajectory of the production center of gravity within the fruit industry.

2.3. The Locational Gini Coefficient

The Gini coefficient, proposed by Italian economist Corrado Gini in 1921 [28], is used to quantitatively determine the degree of imbalance in income distribution [29,30,31]. Numerous scholars have employed this framework to investigate issues related to industrial geographic agglomeration. The calculation formula is shown in Equation (1).

$${GINI}_i={\displaystyle \frac{1}{2m^2\mu }}{\sum }_{k=1}^m{\sum }_{h=1}^m\left|x_k-x_h\right|$$

(1)

In Equation (1), variable i represents the different sectors of the industry, ranging from 0 to 5. Specifically, i = 0 represents the fruit industry and i = 1 represents another sector. The numbers 1–5 symbolize the five primary fruit products, namely oranges, pears, grapes, bananas, and apples, in sequential order. The term ${GINI}_i$ is the locational Gini coefficient for the fruit industry or production of a specific fruit. Variable m indicates the number of provinces involved, whereas k and h represent distinct provinces (with k ≠ h). $x_k$ and $x_h$ denote the proportion of the planted area of the fruit industry or a specific fruit industry in province k and province h, respectively, in relation to the corresponding kinds of fruit-planted areas in the country. Furthermore, $\mu$ represents the average value of the planted area of the fruit industry or a specific fruit industry in each province in relation to the proportion of the planted area in the corresponding fruit-planted area in the country. The locational Gini coefficient for the fruit industry is a metric ranging from [0 to 1], where a higher value signifies increased geographic agglomeration within the industry.

2.4. Industrial Concentration Ratio

The Gini coefficient of location plays a significant role in assessing the balance of industrial distribution across regions but exhibits certain limitations in aspects such as enterprise size and the setting of geographical units [32,33,34]. The industry concentration rate indicates the combined share of the leading provinces in terms of the industrial production scale within the national context. This metric effectively mirrors the agglomeration level and its fluctuations among the top-ranking provinces in terms of the fruit industry’s production scale. The specific formula is shown in Equation (2).

$${CR}_m^i={\sum }_{k=1}^m{x}_k^i\times 100\%$$

(2)

In Equation (2), variable i delineates the different sectors of the industry, with values ranging from 0 to 5. Specifically, i = 0 represents the fruit industry, whereas the values i = 1 through i = 5 correspond to the orange, pear, grape, banana, and apple industries, respectively, highlighting the five principal fruit categories. Variable m represents the number of provinces commonly selected as 1, 3, or 5. The present study emphasizes observing shifts in the concentration level of the fruit industry within the five leading provinces. Consequently, parameter m was assigned a value of 5. $x_k^i$ is defined as the share of the planting area dedicated to the fruit industry or a specific fruit variety in province k of the total planting area for that fruit category across the nation. The industry concentration rate ranged from 0 to 1. A higher value indicates a larger average national market share of the fruit industry in a given province, signifying a greater degree of geographic agglomeration.

2.5. Spatial Autocorrelation Index

Spatial autocorrelation analysis methods are categorized into global and local spatial autocorrelations. The former primarily utilizes global Moran’s I to depict the spatial clustering of pertinent variables across the entire study area, with the degree of spatial agglomeration reflected through data from neighboring provinces. The latter serves as a complement to global autocorrelation analysis, providing a more detailed and specific quantification of spatial relationships [16,35,36,37].

To delve deeper into the spatial correlation of production agglomeration within China’s fruit industry, this study employed a global spatial autocorrelation analysis at the provincial level. The formula for calculating Moran’s I index for a specific fruit is presented in Equation (3).

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

(3)

$$\overline{x}={\displaystyle \frac{1}{n}}{\sum }_{j=1}^n{x}_i$$

(4)

In Equations (3) and (4), variables $x_i$ and $x_j$ denote the production of a specific fruit in provinces i and j, respectively. These variables represent the average fruit production. Variable $w_{ij}$ represents the spatial weight, highlighting the proximity between provincial units i and j, determined according to the neighborhood criterion. Variable n denotes the total number of provincial units. The weight $w_{ij}$ is assigned a value of 1 when provinces i and j are adjacent to each other and 0 otherwise. A positive I value indicates spatial distribution similarity in the fruit industry output across adjacent or nearby provinces or districts. Conversely, a negative I value suggests a lack of spatial distribution similarity in fruit industry output between adjacent or nearby provinces, indicating that similar industries are not agglomerated in these regions.

Analyzing local spatial autocorrelation reveals whether high or low output values cluster in the fruit industry across province i and its neighboring provinces within the entire region. The formula for the local Moran’s I index is presented in Equation (5), where S denotes the standard deviation of the fruit industry output across provinces. The meanings of the other parameters are consistent with those used in the global spatial autocorrelation analysis.

$$I_i={\displaystyle \frac{(x_i-\overline{x})}{S^2}}{\sum }_j{w}_{ij}(x_i-\overline{x})$$

(5)

2.6. Regional Specialization Index

The regional specialization index is an effective measure for assessing the level of agglomeration within the agricultural industry [38]. Fruit production is contingent on immobile factors such as climate and soil, and the specialization index effectively captures the fundamental role of natural conditions in agglomeration. Using mathematical models, the spatial distribution of the industry is quantified into comparable metrics, thereby identifying the core and peripheral production areas of the fruit industry. This provides a foundation for targeted policy investments. This study utilizes location entropy to determine the specialization index of fruit j in a given province and explore the distribution of specialization across provinces within the fruit industry. The formula is as follows:

$$Q_j={\displaystyle \frac{E_j/E_t}{P_j/P_t}}$$

(6)

In Equation (6), the variable $Q_j$ represents the locational entropy of fruit j within a specific province (where j ranges from 1 to 5, corresponding sequentially to the five main fruit types: orange, pear, grape, banana, and apple). $E_j$ represents the output of fruit j within a specific province and $E_t$ denotes the total output of the fruit industry in that province. $P_j$, on the other hand, represents the output of fruit j in the entire country, whereas $P_t$ signifies the total output of the fruit industry in the entire country. If $Q_j$ exceeds 1, the specialization level of fruit j in the province surpasses the national average. If $Q_j$ is equal to 1, the level of specialization of fruit j in the province is comparable to the national average. If $Q_j$ is less than 1, the level of specialization of fruit j in the province falls below the national average. Furthermore, the concentration of high values in a limited number of provinces indicates relative agglomeration within the fruit industry.

2.7. Industrial Center of Gravity Model

The industrial center of gravity refers to the point in geographical space where the distribution of the production scale across regions achieves a balance in terms of torque at a given time. The industrial center of gravity model is a crucial analytical tool for examining spatial v ariations in elements during regional development processes [16,39]. This model effectively illustrates spatial concentration and dispersion trends in China’s fruit industry across production spaces. It also clearly delineates the trajectory of the center of gravity deviation, offering an intuitive depiction of the detailed process of regional changes in fruit production. Equations (7)–(9) represent the model expressions.

$${\overline{X}}_j={\sum }_{i=1}^n(P_{ij}·x_i)/{\sum }_{i=1}^n{P}_{ij}$$

(7)

$${\overline{Y}}_j={\sum }_{i=1}^n(P_{ij}·y_i)/{\sum }_{i=1}^n{P}_{ij}$$

(8)

$$d_{\alpha ~\beta }=c·\sqrt{{(X_{\alpha }-X_{\beta })}^2+{(Y_{\alpha }-Y_{\beta })}^2}$$

(9)

where ${\overline{X}}_j$ and ${\overline{Y}}_j$ represent the longitude and latitude values, respectively, of the geographical location of the center of gravity of fruit production in China in the jth year. Variable n denotes the number of provincial units; $P_{ij}$ represents fruit production in province i in the jth year; and $x_i$ and $y_i$ represent the latitude and longitude, respectively, of the center of gravity of the ith geographic unit. The geographic coordinates of the provincial capital city of each province were selected for this purpose. $d_{\alpha ~\beta }$ represents the distance by which the center of gravity has moved in the year, $\alpha ~\beta$(unit: km), and the different years. The coordinates ($X_{\alpha }$,$Y_{\alpha }$) and ($X_{\beta }$,$Y_{\beta }$) denote the geographic coordinates (longitude and latitude values) of the center of gravity of Chinese fruits in the first year. The constant c is typically assigned a value of 111.111, representing the coefficient for converting geographic coordinate units into distances in kilometers for every 1° of change.

3. Results

3.1. Spatiotemporal Evolution of Degree of Production Agglomeration

Employing the locational Gini coefficient and industry concentration rate, this study evaluates the geographical agglomeration within the fruit industry, uncovering the characteristics and trends of agglomeration. By adopting the share of fruit cultivation area in each province relative to the nationwide total as weights, the analysis yielded a weighted Gini coefficient for the fruit industry in China, revealing a predominantly high degree of agglomeration (

Table 1. Current status of geographic agglomeration of fruit industry production in China, 2020.

Categories	${\mathit{G}\mathit{I}\mathit{N}\mathit{I}}_{\mathit{i}}$	Top Five Provinces and Their Weights
		Top Five Provinces	${\mathit{C}\mathit{R}}_5$ (%)
Fruit Industry	0.4943	Guangxi, Guangdong, Sichuan, Shaanxi, Xinjiang	45.70
Orange	0.6087	Guangxi, Sichuan, Hunan, Jiangxi, Guangdong	75.70
Pear	0.5199	Hebei, Sichuan, Liaoning, Xinjiang, Henan	47.63
Apple	0.6522	Shaanxi, Gansu, Shandong, Shanxi, Liaoning	70.17
Grape	0.4630	Xinjiang, Shaanxi, Sichuan, Hebei, Henan	42.53
Banana	0.7810	Guangdong, Yunnan, Guangxi, Fujian, Guizhou	99.39

). In 2020, the banana production sector was distinguished by the highest locational Gini coefficient at 0.7810, followed by apples at 0.6522, oranges at 0.6087, and pears at 0.5199. Grapes registered the lowest coefficient at only 0.4630. The weighted locational Gini coefficient for the entire fruit industry was 0.4943. The production of China’s fruit industry exhibits a high degree of geographical agglomeration. The values for the five major fruit crops were all significantly high, with bananas reaching up to 99.39%. Only the value for grapes (42.53%) falls below the overall level of the fruit industry (45.70%). Integrating the measurements from the locational Gini coefficient and industry concentration rate reveals a high degree of consistency between these two indicators.

From a geographical perspective, the aggregation patterns across China’s four major regions are characterized by distinct regional gradient features. The proportion of the fruit planting area was sequenced from the eastern, central, western, and northeastern regions (

Table 2. Percentage change in area under fruit cultivation in China (%).

	Eastern Part			Central Part			Western Part			Northwest Part
	1978	2020	Alteration	1978	2020	Alteration	1978	2020	Alteration	1978	2020	Alteration
Fruit Industry	39.59	25.93	−9.66	22.56	18.85	−3.71	21.24	51.84	30.6	16.62	3.37	−13.25
Orange	27.92	17.31	−10.61	41.30	35.23	−6.07	30.77	47.47	16.7	----	----	-----
Pear	38.45	27.99	−10.46	23.16	24.23	1.07	13.17	37.70	24.53	25.22	10.07	−15.15
Apple	38.55	20.57	−17.98	21.84	13.77	−8.07	18.67	57.88	39.21	20.94	7.78	−13.16
Grape	22.55	23.22	0.67	24.56	19.16	−5.4	40.77	50.90	10.13	12.11	5.71	−6.4
Banana	50.00	47.36	−2.64	----	----	----	50.00	52.64	2.64	----	----	----

). Observing the numerical trends of major fruits, the spatial changes in China’s fruit industry production reveal a developmental trend toward aggregation in the western regions, with an overall decline in the high agglomeration levels and regional monopolistic trends. Specifically, in 1978, the top five provinces in China in terms of the fruit planting area (in order, Shandong, Liaoning, Hebei, Henan, and Shaanxi) accounted for 56.06% of the country’s total planting area. By 2020, the top five provinces (in order, Guangxi, Guangdong, Sichuan, Shaanxi, and Xinjiang) saw a 45.70% decrease in the planting area. When examining the production dynamics across different regions, the western regions of China have seen a substantial expansion in the fruit industry’s cultivation area, representing 30.6% of the total national fruit planting area. Currently, western China stands out as the leading region in terms of fruit planting area, constituting a major agglomeration hub for the industry at large, as well as specific fruit sectors. The apple industry has undergone the most significant transformation, with the planting area for apples in the western regions increasing from 13.17% of the total apple planting area in 1978 to 50.90% in 2020. Additionally, possibly due to natural climate conditions, the banana industry’s planting area has remained relatively stable. The main banana-producing areas in China (e.g., Guangdong and Guangxi) have significantly increased their yield per unit in hectares through facility-based farming technology, resulting in greater production concentration. However, the cultivated area has reached saturation and is even declining in some regions due to land constraints. This phenomenon is consistent with the concept of “intensive agglomeration”, suggesting that industrial agglomeration evolves from a “scale expansion” phase to an “efficiency-driven” phase. Consequently, the Gini coefficient and the area share indicators exhibit divergence during the deviation stage.

Changes in the per unit yield of fruits can affect production agglomeration [40]. In this study, we calculated the per unit yields of five major fruits and illustrated their spatiotemporal variations (Figure 1). The conclusions indicate that, over time, both the numerical yields and spatial distribution of the five types of fruit have significantly shifted, signifying a gradual increase in the per unit yield of China’s fruit industry. In terms of spatial distribution, the gray regions on the yield maps have diminished over time, signifying a reduction in provinces that do not cultivate certain fruits. Furthermore, a comparison of the changes in regional colors revealed a trend in the overall production of oranges, apples, grapes, and bananas moving from east to west. Additionally, provinces with higher yields exhibit a certain level of agglomeration, where several provinces within a specific area yield significantly higher yields than those in other provinces. The agglomeration effect is particularly pronounced in the eastern and central regions. Owing to numerous similarities in natural geography and socio-historical contexts, when adjacent provinces experience significant improvements in per unit yield, diffusion phenomena based on demonstration, learning, and imitation occur, generating positive externalities. Naturally, such regional agglomeration effects are influenced by various factors, including economic development, market variations, orchard infrastructure development, and the enhancement of fruit farmers’ skills.

3.2. Spatiotemporal Evolution of Spatial Heterogeneity and Dependency

3.2.1. Global Spatial Autocorrelation Analysis

This study calculates the global and local spatial autocorrelation and specialization indices of China’s fruit industry from 1978 to 2020. We aim to gain a better understanding of the agglomeration levels of different types of fruits across various provinces, with the specific results presented in

Table 3. Autocorrelation results of provincial fruit production distribution in China from 1978 to 2020.

Year	Moran’s I	Z-Score	p-Value
2020	0.161	1.626	0.052
2018	0.151	1.548	0.061
2013	0.200	1.989	0.023
2008	0.233	2.343	0.010
2003	0.236	2.606	0.005
1998	0.298	3.212	0.001
1993	0.035	0.626	0.266
1988	0.050	0.780	0.218
1983	0.132	1.644	0.050
1978	0.175	2.042	0.021

. The findings indicate that, with a few exceptions, China’s fruit industry consistently exhibits significant positive spatial autocorrelation. This suggests a similarity in the spatial distribution of fruit production among neighboring regions. Meanwhile, the global Moran’s I values were positive, indicating that provinces with high levels of fruit production tended to form agglomerations with high values. Reflecting changes in numerical values, the Moran index demonstrates a pattern of initial decrease, followed by an increase, and then another decline, overall indicating a trend toward strengthened spatial autocorrelation and a pronounced agglomeration effect.

3.2.2. Local Spatial Autocorrelation Analysis

Subsequently, utilizing local spatial autocorrelation analysis and Moran scatterplot analysis, potential spatial agglomerations among the provinces were investigated, as depicted in Figure 2. The local Moran scatterplot encompasses four quadrants: the first quadrant represents high–high clustering, indicating areas of high observation values surrounded by similarly high value regions; the second quadrant signifies low–high clustering, where regions of low observation values are encompassed by high value areas; and the third and fourth quadrants correspond to low–low and high–low clustering, respectively. The Moran scatterplots from 1978 to 2020 reveal a gradual dispersion of provinces from the low–low clustering area to the other three quadrants, with a notable increase in the number of provinces within the high–high clustering area. Regionally, provinces in the high–high clustering area are primarily located in the central and eastern parts of China, whereas almost all provinces in the low–low clustering area are in the northeastern and western regions.

3.2.3. Specialization Index Analysis

In addition, this study calculated the specialization index for five major fruits from 1978 to 2021, as detailed in Figure 3. Specifically, although the specialization index for the orange industry shows an overall declining trend, provinces with higher degrees of specialization gradually form relatively concentrated regional agglomerations. Hebei and Qinghai provinces have long been at the forefront of pear production specialization in China, yet the overall specialization level in pear production has significantly decreased, evolving toward a more widespread distribution. Additionally, the number of provinces where apple production specialization exceeds the national average has increased, as shown in the third row of Figure 3, where red areas converge in Gansu, Shaanxi, and other regions. Specialization in grape production has undergone notable changes; however, Xinjiang’s grape industry has consistently maintained a high level of specialization, ranking first nationwide for over 40 years, with industry clustering showing southward movement. Banana production is primarily concentrated in several southern provinces, with a relatively high concentration.

Given that the aforementioned specialization indices do not account for the absolute scale of the industries within each region, regions with high specialization indices could still have small scales. To circumvent potential analytical biases, this study compares the top five provinces by fruit industry output in 1978 and 2021 with the top five provinces ranked by the specialization index (

Table 4. Top five provinces and their changes in China’s fruit production and specialization index, 1978–2021.

	1978		2021
	Production	Specialization Index	Production	Specialization Index
Fruit Industry	Shandong, Liaoning, Hebei, Henan, Shaanxi (61.86%)		Guangxi, Shandong, Guangdong, Henan, Shaanxi (44.16%)
Orange	Sichuan, Guangdong, Zhejiang, Guangxi, Fujian (86.48%)	Jiangxi, Zhejiang, Sichuan, Hunan, Guangdong (6.96)	Guangxi, Sichuan, Hunan, Hubei, Guangdong (74.95%)	Jiangxi, Hunan, Fujian, Guangxi, Hubei (2.84)
Pear	Hebei, Shandong, Liaoning, Jiangsu, Shanxi (69.44%)	Qinghai, Shanghai, Anhui, Jiangsu, Hebei (2.47)	Hebei, Xinjiang, Henan, Liaoning, Sichuan (50.13%)	Hebei, Qinghai, Anhui, Liaoning, Tianjin (2.81)
Apple	Shandong, Liaoning, Hebei, Henan, Shaanxi (84.43%)	Tibet, Liaoning, Shandong, Ningxia, Inner Mongolia (1.80)	Shaanxi, Shandong, Gansu, Shanxi, Henan (76.11%)	Shaanxi, Gansu, Shanxi, Shandong, Liaoning (2.80)
Grape	Xinjiang, Shandong, Liaoning, Henan, Hebei (74.83%)	Xinjiang, Anhui, Inner Mongolia, Tianjin, Ningxia (6.83)	Xinjiang, Hebei, Shandong, Yunnan, Shaanxi (50.71%)	Xinjiang, Tianjin, Tibet, Shanghai, Zhejiang (2.78)
Banana	Guangdong, Yunnan, Guangxi, Fujian, Guizhou (100%)	Guangdong, Yunnan, Fujian, Guangxi, Tibet (5.32)	Guangdong, Guangxi, Yunnan, Fujian, Guizhou (99.54%)	Guangdong, Yunnan, Guangxi, Fujian, Tibet (3.11)

Note: The shares of production in the top five provinces in the national total and average of the specialization indices of the five provinces, respectively, are indicated in the parentheses.

). This comparison reveals a pattern of specialized agglomeration within the provincial distribution of the Chinese fruit industry. In terms of production volume, the proportion of fruit production in the top five provinces significantly decreased, with Liaoning and Hebei provinces experiencing substantial declines in their roles in fruit production. By contrast, regions such as Guangdong and Guangxi have leveraged and expanded their geographical and climatic advantages to form production agglomerations of advantageous fruits, thereby playing a significant role in the development and expansion of China’s fruit industry. From the specialization index perspective, many provinces with high production volumes do not necessarily exhibit a high degree of specialization. Taking grape production in 2021 as an example, the top five provinces in terms of output are Xinjiang, Hebei, Shandong, Yunnan, and Shaanxi; however, only Xinjiang ranks within the top five for the specialization index. Additionally, the average specialization index for oranges, grapes, and bananas among the provinces ranked in the top five experienced a significant decline, indicating a notable reduction in specialization levels. However, there is a trend toward increased local monopolization in the production of certain fruits, with notable examples being apple production in Shaanxi, grape production in Xinjiang, and banana production in Guangdong, where the current level of monopolization is relatively high.

3.3. Spatial Evolution Trajectory of Center of Gravity

By calculating the coordinates of the center of gravity, movement distance, and direction for China’s fruit industry (

Table 5. Coordinates of center of gravity and distance shifted for citrus, pear, and apple production in China, 1978–2021.

Year	Orange Industry				Pear Industry				Apple Industry
	Longitude	Latitude	Distance	Orientation	Longitude	Latitude	Distance	Orientation	Longitude	Latitude	Distance	Orientation
1978	111.73	27.57			115.89	36.40			117.48	38.41
1979	112.24	27.93	69.40	northeast	116.08	36.21	29.88	southeast	117.60	38.11	35.55	southeast
1980	110.53	27.87	190.74	southwest	115.03	35.39	148.06	southwest	117.25	38.18	39.75	northwest
1981	111.47	27.83	105.32	southeast	115.10	35.94	61.52	northeast	117.22	38.04	16.39	southwest
1982	111.75	27.58	41.07	southeast	114.90	35.72	33.11	southwest	116.87	38.08	39.45	northwest
1983	111.17	28.34	105.96	northwest	115.10	36.08	46.05	northeast	116.91	37.95	15.55	southeast
1984	111.24	27.81	58.86	southeast	114.62	35.96	55.20	southwest	116.63	38.08	35.11	northwest
1985	111.56	27.95	38.56	northeast	114.49	36.07	19.37	northwest	115.76	37.86	99.46	southwest
1986	112.11	27.57	74.11	southeast	114.45	35.95	13.78	southwest	115.94	37.81	20.84	southeast
1987	112.57	27.66	52.88	northeast	114.50	35.97	6.12	northeast	116.00	37.69	15.50	southeast
1988	112.45	26.65	113.09	southwest	114.38	35.92	14.40	southwest	115.81	37.73	21.79	northwest
1989	112.86	27.54	108.36	northeast	114.16	35.78	29.85	southwest	115.85	37.76	5.75	northeast
1990	112.98	26.99	61.78	southeast	114.09	35.97	22.46	northwest	116.21	38.11	55.67	northeast
1991	113.32	27.20	43.68	northeast	114.13	36.09	14.36	northeast	116.31	38.04	13.23	southeast
1992	112.70	26.75	84.97	southwest	114.18	36.08	5.59	southeast	116.88	38.24	67.53	northeast
1993	113.02	27.40	80.15	northeast	114.38	36.13	22.70	northeast	117.02	38.22	15.24	southeast
1994	113.49	27.53	54.92	northeast	114.39	36.24	12.74	northeast	116.97	38.12	11.57	southwest
1995	113.67	27.84	39.66	northeast	114.54	36.30	18.34	northeast	117.03	38.10	7.37	southeast
1996	113.80	27.87	14.43	northeast	114.95	36.47	49.14	northeast	117.18	38.12	16.25	northeast
1997	113.93	27.99	20.05	northeast	114.70	36.23	38.44	southwest	116.84	37.98	40.00	southwest
1998	113.28	27.98	72.30	southwest	115.15	36.30	50.49	northeast	117.10	38.10	31.74	northeast
1999	113.78	28.07	56.22	northeast	114.75	36.17	46.05	southwest	117.13	38.10	3.06	east
2000	112.57	27.99	134.92	southwest	114.75	35.96	24.05	south	116.93	38.03	24.02	southwest
2001	113.11	27.81	63.61	southeast	114.58	35.74	30.85	southwest	116.94	38.02	1.36	southeast
2002	113.00	27.77	12.71	southwest	114.32	35.76	30.07	northwest	116.88	38.11	11.76	northwest
2003	112.92	27.90	16.58	northwest	114.58	35.90	33.22	northeast	117.08	38.27	28.53	northeast
2004	112.98	27.85	8.35	southeast	114.55	35.89	3.61	southwest	117.30	38.34	25.13	northeast
2005	112.65	27.77	37.55	southwest	114.23	35.86	35.03	southwest	117.18	38.33	12.95	southwest
2006	112.94	27.69	33.79	southeast	114.12	35.76	16.53	southwest	117.23	38.37	7.48	northeast
2007	112.94	27.76	7.88	north	114.04	35.78	9.31	northwest	117.30	38.36	8.05	southeast
2008	112.98	27.84	9.44	northeast	113.83	36.03	36.47	northwest	117.15	38.35	17.05	southwest
2009	112.78	27.80	22.69	southwest	113.33	35.92	56.99	southwest	117.11	38.42	8.41	northwest
2010	112.57	27.76	24.19	southwest	113.13	35.98	23.08	northwest	117.05	38.45	7.76	northwest
2011	112.57	27.78	2.31	north	114.06	35.76	105.72	southeast	116.95	38.47	10.62	northwest
2012	112.47	27.80	10.84	northwest	113.46	35.86	67.80	northwest	116.82	38.49	14.86	northwest
2013	112.37	27.80	11.78	west	113.46	35.79	7.17	south	116.53	38.46	32.53	southwest
2014	112.32	27.79	4.83	southwest	113.22	35.72	28.04	southwest	116.37	38.41	18.85	southwest
2015	112.46	27.80	14.89	northeast	113.05	35.66	19.62	southwest	116.29	38.41	8.84	west
2016	111.99	27.66	54.01	southwest	113.00	35.64	6.37	southwest	116.11	38.48	21.40	northwest
2017	111.86	27.49	23.58	southwest	112.82	35.65	19.05	northwest	116.05	38.54	8.85	northwest
2018	111.71	27.32	25.68	southwest	112.91	35.53	16.84	southeast	115.91	38.52	15.03	southwest
2019	111.44	27.05	41.82	southwest	112.94	35.60	9.00	northeast	115.86	38.56	7.51	northwest
2020	111.26	26.89	27.60	southwest	112.39	35.62	61.71	northwest	115.78	38.55	9.30	southwest
2021	111.03	26.74	29.96	southwest	111.96	35.64	47.10	northwest	115.53	38.54	27.27	southwest

and

Table 6. Coordinates and shift distances of grapes’, bananas’, and fruits’ overall production center of gravities in China, 1978–2021.

Year	Grape Industry				Banana Industry				Fruit Industry
	Longitude	Latitude	Distance	Orientation	Longitude	Latitude	Distance	Orientation	Longitude	Latitude	Distance	Orientation
1978	103.98	40.15			112.47	23.37			115.65	35.96
1979	104.03	40.10	7.24	southeast	111.99	23.45	53.87	northwest	115.76	35.74	27.36	southeast
1980	103.06	40.21	107.90	northwest	111.71	23.60	35.85	northwest	114.98	35.27	100.98	southwest
1981	102.76	40.24	34.05	northwest	112.92	23.51	135.90	southeast	115.02	35.17	11.41	southeast
1982	103.46	39.99	82.88	southeast	112.81	23.51	12.45	west	114.46	34.69	81.63	southwest
1983	104.57	39.79	125.74	southeast	112.32	23.46	55.09	southwest	114.49	34.95	28.82	northeast
1984	105.35	39.81	86.84	northeast	112.75	23.43	47.52	southeast	113.99	34.45	78.77	southwest
1985	104.90	39.85	50.88	northwest	112.70	23.41	5.51	southwest	113.66	34.09	54.31	southwest
1986	105.95	39.50	123.23	southeast	112.44	23.32	31.37	southwest	113.42	32.92	132.97	southwest
1987	106.78	39.31	94.11	southeast	112.52	23.30	9.35	southeast	113.53	32.65	32.60	southeast
1988	108.13	39.06	153.20	southeast	112.80	23.37	32.02	northeast	113.59	32.90	29.55	northeast
1989	107.82	39.25	40.79	northwest	113.09	23.46	33.84	northeast	113.55	32.62	32.22	southwest
1990	106.12	39.73	195.75	northwest	112.96	23.44	14.52	southwest	113.54	32.40	23.67	southwest
1991	106.12	39.79	6.42	north	112.77	23.42	21.65	southwest	113.64	31.95	51.92	southeast
1992	106.59	39.44	64.61	southeast	112.35	23.41	46.41	southwest	113.84	32.56	72.20	northeast
1993	107.50	39.18	104.67	southeast	112.41	23.45	8.22	northeast	114.21	32.96	60.94	northeast
1994	108.37	38.81	105.13	southeast	112.54	23.53	17.11	northeast	114.42	33.19	33.53	northeast
1995	108.68	38.62	40.49	southeast	112.53	23.55	2.20	northwest	114.63	33.32	27.83	northeast
1996	109.03	38.49	42.03	southeast	113.05	23.77	62.32	northeast	115.17	33.93	91.02	northeast
1997	109.51	38.12	67.52	southeast	112.90	23.73	16.87	southwest	115.12	33.86	9.82	southwest
1998	109.82	38.20	35.70	northeast	112.84	23.69	7.92	southwest	115.16	34.08	24.36	northeast
1999	109.99	38.04	25.28	southeast	112.83	23.62	8.51	southwest	114.95	33.81	38.39	southwest
2000	110.62	38.05	70.15	northeast	112.81	23.57	6.03	southwest	114.57	33.32	68.50	southwest
2001	111.19	37.62	78.55	southeast	112.85	23.57	4.55	east	115.03	33.80	73.95	northeast
2002	110.55	37.90	77.59	northwest	112.88	23.54	5.31	southeast	114.79	33.93	30.43	northwest
2003	110.50	37.92	6.20	northwest	112.97	23.56	10.28	northeast	114.88	33.97	11.49	northeast
2004	110.16	37.95	37.08	northwest	112.85	23.56	13.69	west	114.79	33.91	12.02	southwest
2005	109.87	37.85	34.04	southwest	112.74	23.56	11.42	west	114.62	33.95	19.48	northwest
2006	109.35	37.82	58.27	southwest	112.58	23.56	18.64	west	114.62	34.02	7.07	north
2007	109.18	37.82	19.21	west	112.21	23.56	40.46	west	114.26	33.84	44.72	southwest
2008	109.43	37.57	39.34	southeast	111.89	23.65	37.40	northwest	114.33	33.85	7.44	northeast
2009	108.90	37.56	58.97	southwest	111.41	23.60	53.73	southwest	114.03	33.77	34.20	southwest
2010	109.13	37.29	39.46	southeast	111.12	23.57	31.78	southwest	113.81	33.86	26.16	northwest
2011	109.92	36.61	115.86	southeast	110.80	23.57	35.67	west	113.78	33.75	12.69	southwest
2012	109.58	36.41	44.33	southwest	110.48	23.59	35.89	northwest	113.47	33.68	34.43	southwest
2013	109.51	36.22	21.69	southwest	110.03	23.63	50.15	northwest	113.36	33.62	13.59	southwest
2014	109.45	35.94	31.80	southwest	109.85	23.63	20.14	west	113.18	33.60	20.34	southwest
2015	109.18	35.82	33.79	southwest	109.95	23.56	13.99	southeast	113.07	33.47	19.26	southwest
2016	109.11	35.47	39.39	southwest	109.86	23.55	10.50	southwest	112.85	33.50	25.13	northwest
2017	109.30	35.54	22.70	northeast	110.11	23.47	28.68	southeast	112.90	33.38	14.42	southeast
2018	108.80	35.92	69.05	northwest	110.10	23.54	7.62	northwest	112.70	33.10	38.06	southwest
2019	108.61	35.99	22.88	northwest	110.19	23.56	10.88	northeast	112.57	33.04	15.96	southwest
2020	108.83	35.86	28.85	southeast	110.31	23.55	12.60	southeast	112.45	32.83	26.09	southwest
2021	108.59	35.84	26.90	southwest	110.29	23.56	1.80	northwest	112.33	32.65	24.65	southwest

), it is apparent that the industry’s center of gravity has undergone significant shifts since the beginning of the reform and opening-up policy. Overall, the center of gravity of China’s fruit industry has moved from 115.65° E, 35.96° N (Liaocheng City, Shandong Province) in 1978 to 112.33° E, 32.65° N (Nanyang City, Henan Province), translating to a southwestward shift of 521.30 km. This aligns with findings in the literature [41,42], which indicate that the center of gravity for production has been shifting from east to west. The centers of gravity for the other four major fruits (oranges, pears, apples, and bananas) and the overall fruit industry have shifted westward. Notably, the center of gravity for apple production has moved northwestward by 216.27 km, illustrating a “westward and northward expansion” trend for apple production.

Building on this, the present study utilized Python to plot the evolution of the center of gravity’s movement trajectories over various years. This analysis delves into the movement distance and direction of the centers of gravity for China’s five major fruits and the overall fruit industry, with detailed illustrations provided in Figure 4. First, the center of gravity for orange production has the smallest overall movement distance but exhibits considerable spatial fluctuations. Since 2016, the center of gravity of orange production has rapidly shifted toward the southwest in a linear trajectory. Second, the center of gravity for pear production has experienced significant changes in longitude, shifting from 115.89° in 1978 to 111.96° in 2021. The largest annual movement occurred between 1979 and 1980, covering a distance of 148.06 km. From 2012 to 2018, the geographical distribution of pear production shifted notably toward the southwest. Since 2019, this center has generally exhibited a horizontal westward trend.

Third, in the investigation of fresh agricultural products suitable for consumers of all ages, apples emerged as the foremost choice, indicating their suitability for people across all age groups [43]. As a pivotal component of China’s fruit industry, the spatial distribution changes in apple production can be broadly categorized into the following three phases: (1) From 1978 to 1987, the center of gravity of apple production moved southwestward, covering a distance of 182.23 km. This period witnessed the largest annual movement, with the center shifting southwestward by 99.46 km between 1984 and 1985. (2) In the subsequent two decades (1988–2007), the production center gradually moved northeastward, with the coordinates in 2007 almost returning to those of 1978. (3) From 2008 to 2021, there was a notable “westward shift” in the apple production center of gravity, moving northwestward by 180.68 km, with an average annual movement speed of 13.90 km. Fourth, the center of gravity for grape production exhibited the most substantial change in spatial distribution, shifting southeastward by 701.09 km. This movement presents a slanted “J”-shaped variation. Fifth, the shift in the center of gravity for banana production was the smallest among the five major fruits. Before 2000, changes in the center were primarily observable in latitude fluctuations (i.e., in the north–south direction). After 2000, the center began to shift westward, with the movement between 2003 and 2007 being almost due west. Finally, the overall trajectory of the production center for the fruit industry outlines a “6” shape. From 1978 to 1991, the industry’s center of gravity swiftly moved southwestward, with a significant change in latitude from 35.96° in 1978 to 31.95° in 1991. However, from 1992 to the end of the 20th century, the center of gravity gradually shifted northeastward. Since the beginning of the 21st century, the spatial center of gravity of China’s fruit industry has exhibited an almost annual westward shift. The directional evolution of the fruit industry’s production center of gravity is largely consistent with changes in the apple production centroid. As a whole, the center of gravity of China’s fruit industry demonstrates a persistent westward shift; however, the migration trajectories of different fruit categories exhibit phase-dependent fluctuations due to natural resource endowments, government policies, and additional influencing factors.

4. Discussion and Conclusions

The layout of China’s agricultural industry is characterized by regionalization, continuity, and agglomeration development, culminating in the formation of industrial belts and block-style production. This article meticulously investigates the production agglomeration characteristics and trends within China’s fruit industry from three perspectives: the degree of production agglomeration, industry spatial heterogeneity and dependency, and the spatial evolution trajectory of the industry’s center of gravity. The analysis revealed a significant agglomeration phenomenon across the overall fruit industry in China, characterized by a high concentration. There are marked differences in the degree of agglomeration across China’s four major regions, which display distinct regional gradient features. Additionally, significant heterogeneity exists among the different types of fruits. Among these fruits, the concentration of apple production remained notably prominent; however, the agglomeration trends for orange, pear, and grape production gradually declined. Banana production was consistently concentrated in five provinces located in subtropical and tropical regions: Guangdong, Guangxi, Yunnan, Fujian, and Guizhou. The aforementioned phenomena indicate, to a certain extent, the significant and strong dependency of the fruit industry’s development on environmental and climatic conditions.

Sustainability challenges, including climate adaptation, greenhouse gas emissions, and food security have raised concerns [44]. Influenced by a variety of factors, including climate change, population growth, technological advancements, and other factors [40,45], the agglomeration of fruit production has undergone subtle shifts. The intensifying situation of global warming—along with a surge in natural disasters and pestilences stemming from extreme weather conditions—poses a severe threat to agriculture, particularly affecting the production and market stability of the cultivation sector, thereby significantly affecting the agglomeration of China’s fruit industry. While natural geographical conditions primarily influence production agglomeration, the number and quality of professionals in the industry, the technological standards across various segments of the industrial chain, the support provided by national and government entities, and the trends in both domestic and international markets also exert considerable influences that cannot be overlooked [46]. Furthermore, most fruit commodities are characterized by their perennial crop habits, fresh and perishable storage characteristics, labor-intensive production processes, among other distinctive features. Therefore, it can be posited that a range of factors, including natural endowments, labor and transportation costs, among others, significantly influence the concentration of production within the fruit industry.

Our findings also indicate that spatial autocorrelation within China’s fruit industry is progressively strengthening, signifying that production levels are spatially correlated. Overall, the specialization index has declined yet localized relative agglomeration phenomena still exist. Finally, calculations of the industrial production center reveal a gradual shift from east to west overall. Among these fruits, the center of gravity for orange and pear production is shifting southwestward, whereas apple and banana production demonstrates westward movement with northward expansion. Additionally, the production center for grapes trends southeastward. As evidenced by recent studies [47,48,49], China has experienced a steady increase in forest area. This expansion has facilitated the optimization of land resources in the western region and has played a pivotal role in mitigating drought conditions [50]. The long-term “westward shift” in China’s fruit industry production is influenced by a confluence of factors such as climate change, enhancements in infrastructure, and socioeconomic advancements in the western region.

The factors driving the shift in the center of gravity and transformation in the agglomeration pattern in the fruit industry can be examined through natural conditions, industrial policies, scientific and technological innovation, and market demands.

First, natural endowments are among the most critical factors influencing the development of the fruit industry. Compared to other agricultural products, fruits possess distinctive biological traits, and each type has a unique optimal cultivation area. The unique topographical conditions of the western region significantly hinder local socioeconomic development [42]. A field study conducted by our research group indicates that certain regions of Gansu exhibit a favorable climatic environment for fruit cultivation. However, geographical constraints limit the scalability of mechanized production in the local fruit industry.

Second, governments at all levels in China have consistently emphasized the high-quality development of the fruit industry. Major national ecological restoration projects have been vigorously implemented [21], with governments at various levels encouraging the cultivation of fruit and other cash crops to foster high-quality agricultural development. A range of initiatives serve as ongoing guarantees for the development of the fruit industry, including the provision of special financial support, research and development for new technologies, the construction of high-quality trading markets, and the cultivation and building of special brands.

Third, the input–output ratio of fruit production has increased significantly due to advancements in fruit varieties [41], agricultural technological innovations, and spillover effects [51,52]. Covering the entire process from research and development in breeding to production, planting, harvesting, and storage, a diverse range of technologies and models aimed at enhancing productivity and efficiency are gradually overcoming the constraints imposed by natural conditions. These advancements provide technological momentum for the rapid development of the fruit industry.

Fourth, social progress and economic development have driven increasing consumer demand for fruit products, shifting the focus from quantity to quality and thereby enhancing effective supply. However, significant disparities in natural, economic, and social conditions exist across regions and continue to evolve. These factors drive the shift in the fruit industry’s production center of gravity and the transformation of its production agglomeration patterns.

Changes in fruit industry agglomeration influence both smallholders and regional inequality in a dual manner. Specifically, a weakening of fruit industry agglomeration may result in higher costs, reduced market competitiveness, diminished bargaining power, technological stagnation, and constrained innovation capacity for smallholders. Conversely, the rise in e-commerce has facilitated a shift toward more decentralized market opportunities, allowing smallholders to transition to the production of high value products (e.g., organic fruits and GI agricultural products) and adopt more specialized and differentiated market strategies, thereby reducing their reliance on economies of scale. Regarding regional inequality, the fruit industry tends to concentrate in densely populated areas of developed regions, potentially driving employment growth and infrastructure expansion, thereby contributing to regional economic balance and resource reallocation. However, this process may also lead to the decline of traditionally advantageous regions. Amid the dynamic evolution of fruit industry agglomeration, policy intervention, regulatory frameworks, and agricultural technological support play a crucial role. This raises the question of how to formulate differentiated industrial policies and promote regional synergy in development.

To promote high-quality development in the fruit industry and ensure global food security, this study, based on its research findings, proposes the following strategic recommendations.

First, identifying evolutionary trends, establishing a dynamic monitoring and early warning system, and implementing precise policy regulations are essential. The spatiotemporal characteristics and changing trends of production agglomeration not only provide crucial guidance for the future development of the fruit industry but also profoundly influence the innovation of China’s agricultural policy system. From the perspective of policy regulation, the spatiotemporal evolution of production agglomeration offers new insights into the structural reform of agricultural supply. On the one hand, policy formulation should shift from extensive management to precise regulation, integrating data sources through big data platforms to establish a comprehensive policy response mechanism encompassing meteorological early warning, soil moisture monitoring, and market supply–demand forecasting. This approach ensures the accurate alignment of policy instruments, such as fiscal subsidies and technology promotion, with regional industrial characteristics. On the other hand, the restructuring of the agricultural spatial governance system should be promoted by dynamically monitoring production agglomeration intensity. This would guide the allocation of policy resources toward advantageous production regions, ensuring a more targeted and effective policy framework. For example, targeted support for cold chain logistics was implemented in the Shaanxi apple production belt, while legislative measures for brand protection were strengthened in the primary production areas of Gannan navel oranges. These initiatives contribute to the formation of a positive feedback mechanism between agglomeration and policy. At a deeper level, this transformation is reshaping the governance paradigm of China’s agricultural sector. The integration of a dynamic monitoring system with a differentiated policy toolkit marks a shift in agricultural policy from reactive responses to proactive governance.

Second, efforts should be made to deepen the development of regional characteristics and strengthen the location advantage of fruit products. The formulation of criteria for identifying areas of superiority for specialty fruit products is a potential avenue for exploration, as is the establishment of a fund for cultivating specialty fruit industry clusters. The deepening of the cluster strategy for high agglomeration category (apple) fruit should be implemented through the following measures: the establishment of cross-provincial and regional technical cooperation platforms; the targeted promotion of agricultural technology; and the building of a vertical division of labor network, a “planting–processing–e-commerce” integration mode. The main producing areas should be guided toward the deep processing of fruits and the extension of cold chain logistics. Layout strategies for low agglomeration categories (grapes) of fruits should be optimized by delineating the functional zoning of varieties and establishing a compensation mechanism for regional collaboration.

Third, the development model for the westward movement and upgrading of the fruit industry should be innovated. Implementing the “Western Fruit to the East” quality improvement plan in the primary fruit-producing regions of the west will involve establishing digital sorting centers and creating a “Western High-Quality Fruit Special Zone” in the eastern consumer market to streamline distribution channels. Furthermore, enterprises in the eastern regions are encouraged to invest in the development of standardized orchards in the western region, supporting the “technology share + guaranteed purchase” cooperation model and implementing advanced management standards from the east (such as the Shandong apple management system) to expand and extend the fruit industry value chain.

While this study makes a marginal contribution to the future development of the fruit industry, it has certain limitations. In terms of methodological innovation, the current approaches to measuring agricultural industry agglomeration are somewhat traditional and may not fully apply to the agglomeration measurement of different types of crops, necessitating imminent innovation at the methodological level. Moreover, further exploration of the factors influencing fruit industry agglomeration is essential, with a pressing need to unravel the mechanisms among internal and external factors and industry agglomeration. These areas will guide future research studying agglomeration in the fruit industry.