The security and stability of grain production form the foundation for the high-quality development of a nation’s economy and society, holding strategic significance for social stability and sustainable economic growth. This study employs the core variable method to measure and analyze the spatiotemporal pattern and evolutionary characteristics of China’s county-level grain production resilience from 2001 to 2023. A geographically weighted regression (GWR) model is further constructed to explore the influencing factors of grain production resilience and their spatiotemporal heterogeneity. The results indicate that: (1) China’s grain production resilience shows a fluctuating upward trend, with significant regional disparities—exhibiting a spatial gradient characterized by a high-value core in the northeast, followed by the eastern and central regions, and lagging performance in the west. (2) The overall disparity in grain production resilience is substantial, with interregional differences serving as the primary source of variation. Internal disparities are largest in the eastern region, followed by the northeast and west, while the central region displays relatively minor internal differences. (3) Grain production resilience exhibits significant spatiotemporal heterogeneity. The positive effects of grain yield per unit area, multiple cropping index, agricultural mechanization level, agricultural informatization, and agricultural investment have continued to strengthen, whereas the negative impact of the proportion of employment in the primary industry has become increasingly evident, suggesting that traditional labor-intensive agricultural models increasingly constrain grain production resilience.
geographically weighted regressiongrain production resilienceinfluencing factorsregional disparityspatiotemporal evolutionThe author(s) declared that financial support was received for this work and/or its publication. This research was supported by the National Social Science Fund of China (General Project No. 24BGL174), titled “Research on the Risk Prevention of China’s Global Food Supply Chain under Major Emergencies.”section-at-acceptanceLand, Livelihoods and Food SecurityIntroduction
The security and stability of grain production constitute the foundation for the high-quality development of a nation’s economy and society, and are of strategic significance for social stability and sustainable economic growth. Since the 20th National Congress of the Communist Party of China, ensuring national food security and strengthening the foundation of grain security have been identified as major national tasks. China’s total grain output has remained above 1.3 trillion jin for nine consecutive years. However, at this new stage of development, China’s grain production still faces multiple challenges. Global geopolitical conflicts, slowing economic growth, and the increasing frequency of extreme weather events have placed considerable pressure on the stability of the global food system (UNICEF, 2021; Hertel, 2022). Meanwhile, under the influence of the urban–rural dual structure, domestic grain production continues to be dominated by smallholder farming, characterized by a weak foundation and limited capacity to respond to natural disasters and market fluctuations (Fan and Zhang, 2020). Enhancing grain production resilience has become essential for maintaining production stability and safeguarding national food security. This has significant theoretical and practical implications for future policy formulation and capacity building in food security.
The concept of “resilience” was first introduced by Holling (1973) to describe the capacity of ecosystems to recover from and adapt to external disturbances. Tendall et al. (2015) later extended this concept to food system research by proposing a food system resilience framework, emphasizing the system’s ability to withstand and adapt to natural, political, or economic shocks. Methodologically, two main approaches are commonly used to measure grain production resilience: the indicator system method and the core variable method. The indicator system method is the most widely applied approach in existing studies. It evaluates system resilience comprehensively by constructing a multidimensional index system encompassing resistance, recovery, and adaptability. However, this method involves a degree of subjectivity and limitations in indicator selection and weight determination (Zhang et al., 2022). The core variable method, initially applied in regional economic resilience research (Martin and Sunley, 2015), captures a system’s sensitivity to external shocks through the fluctuations of key variables such as output and employment. Although its direct application in grain production resilience studies remains limited, the method’s simplicity and intuitive nature provide a feasible analytical framework for county-level empirical studies based on large-sample, long-term panel data. At the case study level, some scholars have focused on specific regions, such as major grain-producing areas (Liu et al., 2021a,b) and the Yellow River Basin (Wang Q. et al., 2023; Wang X. et al., 2023). Most existing studies use provinces or single regions as the unit of analysis. While these approaches capture macro-level patterns, they fail to reveal the spatial heterogeneity and micro-mechanisms at the county level (Li and Li, 2023; Yu et al., 2022). This paper uses the core variable method using county-level data. It offers the advantage of requiring only a single key indicator readily available across counties and years, which reduces data demands and subjective weighting in a long-term panel study. Focusing on one variable may overlook other resilience dimensions; we mitigate this by complementing the core-output measure with a multifactor analysis to ensure a more comprehensive assessment.
Regarding the influencing factors of grain production resilience, research scope has gradually expanded in response to rising uncertainty and risk. Within a resilience framework, existing determinants can be systematically organised into four interrelated dimensions: (i) exposure and baseline constraints from the natural environment, (ii) production capacity and input endowments, (iii) socio-economic structure and factor mobility, and (iv) institutional and service-based risk management. Specifically, natural-environmental factors shape the intensity and frequency of shocks (Lesk et al., 2021); capacity and input factors such as high-standard farmland construction and technological innovation enhance resistance and recovery by improving productivity and buffering capabilities (Zhang et al., 2021; Fang et al., 2021); socio-economic and labour factors, including labour transfer, influence resilience through changes in labour availability, human capital, and farm management practices (Chen et al., 2020); and institutional and service factors, such as technological training, agricultural production services, and agricultural insurance, strengthen coping capacity by facilitating technology adoption, stabilising expectations, and sharing risks (Wang and Zhao, 2020; Liu et al., 2022; Zhou et al., 2022). This classification clarifies the theoretical logic for variable selection in the subsequent empirical models by linking each factor group to specific resilience functions (resistance, recovery, and reconfiguration). Most existing research has focused on the independent effects of specific factors, with limited attention to how these dimensions interact (e.g., services and training conditioning the effectiveness of infrastructure and innovation, or labour mobility reshaping technology adoption and risk exposure). Moreover, substantial variations in research scope, methodology, and data sources have resulted in significant heterogeneity in conclusions, while spatial disparities, scale differences, and spatiotemporal dynamics have yet to be fully integrated.
Although existing research has made substantial progress in measuring grain production resilience and identifying its influencing factors, there remains considerable room for improvement. On the one hand, most studies have overlooked the interactive effects among multiple factors and the regional heterogeneity of resilience, making it difficult to comprehensively reveal the underlying formation mechanisms. On the other hand, the research scale has largely remained at the provincial or specific regional level (Wang Q. et al., 2023; Wang X. et al., 2023), limiting the ability to capture spatial disparities and dynamic evolution at the county level and thereby weakening the support for regionally targeted policy formulation. In this context, this study uses counties as the basic unit of analysis and, based on long-term panel data from 2001 to 2023, systematically examines the spatiotemporal evolution and regional disparities of China’s grain production resilience. Compared with existing studies that are predominantly conducted at the provincial or basin scale, this research provides a fine-grained county-level perspective and thus uncovers substantial intra-regional heterogeneity that is obscured at higher spatial scales. Furthermore, by integrating the core-variable approach with a geographically weighted regression framework, this study captures the spatial heterogeneity of resilience drivers, offering a more dynamic and mechanism-oriented understanding of grain production resilience formation. On this basis, it explores the driving factors and proposes differentiated policy recommendations, thereby contributing both methodological advancement and policy-relevant insights to the literature on agricultural resilience and national food security.
Research designData source
The data used in this study are primarily derived from the China City Statistical Yearbook (2001–2023), the China County Statistical Yearbook (2001–2023), and other statistical materials published by the National Bureau of Statistics. Additional information was obtained from the statistical yearbooks of each province (autonomous region and municipality) and the Statistical Bulletins of National Economic and Social Development released by local governments. Supplementary Data were collected from the EPS Statistical Database and official public datasets of relevant regions. Considering variations in data availability and administrative boundary adjustments during the study period, the original data were systematically integrated and cleaned to ensure continuity and comparability. Considering variations in data availability and frequent administrative boundary adjustments during the study period, a constant-boundary principle was adopted to ensure the temporal consistency of county-level units. Specifically, counties were harmonized to a unified spatial reference by tracing administrative changes over time. For counties that were merged, historical data were aggregated to the post-merger administrative unit; for counties that were split, historical data were reassigned to the original parent county to maintain continuity; and for newly established counties, observations were included only after their official establishment, with earlier periods treated as missing rather than interpolated. This rule-based harmonization avoids artificial structural breaks caused by administrative changes and preserves the comparability of long-term county-level time series. After processing, a total of 2,646 county-level administrative units (excluding Hong Kong, Macao, and Taiwan) were included as research samples, comprising 59,859 county–year observations from 2001 to 2023. These data provide a robust empirical foundation for the subsequent measurement and spatiotemporal analysis of grain production resilience.
MethodsMethods for measuring food production resilience
The scientific and accurate measurement of grain production resilience remains a key frontier issue in academic research. Martin classified economic resilience into three dimensions—resistance, recovery, and adaptability—and proposed the use of the core variable method to quantitatively analyze resistance and recovery, thereby effectively characterizing regional economic resilience (Martin and Sunley, 2015). Accordingly, this study adopts the core variable method and selects grain output as the core indicator. This choice is supported by both solid theoretical reasoning and practical considerations:
First, grain output represents the final outcome of the grain production system and comprehensively reflects the combined effects of multiple factors, including natural conditions, input elements, and institutional environments. When the system experiences shocks such as extreme weather events, market fluctuations, or policy adjustments, changes in output serve as the most direct and sensitive indicator of system response. Thus, fluctuations in grain output objectively capture the system’s capacity for resistance and recovery. Second, from a theoretical perspective, Adger et al. (2011) proposed in studies of economic resilience that a system’s performance under shocks should be reflected through its key output variables. The core variable method is built upon this theoretical logic. Applying this method to grain production resilience not only aligns with the theoretical connotation of resilience measurement but also corresponds to the practical characteristics of agricultural systems. Third, from an academic perspective, existing research on agricultural and food system resilience commonly employs output or income as core variables (Meuwissen et al., 2019), indicating that this method possesses strong operability and broad academic recognition.
The calculation formula is as follows (Equation 1):
(ΔRit+k)=Σ(Rit·Gt+k)
where: (ΔRit+k) denotes the change in grain output of region i during the expected period (t + k); Rit represents the grain output of region i in year t; Gt+k indicates the change in national grain output during the same period (t + k).
The formula for calculating resilience is therefore given by (Equation 2):
Resilience=ΔRi−(ΔRi)α∣(ΔRi)α∣
where: Resilience denotes the grain production resilience; ΔRi represents the actual change in grain output of region i; (ΔRi)α indicates the expected change in grain output of region i, calculated based on the national average level in the same year. If the grain production resilience of a region exceeds the national expected average, the resilience value will be greater than zero, indicating a higher level of resilience. Conversely, if the resilience level of a region falls below the national expectation, the resilience value will be less than zero, suggesting a lower level of resilience.
It is important to note that the resilience value represents a quantitative measure, while the resilience level refers to a qualitative classification based on the interval of resilience values. These two concepts are not independent but are logically connected, reflecting the transformation from quantitative assessment to qualitative interpretation. This method not only ensures the objectivity and comparability of the measurement process but also facilitates subsequent analyses of spatial distribution patterns and regression testing of influencing factors.
Kernel density estimation
This study employs kernel density estimation to examine the dynamic evolution trend of grain production resilience, and the functional form is expressed as follows:
f(x)=1nh∑i=1nk(xi−x¯h)
In Equation 3, n denotes the number of observations (i.e., the total number of counties), and i represents each individual county, indicating independent and identically distributed observations. x refers to the mean value, and f(x) denotes the density function. k(xi−x¯h) is the kernel function, while h represents the bandwidth, whose magnitude determines the accuracy of the kernel density estimation and the smoothness of the density curve. Based on the selection of the optimal bandwidth, this study applies the Gaussian kernel density function to analyze the dynamic evolution characteristics of grain production resilience at the county level in China. The specific functional expression is as follows (Equation 4):
xik(x)=12πe−12t2Theil index
This study constructs a Theil index–based measurement framework to systematically analyze the spatial heterogeneity and driving mechanisms of grain production resilience across China’s counties. Based on the Theil index, a decomposition model is further developed to quantitatively identify the contributions of inter-regional and intra-regional differences. The Theil index is utilized instead of the Gini coefficient because Theil can be decomposed into inter-regional and intra-regional components, enabling a clearer identification of disparity sources across scales. For regional division, the country is classified into four major regions—Northeast, East, Central, and West—to explore the sources of regional disparities in county-level grain production resilience in China.
Geographically weighted regression model
Traditional linear regression models perform global estimations on the sample and parameters without accounting for spatial factors. In contrast, the geographically weighted regression (GWR) model enables localized estimation of influencing factors across different regions, thereby revealing spatial variations in their effects (Fotheringham et al., 2002; Wu, 2020). To examine the spatial heterogeneity of factors influencing county-level grain production resilience in China, this study employs the GWR model to conduct regression analysis for each variable.
The model is specified as follows (Equation 5):
yi=β0(ui,vi)+∑kβk(uj,vi)xik+εi
where: yi denotes the dependent variable; β0(ui,vi) represents the intercept at location i, indicating the baseline influence of explanatory variables on the dependent variable; (uj,vi) are the geographic coordinates of sample i; βk(uj,vi) is a continuous function representing the estimated local regression coefficient of the kth explanatory variable at location i; xik denotes the value of the kth independent variable at location i; and εi represents the random error term.
Variable selection
The formation of grain production resilience is influenced by a complex interplay of diverse factors. Existing studies have explored its key driving elements from multiple perspectives; differences in regional scope, methodological approaches, and data selection have led to inconsistencies in empirical findings and divergent interpretations of the underlying mechanisms. Based on the conceptual framework and theoretical connotation of grain production resilience, this study conducts a systematic analysis of its influencing factors and mechanisms of action. Cheng et al. (2024) argued that grain production resilience is a continuously evolving process encompassing three stages: resistance, recovery, and reconfiguration. This process is shaped by both intrinsic and extrinsic factors—the former including climate change, soil quality, and ecosystem conditions, and the latter encompassing technological innovation, farmland infrastructure, and agricultural insurance (Huang et al., 2022; Zhang et al., 2023).
Resistance reflects the capacity of the agricultural system to maintain a stable level of production under external shocks. Existing studies have shown that grain yield per unit area is a key indicator for measuring agricultural productivity. Higher yields typically indicate advantages in technological progress, efficiency of factor allocation, and crop variety improvement, enabling the system to sustain high output even under risks such as abnormal climate conditions and pest outbreaks (Fischer et al., 2021; Gollin et al., 2021; Han et al., 2024). Meanwhile, the level of agricultural mechanization reflects the degree of dependence on machinery and equipment. A higher mechanization level not only significantly improves labor productivity but also ensures the stable operation of key production processes—such as sowing, fertilization, and harvesting—under extreme weather or labor shortages, thereby enhancing the system’s capacity to withstand shocks (Zhao et al., 2023). Based on this, the present study selects grain yield per unit area and the level of agricultural mechanization as indicators of resistance and hypothesizes that both are positively correlated with grain production resilience (Li et al., 2023).
Recovery reflects the agricultural system’s ability to restore its functions after being disrupted through internal adjustment or external intervention. The multiple cropping index captures the flexibility of farmland utilization and crop rotation; a higher index indicates greater potential for the agricultural system to achieve rapid recovery following disasters (Li et al., 2022). However, an excessively high proportion of employment in the primary industry often suggests redundant agricultural labor and insufficient investment in capital and technology, which can constrain the system’s recovery capacity (Dorward, 2020; Wang Q. et al., 2023; Wang X. et al., 2023). This study selects the multiple cropping index and the proportion of employment in the primary industry as indicators of recovery, hypothesizing that the multiple cropping index is positively correlated with grain production resilience, while the proportion of employment in the primary industry is negatively correlated with it.
Reconfiguration emphasizes the agricultural system’s ability to enhance its functions through optimization and restructuring after disturbances. The level of rural informatization reflects the degree of digital technology application. The development of informatization not only promotes the intelligence and precision of agricultural production but also strengthens farmers’ access to market and weather information, thereby improving their capacity for risk early warning and adaptive response (Klerkx and Rose, 2020; Zhang et al., 2024). Meanwhile, a higher level of investment contributes to system restructuring by improving infrastructure, fostering technological innovation, and advancing institutional reform, thus providing financial and material support for the system’s transformation toward higher quality and sustainability (Liu et al., 2021a,b; Cai et al., 2025). Accordingly, this study selects the level of rural informatization and the level of investment as indicators of reconfiguration, hypothesizing that both are positively correlated with grain production resilience (see Figure 1).
Trends of resilience in grain production in China and four major regions.
Line graph depicting the average resilience of grain production from 2001 to 2023 across different regions: Northeast, Eastern, Central, Western, and the National average. The Northeast shows significant fluctuations, peaking around 2004 and 2020. Other regions remain relatively stable with minimal changes.ResultsTemporal evolution of resilience in China’s grain production
From 2001 to 2023, China’s grain production resilience exhibited a fluctuating yet upward trend. In the early years, the resilience level fluctuated between 2001 and 2005 due to the combined influence of natural disasters and unstable market conditions. Subsequently, as agricultural mechanization continued to improve and high-standard farmland construction advanced, grain production resilience steadily strengthened, showing a relatively stable upward trajectory, particularly after 2015 (see Table 1).
The description of independent variables.
Variable
Label
Specific indicators
Direction
Yield per unit area
X1
Total grain production/total cropland area
+
Level of Agricultural Mechanization
X2
Total agricultural machinery power/total cropland area
+
Percentage of workforce employed in primary industries
X3
Number of persons engaged in agriculture, forestry, animal husbandry, and fisheries/number of rural residents
−
Crop rotation index
X4
Total cropland area/arable land area
+
Level of Rural Informatization
X5
Broadband subscriber count/total households at year-end
+
Investment Level
X6
Total fixed asset investment/rural population
+
It should be noted, that despite the improvement nationwide, significant disparities persist among counties. Some major grain-producing counties have maintained high resilience values owing to favorable land resources and strong policy support. In contrast, counties characterized by fragile ecological environments and weak infrastructure continue to record low resilience values. This intra-county differentiation suggests that while national resilience has improved, it is accompanied by pronounced internal disparities.
(1) 2001–2018: The Low-Level Fluctuation Stage. During this period, the mean level of grain production resilience remained in the low-to-medium range, indicating a limited capacity of the system to withstand external shocks. In 2004, resilience temporarily peaked as a result of combined effects from increased grain output, favorable policies, and suitable climatic conditions. Nevertheless, the improvement proved unsustainable due to weak agricultural infrastructure and insufficient levels of mechanization and informatization, leading to prolonged volatility. This stage demonstrates that under the dominance of smallholder farming and the impact of external disturbances, the grain production system exhibited pronounced vulnerability and lacked a stable foundation for resilience-driven growth.
(2) 2019–2023: The Significant Improvement Stage. With the in-depth implementation of the “Storing Grain in Land and Storing Grain in Technology” strategy, the construction of high-standard farmland accelerated, and both mechanization and informatization levels improved markedly. These advances promoted a stepwise leap in the grain production system, keeping resilience at a relatively high level. Meanwhile, policy investment and technological progress enhanced the system’s adaptive capacity; extreme climate events and market fluctuations continued to introduce uncertainty, causing moderate volatility in certain years. This indicates that modernization and policy guidance have become the core driving forces behind the current improvement in grain production resilience, yet their stability remains constrained by external environmental factors.
At the regional level, the Northeast region exhibits the highest grain production resilience, which rose rapidly in the early stage and has since remained high, reflecting its advantage as China’s primary grain-producing area. Nonetheless, its resilience has shown intermittent instability due to climatic fluctuations and a relatively homogeneous industrial structure. The Central region generally aligns with the national average, benefiting from favorable arable-land and infrastructure conditions and achieving steady improvement under policy support, displaying a pattern of continuous convergence. The Eastern and Western regions maintain relatively low resilience levels—the former constrained by limited arable land and accelerated urbanization, and the latter restricted by harsh natural conditions and water scarcity, with grain production occupying a smaller share of the regional economy. China’s grain production resilience exhibits a stable gradient pattern of “Northeast leading—Central catching up—East and West lagging,” underscoring the pronounced regional imbalance arising from differences in resource endowments, industrial structures, and policy environments.
Figure 2 presents the results of the kernel density estimation, showing that the distribution curves of county-level grain production resilience from 2001 to 2023 exhibit a distinct multi-peak pattern. This indicates that disparities among counties have not disappeared with the improvement in resilience levels. On the contrary, some counties have gradually clustered toward higher resilience levels, while others have remained persistently in low-resilience zones. This phenomenon of “divergence with coexistence” suggests that disparities in resource endowment, technological investment, and policy effectiveness across counties are becoming increasingly pronounced.
Kernel density estimation of grain production resilience in the four major regions of China.
Four 3D plots (a, b, c, d) showing the nuclear density distribution of grain production resilience levels from 2001 to 2023. Each graph varies in peak height and distribution. The x-axis represents resilience levels (negative to positive), the y-axis denotes years (2001 to 2023), and the z-axis shows nuclear density.
Further observation of changes over time reveals that in 2001 most counties were concentrated at low-to-medium resilience levels; by 2010, a portion of counties had clearly shifted toward higher levels; in 2020, the impacts of the COVID-19 pandemic and extreme weather led to a contraction in the number of high-resilience counties; and by 2023, resilience levels rebounded, forming several clusters of high-value regions. These dynamics indicate that county-level grain production resilience does not increase in a strictly linear manner over time but rather evolves through intertwined processes of shock and recovery.
The four regions display certain similarities in the evolution of their main peaks, distributional spread, and polarization trends, yet their trajectories differ markedly. From the perspective of peak evolution, all regions experienced an initial rise followed by a decline, with the width of the main peak first narrowing and then widening. By the end of the study period, both the eastern and western regions exhibited higher and more concentrated peaks, suggesting increasing aggregation of high-resilience areas. In contrast, the central and northeastern regions showed lower and more dispersed peaks, implying growing internal disparities. The eastern region achieved a high degree of concentration driven by technological and market advantages, whereas the western region improved notably in some areas through policy and technological support. The central region experienced widening disparities due to crop-structure shifts and labor outflow, while the northeast showed significant differentiation constrained by resource overuse and climate variability.
In terms of distributional extension, all four regions display bimodal tails, with the eastern and northeastern regions exhibiting longer right-hand tails, indicating a larger proportion of high-resilience counties. These areas, supported by agricultural modernization, structural adjustment, and climate adaptation, have generated a significant spillover effect. Regarding polarization, all four regions show a multi-polar pattern. The eastern and central regions remain relatively stable, though stratified differences persist; the western region shows pronounced high-low contrasts; and the northeast demonstrates the most distinct step-wise polarization. High-resilience counties there are concentrated in areas with well-developed infrastructure and advanced technology, while traditional agricultural counties remain severely constrained, forming a pattern of “dual polarization.” This spatial configuration aligns with the core–periphery theory and the concept of path dependence, reflecting the long-term combined effects of natural conditions, technological progress, policy support, and market dynamics.
Spatial characteristics of food production resilience
To further reveal the spatial distribution of grain production resilience at the county level in China, this study uses ArcGIS to visualize the resilience levels for each year. Considering the limitations of paper length, four representative years—2001, 2010, 2020, and 2023—were selected for analysis. The resilience levels of all counties were classified into five categories: high, sub-high, medium, sub-low, and low. Based on this classification, the study further examines the temporal evolution patterns and spatial distribution characteristics of counties across different resilience levels (see Figure 3).
Spatial distribution of resilience in grain production in Chinese counties (2001–2023). This map is based on the standard map GS (2019)1822 from the Ministry of Natural Resources’ Standard Map Service website, with no modifications to the base map boundaries.
Four maps of China from 2001, 2010, 2020, and 2023 show data values using color coding. The categories are no data, low, second-lowest, median, second-highest, and high values, represented by white, light green, green, yellow, orange, and red, respectively. Over time, there is an increase in higher value areas, particularly noticeable in the northern regions. A scale of zero to five hundred kilometers is included for reference.
Based on county-level analysis, China’s grain production resilience from 2001 to 2023 has shown a steady upward evolution from low to high levels, with a spatial pattern characterized by “highest in the Northeast, moderate in the East and Central regions, and relatively low in the West.” In 2001, low and sub-low resilience zones dominated, primarily concentrated in the arid western regions and the hilly areas of the south, displaying a clear spatial clustering pattern. Specifically, counties with low and sub-low resilience were mainly distributed across the western arid belt and southern hilly regions; medium-level counties were more widespread, concentrated in the Central Plains and the Huang–Huai–Hai area; while high-resilience counties were sparsely scattered across the Northeast Plain and portions of the eastern coastal grain-producing belt, forming an spatial pattern of “broad low values and sparse high values.” By 2010, medium- and high-resilience zones had increased significantly, with high-resilience areas expanding from the core grain-producing regions of Northeast China toward the Beijing–Tianjin–Hebei area. Meanwhile, the southeastern coastal region remained at a low level, constrained by frequent meteorological disasters and the dual pressures of urbanization-driven farmland loss and an export-oriented agricultural structure, which increased dependence on external supply and hindered improvements in resilience. In 2020, overlapping shocks from the COVID-19 pandemic and natural disasters—such as delayed farming seasons and agricultural losses—substantially weakened system stability, resulting in the contraction of high-resilience zones and a corresponding increase in the proportion of low and sub-low areas. By 2023, high-resilience zones further concentrated in the Northeast, North China, and the Huang–Huai–Hai Plain, benefiting from the implementation of the “Storing Grain in Land and Technology” strategy and the advancement of high-standard farmland, mechanization, and informatization. In contrast, the western and remote regions, constrained by limited resource endowments, lagging infrastructure, and restricted technological diffusion, continued to exhibit low-value clusters, reflecting a persistent trend of regional divergence.
From a regional perspective, the Northeast maintains the highest resilience, the Central region has experienced steady improvement, while the East has progressed more slowly due to constraints from limited arable land and the crowding-out effect of non-agricultural industries. The West remains relatively lagging as a result of unfavorable natural conditions and insufficient investment. Nevertheless, significant intra-regional disparities exist at the county level. For example, in the Northeast, core counties in the Songnen Plain exhibit substantially higher resilience than peripheral mountainous counties due to superior arable land resources and large-scale agricultural operations. In the eastern coastal region, some counties demonstrate strong adaptive capacity supported by informatization and facility agriculture, whereas others remain weaker due to land scarcity and fragmented farm structures. While macro-regional divisions help illustrate the spatial structure, a fine-grained county-level analysis is indispensable for a deeper understanding of spatial disparities in grain production resilience.
Regional variations in food production resilience
The preceding analysis examined the temporal and spatial evolution of county-level grain production resilience in China; it did not reveal the extent of regional disparities or the primary factors driving them. To address this gap, the present study employs the Theil index to measure the level of disparity in grain production resilience across counties. Furthermore, China is divided into four regions—Eastern, Central, Western, and Northeastern—to enable regional decomposition of the Theil index. This decomposition distinguishes between inter-regional and intra-regional disparities, thereby providing a basis for exploring the underlying mechanisms and key determinants of spatial differences in county-level grain production resilience.
Figure 4 illustrates that from 2001 to 2023, China’s disparity in grain production resilience exhibited a fluctuating upward trend. Although the level remained relatively low, disparities expanded significantly after 2018, reaching their peak in 2023. The sources of these disparities underwent a transformation—from being primarily driven by intra-regional differences to being dominated by inter-regional disparities. During the initial stage (2001–2005), the Theil index was low but expanded rapidly, mainly due to agricultural structural adjustment and the advancement of modernization. Between 2005 and 2015, the disparities entered a relatively stable phase, reflecting the equalizing effects of institutional reforms and technological diffusion. From 2018 to 2023, the disparities widened sharply and peaked in 2023, closely associated with extreme weather events, fluctuations in agricultural input prices, and the impacts of the COVID-19 pandemic.
Spatial differences and decomposition of grain production resilience in Chinese counties.
Line and bar graph depicting regional differences and contribution rates in China from 2001 to 2023. Bars represent different regions (Eastern, Western, Northeast, Central) with slight fluctuations. Lines show the decline of regional contribution rate from about 70% to 40%, and interregional contribution rate fluctuating around 1.2.
The decomposition results of the Theil index indicate that inter-regional disparities constitute the primary source of national differences in grain production resilience. At the same time, intra-regional disparities among counties have also continued to expand. In the eastern region, under the dual influences of urbanization and industrial transformation, some counties have maintained high resilience through agricultural technology advancement and facility-based development, while others remain at low levels due to severe farmland loss. Although the northeastern region maintains the highest resilience, internal disparities have gradually widened, manifesting as differentiation between core grain-producing areas and peripheral counties. The central and western regions show relatively smaller disparities, yet uneven development among counties has become increasingly evident. The spatiotemporal evolution of grain production resilience in China exhibits a “dual characteristic”: on the one hand, significant differences exist among macro-regions; on the other, intra-regional disparities at the county level are equally noteworthy and should not be overlooked.
Factors influencing the resilience of grain productionSelection of influencing factors
This study uses a two-way fixed effects model to estimate the parameters of the selected influencing factors, and the estimation results are presented in Table 2. The results indicate that from 2001 to 2023, grain yield per unit area, rural informatization level, proportion of employment in the primary industry, multiple cropping index, agricultural mechanization level, and investment level all had significant effects on county-level grain production resilience in China.
Estimation of bidirectional fixed effects on resilience in grain production.
Variable
Coefficient
Standard error
T-value
p-value
Constant
−231.875
8.420
−27.54
0.00***
Yield per unit area
30.823
0.904
34.08
0.00***
Level of agricultural mechanization
11.899
0.971
12.25
0.00***
Percentage of workforce employed in primary industries
−4.362
2.093
−2.08
0.00***
Crop rotation index
5.347
1.473
3.63
0.00***
Level of rural informatization
1.757
1.053
1.67
0.037**
Investment level
−3.183
0.509
−6.26
0.00***
*, **, and *** indicate significance at the 10, 5, and 1% levels, respectively.
By incorporating the influencing factors as independent variables and grain production resilience as the dependent variable, the Geographically Weighted Regression model was applied to further analyze the spatiotemporal evolution characteristics of the factors affecting grain production resilience in China. The regression coefficients obtained from the GWR model were used to construct Figure 5, which visualizes the spatial and temporal patterns of these influencing factors.
Spatial distribution of regression coefficients of influencing factors (2001–2023).
Eight maps of China display agricultural data from 2001 and 2024. Maps a and b show grain yield per unit area, with varying shades indicating different yield levels. Maps c and d illustrate the level of agricultural mechanization. Maps e and f present the percentage of population employed in primary industry. Maps g and h depict the crop rotation index. Colors range from green to red, representing different data intervals. Four maps of China show changes in rural informatization and investment levels from 2001 to 2024. Each map is color-coded: green indicates lower levels, while red indicates higher levels. Maps i and j display the level of rural informatization for 2001 and 2024, respectively. Maps k and l illustrate the investment levels for the same years. A scale bar is included for reference.Analysis of heterogeneity in influencing factorsGrain yield per unit area
Grain yield per unit area has a significant positive effect on grain production resilience that has steadily strengthened over time (Figures 5a,b). In 2001, high-impact areas were concentrated in western China, where yield increases effectively mitigated risks associated with fragile natural conditions. In contrast, the eastern region, characterized by a higher level of modernization, showed lower dependence on yield improvement. By 2023, the areas of strong influence had shifted to the Northeast, where warming climates and expanded sown areas substantially enhanced regional risk resistance through yield gains.
Agricultural mechanization level
Agricultural mechanization exhibits a generally positive effect on resilience (Figures 5c,d), with the most pronounced impact observed in the Northeast. The predominance of large-scale field crops and flat terrain in this region enables mechanization to significantly improve production efficiency and stability. In comparison, the central and western regions—constrained by fragmented landholdings, harsher natural conditions, and limited financial resources—benefit less from mechanization, resulting in weaker effects.
Proportion of employment in the primary industry
The proportion of employment in the primary industry has an negative influence on resilience, though the magnitude of this effect weakens over time (Figures 5e,f). In 2001, the negative impact was concentrated in western regions, where single-industry structures and labor-intensive production patterns constrained risk resistance. By 2023, the negative effect had shifted toward the Northeast and Inner Mongolia, reflecting the combined influence of labor outflow and insufficient industrialization. In parts of Southwest and Western China, higher labor concentration partially offset the disadvantages of resource scarcity, resulting in localized positive effects.
Multiple cropping index
The multiple cropping index has a consistently significant positive impact on resilience (Figures 5g,h), with its influence intensifying over time. In 2001, high-impact areas were mainly distributed along the eastern coastal regions, where higher cropping intensity directly translated into greater yields and resilience. By 2023, high-impact areas expanded toward the Northeast and several southern provinces, where agricultural modernization and improved land-use efficiency made the marginal benefits of multiple cropping more pronounced.
Rural informatization level
Rural informatization has a strong and gradually increasing positive influence on resilience that gradually increased over time (Figures 5i,j). In 2001, high-impact areas were concentrated in the Northeast, where informatization was initiated early and agricultural operations were highly scaled. By 2023, high-value areas expanded significantly, forming a “Northeast–Northwest belt-shaped cluster.” Enhanced informatization has strengthened production efficiency, early-warning capabilities, and risk prevention; the central region still lags behind, revealing a persistent “digital divide.”
Investment level
Investment level has a significant and continuously increasing effect on resilience that continuously increased over time (Figures 5k,l). In 2001, investment exhibited a “west-high, east-low” pattern, with some southwestern and northwestern provinces emerging as major high-value regions. By 2023, high-value areas had shifted to the eastern coastal region, the Yangtze River Delta, and the Huang–Huai–Hai Plain, where resources were increasingly concentrated in high-efficiency agricultural zones. Although investment has facilitated structural optimization and recovery capacity in the East, insufficient capital input in parts of the Central and Western regions has created “low-resilience belts,” thereby widening regional disparities.
The regression results indicate that grain yield per unit area, agricultural mechanization, multiple cropping index, and rural informatization all have significant positive effects on grain production resilience. Collectively, these factors enhance both the resistance and adaptability of the agricultural system. However, the proportion of agricultural labor and investment level deviate partially from theoretical expectations.
First, the negative coefficient of agricultural labor share contradicts the conventional assumption that labor input promotes productivity. This finding is nonetheless reasonable in light of China’s current rural labor conditions. With ongoing urbanization and labor migration, a large share of young and skilled workers have shifted to non-agricultural sectors, leaving behind an older and less-educated labor force. This demographic group has limited capacity to adopt advanced technologies and modern farming practices, thereby contributing little to resilience improvement. Furthermore, excessive reliance on manual labor rather than mechanization and technological input perpetuates a “smallholder inefficiency” pattern, weakening the system’s resilience. Hence, an increase in labor quantity does not necessarily translate into higher resilience; rather, labor quality and structural composition are more crucial.
Second, investment level showed unexpected negative effects in the two-way fixed effects model and in some GWR results. Three potential explanations can be proposed: (1) Structural imbalance of investment—a substantial portion of capital flows into rural infrastructure or non-agricultural sectors, with limited short-term effects on grain production resilience; (2) Low investment efficiency—mismatches exist between financial allocation and farmers’ actual needs, diminishing marginal returns; (3) Regional structural bias—in economically developed regions, investments are often directed toward high value-added industries rather than grain production, producing an apparent negative statistical association.
Yield, mechanization, multiple cropping, informatization, and investment generally enhance grain production resilience, whereas a high proportion of agricultural labor tends to weaken it. The intensity and spatial configuration of these effects demonstrate significant heterogeneity and interdependence: labor structure affects the efficiency of mechanization and informatization; informatization and mechanization promote yield and cropping intensity; and investment strengthens resilience by improving foundational conditions. Regionally, the Northeast should capitalize on its advantages in yield and mechanization, the Central and Western regions should accelerate infrastructure development and technology adoption, and the Western and Southwestern regions should prioritize improvements in informatization and water conservancy. Overall, grain production resilience in China is jointly shaped by regional endowments, technological progress, and policy interventions, displaying marked spatial differentiation at the national scale.
Discussion
The above results reflect how major policy shifts and external events have shaped grain production resilience over time. The identified stage-wise trends in resilience—for example, notable improvements around 2004 and after 2019—correspond to critical periods in China’s agricultural policy and development. From 2001 to 2004, resilience gains coincided with the introduction of supportive grain policies, whereas the 2004–2019 interval saw only modest improvements amid structural adjustments and market fluctuations. The surge in resilience after 2019 aligns with the implementation of the “Storing Grain in Land and Storing Grain in Technology” strategy and accelerated high-standard farmland construction, although external shocks like the COVID-19 pandemic introduced volatility. This temporal context suggests that policy interventions and extraordinary events have a discernible impact on resilience trajectories.
Regional disparities in resilience, as revealed by our analysis, are consistent with long-standing imbalances in resource endowments and development across China. We found that inter-regional differences (particularly the Northeast’s sustained high resilience versus the West’s lagging levels) dominate the national disparity, a pattern also noted by Chang and Jiang (2023) and Li et al. (2024) in regional studies. At the same time, significant intra-regional heterogeneity—for instance, within the Northeast and Eastern regions – becomes evident at the county scale, highlighting micro-level variations that province-level analyses (Yu et al., 2025) tend to obscure. By conducting a fine-grained county-level evaluation, this study offers a more nuanced picture of resilience patterns, underscoring the importance of scale in resilience research.
Our examination of the factors influencing resilience both confirms and extends prior findings in the literature. The positive impacts of grain yield per unit area, mechanization, multiple cropping intensity, and rural informatization on resilience align with theoretical expectations and empirical evidence (Fischer et al., 2021; Li et al., 2022; Zhao et al., 2023; Klerkx and Rose, 2020). These factors enhance agricultural systems by boosting productivity, efficiency, and adaptive capacity, thereby strengthening both resistance and recovery. In contrast, the proportion of labor in the primary industry shows a significant negative relationship with resilience, a result that appears counterintuitive but is supported by studies highlighting labor quality issues in agriculture (Dorward, 2020). In our context, this finding suggests that regions with a high share of agricultural employment often face an aging or less-skilled workforce that struggles to adopt modern techniques, leading to lower resilience. Thus, improving labor structure and human capital is more crucial than simply increasing agricultural labor input.
The role of investment in resilience emerges as complex. While we hypothesized a positive effect, our results show that higher investment levels do not uniformly translate into higher resilience and can even coincide with lower resilience in some areas. This outcome can be attributed to several factors. First, a substantial portion of “investment” in rural areas may flow into infrastructure or non-grain sectors whose benefits to grain resilience materialize only in the long term. Second, inefficiencies in investment allocation—such as funds not aligning with farmers’ needs or being absorbed by administrative costs—can diminish the immediate impact on production resilience. Third, there is a regional bias: more developed regions often channel investments into high value-added industries rather than grain production (Cai et al., 2025), which may explain the observed weak correlation between investment and resilience. In other words, simply increasing financial input is insufficient; the effectiveness and targeting of that investment are key. This finding adds a nuanced perspective to the literature, emphasizing that institutional factors and investment quality must be considered when bolstering resilience.
This highlights that advancing technological and productive capacities consistently bolsters grain production resilience, which resonates with the broader literature on agricultural modernization and resilience. Structural challenges—notably, an unfavorable labor composition and suboptimal investment efficiency—can undermine resilience, aligning with calls in recent research for addressing human and institutional factors (Chen et al., 2020; Zhou et al., 2022). By situating these findings in the context of existing studies, our analysis underscores the importance of a multifaceted approach to strengthening grain production resilience: one that combines technological upgrades with structural reforms. This study contributes to the ongoing scholarly dialogue by providing empirical evidence at the micro level and linking it to macro-level policy implications, thereby deepening the understanding of how and why grain production systems succeed or struggle in the face of disturbances.
Conclusions and recommendationsKey findings
Using the core-variable method, this study measured the level of grain production resilience across Chinese counties and constructed the GWR model to identify the influencing factors. Based on this framework, the spatiotemporal evolution of grain production resilience and the spatial heterogeneity of its determinants from 2001 to 2023 were comprehensively analyzed. The results reveal the following:
(1) Over the study period, China’s grain production resilience exhibited a fluctuating yet upward trend. Regionally, from 2001 to 2004, resilience increased markedly in the Northeast, East, and Central regions, while the Western region experienced only limited growth. Between 2004 and 2019, resilience in the first three regions declined slightly before stabilizing, whereas changes in the West remained minimal. From 2019 to 2023, resilience in the Northeast, East, and Central regions first rose and then fell, while the Western region remained largely stable. The spatial pattern shows a long-term high level in the Northeast, a moderate level in the Central region, and a gradual decline from East to West, reflecting pronounced regional imbalance in evolutionary trajectories.
(2) The disparities in grain production resilience were substantial throughout the study period, displaying repeated fluctuations and distinct stage characteristics. These disparities mainly stemmed from inter-regional differences. Among the four regions, the East exhibited the greatest internal variation, followed by the Northeast and West—both lower in resilience than the East—while the Central region showed the smallest internal gap, though it has gradually widened in recent years.
(3) Grain production resilience was influenced by multidimensional factors with evident spatiotemporal effects. Grain yield per unit area, mechanization level, informatization level, and investment level all exerted significant positive impacts, collectively strengthening system resistance and adaptability. In contrast, the proportion of employment in the primary industry consistently showed a significant negative effect, with its spatial center of influence shifting from the Northwest to the Northeast and its negative impact becoming increasingly pronounced over time.
Policy recommendations
Grain production resilience in China exhibits both national and regional characteristics. Different regions demonstrate considerable variation in their capacity to withstand and respond to climate anomalies, market fluctuations, and the non-agriculturalization of production factors. Based on the main findings of this study, the following policy recommendations are proposed:
(1) Enhance region-specific strategies and optimize structural transformation. Each region should formulate differentiated policies according to its dominant factors and local conditions. Efforts should focus on optimizing cropping structures, appropriately integrating external resources and markets, and improving technological and institutional innovation. The eastern and central regions should accelerate the upgrading of agricultural structures, strengthen the role of new agricultural business entities, and promote the transformation of government functions toward service-oriented agricultural governance. These regions should pursue green and high-quality development to enhance adaptive and recovery capacities. The western and northeastern regions should prioritize the introduction of financial investment, advanced technologies, and mechanization services; improve informatization and water-saving technologies; and accelerate structural transformation to consolidate resistance and reinforce recovery capacity simultaneously.
(2) Strengthen regional coordination and spatial linkages. Given the significant spatial correlation of grain production resilience, it is essential to establish regional coordination organizations and improve interregional mechanisms for high-standard farmland construction, irrigation and drainage systems, and storage–logistics facilities. A well-integrated production–marketing and information-sharing network should also be developed to promote interprovincial exchanges and factor mobility. By leveraging the spatial spillover effects of high-resilience regions, surrounding areas can be guided toward synchronized improvement, achieving both regional coordination and high-quality development in national food security.
This study, using the county level as the unit of analysis, systematically measured the spatiotemporal evolution of grain production resilience in China from 2001 to 2023 and examined its key influencing factors. Despite these contributions, several limitations remain. First, constrained by data availability, this research employed the core-variable method, using grain yield fluctuation as the proxy indicator. Although this metric effectively captures the system’s resistance to external shocks, it does not comprehensively represent the social, ecological, and institutional dimensions of resilience. Second, while the GWR model revealed spatial heterogeneity in influencing factors, it did not fully account for spatial spillover effects or regional interdependencies. Future research could apply spatial Durbin models or multiscale geographically weighted regression to further explore multi-level spatial dynamics. Third, this study did not explicitly incorporate external shocks such as extreme climate events or international grain price volatility, leaving the dynamic responses of resilience insufficiently examined. Future studies could integrate multisource data—including climatic, market, and policy information—to investigate adaptive evolution mechanisms of the grain system under multiple risk scenarios. At the micro level, household surveys or remote sensing data could be incorporated to uncover the mechanisms of resilience formation from behavioral and land-use perspectives. Overall, future research should strengthen data integration and model innovation to deepen understanding of the dynamic, spatial, and institutional coupling mechanisms of grain production resilience, thereby providing more robust scientific support for ensuring national food security and promoting sustainable agricultural development.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
CD: Writing – original draft. ST: Writing – original draft.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsufs.2025.1740869/full#supplementary-material
ReferencesAdgerW. N.BrownK.NelsonD. R. (2011). Resilience implications of policy responses to climate change. Wiley Interdiscip. Rev. Clim. Change2, 757–766. doi: 10.1002/wcc.133CaiD.WangY.ZhaoX.WuJ.LuY.XiB. (2025). A paradigm shift driven by multi-source data, mechanistic insights, and adaptive machine intelligence for sustainable agriculture. Crit. Rev. Environ. Sci. Technol.55, 1578–1598. doi: 10.1080/10643389.2025.2557306ChangJ.JiangH. (2023). Spatio-temporal differentiations and influence factors in China’s grain supply chain resilience. Sustainability15:8074. doi: 10.3390/su15108074ChenJ.LiX.HuangY. (2020). Labor migration, mechanization, and agricultural productivity: evidence from China. World Dev.136:105110.ChengW.LiY.ZuoW.DuG.StannyM. (2024). Spatio-temporal detection of agricultural disaster vulnerability in the world and implications for developing climate-resilient agriculture. Sci. Total Environ.928:172412. doi: 10.1016/j.scitotenv.2024.172412, 38614341DorwardA. (2020). Agricultural labour productivity and sustainable development in the digital age. Food Policy97:101911.FanS.ZhangX. (2020). Resilient food systems for sustainable development. Nat. Food1, 8–9.FangY.ZhouT.WangL. (2021). Agricultural technological innovation and production resilience. Technol. Forecast. Soc. Change168:120781.FischerR. A.ByerleeD.EdmeadesG. O. (2021). Crop yields and global food security: will yield increase continue?Field Crop Res.263:108104.FotheringhamA. S.BrunsdonC.CharltonM. (2002). Geographically weighted regression: the analysis of spatially varying relationships. Chichester: Wiley.GollinD.LagakosD.WaughM. E. (2021). Agricultural productivity and development: revisiting cross-country patterns. J. Dev. Econ.150:102623.HanY.ZhangL.ZhaoX. (2024). Assessing agricultural resilience to climate change in major grain-producing areas of China. Agric. Syst.225:103652.HertelT. W. (2022). The challenges of ensuring global food security under climate change. Nat. Food3, 408–416.HollingC. S. (1973). Resilience and stability of ecological systems. Annu. Rev. Ecol. Evol. Syst.4, 1–23. doi: 10.1146/annurev.es.04.110173.000245HuangJ.HuR.RozelleS.LiY. (2022). Agricultural insurance and resilience enhancement under climate uncertainty. Glob. Environ. Change74:102513.KlerkxL.RoseD. (2020). Dealing with the game-changing technologies of agriculture 4.0: how do we manage diversity and responsibility in food systems innovation?Glob. Food Secur.24:100347. doi: 10.1016/j.gfs.2019.100347LeskC.RowhaniP.RamankuttyN. (2021). Influence of extreme weather disasters on global crop production. Nat. Commun.12, 1–7.LiW.GuoJ.TangY.ZhangP. (2024). Resilience of agricultural development in China’s major grain-producing areas under the double security goals of “grain ecology”. Environ. Sci. Pollut. Res.31, 5881–5895. doi: 10.1007/s11356-023-31316-8, 38133757LiT.LiZ. (2023). Exploring spatial heterogeneity of agricultural production resilience in China. J. Clean. Prod.396:136452.LiF.ZhaoZ.WangY. (2023). Measuring agricultural production resilience under multiple shocks: a spatial perspective. Ecol. Indic.156:111072.LiX.ZhouY.TangH. (2022). Multiple cropping intensity and agricultural sustainability in China: spatiotemporal patterns and driving factors. Land Use Policy120:106288.LiuJ.ChenH.WangX. (2021a). Regional patterns of agricultural resilience in China’s main grain production areas. Land Use Policy109:105654.LiuY.HuS.ChenX. (2022). Agricultural service provision and production resilience in smallholder systems. Agric. Syst.199:103407.LiuJ.ZhangD.LiY. (2021b). Agricultural investment, infrastructure, and food security in China. Land Use Policy109:105685.MartinR.SunleyP. (2015). On the notion of regional economic resilience: conceptualization and explanation. J. Econ. Geogr.15, 1–42.MeuwissenM. P. M.FeindtP. H.SpiegelA.TermeerC. J. A. M.MathijsE.MeyY. d.. (2019). A framework to assess the resilience of farming systems. Agric. Syst.176:102656. doi: 10.1016/j.agsy.2019.102656TendallD. M.JoerinJ.KopainskyB.EdwardsP.ShreckA.LeQ. B.. (2015). Food system resilience: defining the concept. Glob. Food Secur.6, 17–23. doi: 10.1016/j.gfs.2015.08.001UNICEF (2021). The state of food security and nutrition in the world 2021. FAO. doi: 10.4060/cb4474enWangX.LiQ.ChenJ. (2023). The role of labor structure and technology adoption in agricultural sustainability: evidence from China. J. Clean. Prod.398:136729.WangF.ZhaoW. (2020). Agricultural training and resilience building in rural China. Sustainability12:5678.WangQ.ZhaoY.XuL. (2023). Spatiotemporal evolution and determinants of agricultural resilience in the Yellow River Basin. Environ. Sci. Pollut. Res.30, 57632–57648.WuD. (2020). Spatially and temporally varying relationships between ecological footprint and influencing factors in China’s provinces using GWR. J. Clean. Prod.261:121089. doi: 10.1016/j.jclepro.2020.121089YuX.XiaoR.ZhangZ.ZhouF.MoF.XuZ.. (2025). Rising disparities in grain self-sufficiency across China: provincial divergence amidst overall national improvement. Environ. Impact Assess. Rev.114:107942. doi: 10.1016/j.eiar.2025.107942YuH.ZhangX.WangY. (2022). Measuring agricultural system resilience under climate shocks. Agric. Syst.203:103507.ZhangK.LiuY.FangY. (2021). The role of farmland infrastructure in enhancing agricultural productivity resilience. Land10:812.ZhangY.WangL.LiJ. (2022). Assessing agricultural resilience using multidimensional indicators: evidence from China. Sustainability14:1229.ZhangF.WangH.QinT. (2023). Towards sustainable management of agricultural resources: assessing the relationship between water, soil, and food security. J. Environ. Manag.344:118401. doi: 10.1016/j.jenvman.2023.118401ZhangL.YuX.SunS. (2024). Digital agriculture and rural resilience: evidence from China’s county-level data. Comput. Electron. Agric.222:109678.ZhaoJ.ChenY.LiangS. (2023). Impacts of mechanization and digitalization on sustainable agricultural productivity in China. Resour. Conserv. Recycl.197:107156.ZhouJ.XuL.WangH. (2022). Does agricultural insurance enhance the resilience of food systems?Food Policy112:102340.
Edited by: Kai Zhu, Hubei University, China
Reviewed by: Xiangjin Shen, Chinese Academy of Sciences (CAS), China