Spatiotemporal Differentiation Characteristics and Zoning of Cultivated Land System Resilience in the Songnen Plain

Abstract

Enhancing cultivated land system resilience is a fundamental prerequisite for improving land use efficiency and thus addressing climate change. Taking the Songnen Plain—a major grain production area in China—as the study region, this study constructs a definition of cultivated land system resilience from three dimensions (“resistance–adaptability–reconstruction”). An index system for resilience evaluation is established, and methods such as three-dimensional Euclidean distance and K-means clustering are employed to investigate the spatiotemporal differentiation characteristics of cultivated land resilience in the Songnen Plain from 2001 to 2021. Based on the findings, zoning is performed and corresponding management strategies are proposed: (1) Overall resilience in the Songnen Plain increased from 0.4450 in 2001 to 0.7469 in 2021; enhanced resistance played the most significant role in promoting this increase. (2) The Songnen Plain exhibited pronounced spatial differentiation in cultivated land resilience, characterized by higher resilience in the eastern region and weaker resilience in the central and western regions. (3) The zoning results reveal significant disparities in resilience levels within the study area; targeted measures are thus required to address key problems in each zone. This study provides theoretical insights and empirical conclusions for formulating differentiated protection policies for cultivated land systems, thereby ensuring the sustainable development of the Songnen Plain’s cultivated land system.

1. Introduction

In the 21st century, intensified climate change and human activities have reshaped global cultivated land systems, leading to unprecedented pressures and challenges [1]. Studies have shown that problems such as soil quality degradation and biodiversity loss are becoming increasingly severe, posing serious threats to global food security and ecosystem services [2,3]. External disturbances, including extreme weather events and unsustainable land use practices, further exacerbate the vulnerability of cultivated land systems [4,5]. Traditional approaches to cultivated land management mainly target single issues—such as improving soil quality or mitigating land degradation—yet have proved insufficient in confronting the increasingly complex, dynamic, and multi-dimensional external disturbances. To address these challenges, the concept of resilience has emerged as a key framework for understanding how cultivated land systems resist external disturbances, adapt to environmental changes, and achieve a new equilibrium through transformation [6,7,8]. Resilience provides a fresh perspective on how land systems endure external pressures and maintain long-term productivity in the face of climate change and anthropogenic stress [9,10]. Consequently, resilience has gradually been acknowledged by the global scientific community and policymakers as a critical indicator for sustainable land management.

Located in Northeast China, the Songnen Plain is one of the country’s key agricultural production areas. However, it faces multiple challenges—including intensified climate change, land degradation, and urbanization pressures—which threaten the sustainability of agricultural production [11,12]. Hence, there is an urgent need for sustainable land management strategies and adaptive agricultural practices to bolster the resilience of farmland systems in this region and mitigate the pressures imposed by external disturbances.

Building on the above discussion and incorporating the processes that constitute resilience, this study examines cultivated land system resilience from the three dimensions of resistance, adaptability, and reconstruction. Using the Songnen Plain as a case study, we construct an evaluation index system for cultivated land system resilience, analyzing the system’s spatiotemporal evolution and regional differentiation in 2001, 2011, and 2021. This study aims to explore the impacts of modern economic and technological inputs on farmland utilization, offering fresh insights into how external risks can be managed and providing valuable references for sustainable farmland development. Such research is crucial for formulating effective management strategies, restoring and enhancing resilience, strengthening the capacity to withstand disturbances, and achieving the sustainable use of farmland resources—thereby ensuring the long-term sustainability of agricultural productivity, ecological stability, and land resources in the Songnen Plain.

2. The Evolution and Functional Mechanisms of Cultivated Land System Resilience

2.1. The Evolution of Cultivated Land System Resilience

Resilience theory has broadly undergone three developmental stages: engineering resilience (prior to 1973) [13,14], ecological resilience (from 1973 to the early 21st century) [15], and socio-ecological system (SES) resilience (from the early 21st century to the present) [16,17]. From the perspective of its origin, Holling [15] was the first to introduce the concept of “resilience” into the field of ecology in 1973, gradually shifting the focus toward the functional dynamics of ecosystems. In recent years, under the dual context of global climate change and sustainable development, resilience has evolved significantly and is increasingly recognized as a key attribute that enables systems to effectively respond to future uncertainties and potential shocks by adopting adaptive strategies. This evolution has integrated perspectives from ecology, economics, sociology, and other disciplines. The shift toward systems thinking and holistic analyses has not only deepened the theoretical understanding of resilience but also broadened its practical applications across various fields.

Currently, the incorporation of resilience theory into cultivated land system research remains limited. Existing studies on cultivated land—both domestic and international—have primarily focused on farmland quality [18,19,20], ecological security [21,22,23], and land use efficiency [24,25,26], emphasizing the external drivers of land system change, such as land use transition and climate impacts. However, they often overlook the self-regulating and disturbance-resistant capacities that farmland systems exhibit throughout their evolution [27,28]. In 2012, Zhao Huafu et al. [29] first introduced the concept of resilience into cultivated land research, highlighting the land system as an elastic body with self-organizing characteristics. In recent years, the number of studies on cultivated land use system resilience has gradually increased [30], though there is still no widely accepted or unified definition of its core connotation. From a methodological perspective, the main assessment approaches include the key variable method, composite index method, and participatory assessment [31,32]. Among them, the composite index method can be further classified into three categories: frameworks based on resilience functions [33], research subjects [34], and processes through which resilience occurs [35]. Regarding its core functional dimensions, beyond traditional functional analysis, recent research has expanded to cover the resource resilience, production resilience (PR), ecological resilience (ER), and structural resilience of cultivated land scale [30], as well as economic resilience (CR) [10]. The key roles of cultivated land system resilience lie in maintaining agricultural productivity, preserving ecological functions, coping with external disturbances, and enhancing long-term system stability and sustainability [36,37]. The evolution of cultivated land resilience theory—grounded in both resilience and sustainability science—has followed a trajectory from a single equilibrium to multi-feedback loops and adaptive cycles. It has since been widely extended to studies on ecosystem security [38], land security and thresholds [39], and agricultural resilience mechanisms [40], giving rise to derivative concepts such as agricultural resilience, land–ecology–economy resilience, and resilience indices [41,42]. Regarding indicator systems, scholars widely agree that the natural attributes of cultivated land, farming behaviors, and agricultural pollution all affect the use and protection of cultivated land [43,44]. Accordingly, multi-dimensional indicator systems based on the pressure–state–response (PSR) framework have been developed to evaluate cultivated land system resilience from multiple perspectives, such as the “functional structure–stability–redundancy–adaptive cycle” and “productivity–health–farming–use” [45,46].

Compared with the advances made in other fields, existing studies on cultivated land use system resilience lack analytical frameworks from a process-based perspective. Most resilience research has been conducted at the provincial [47] or municipal [48] scale, with relatively few studies focusing on the county-level, micro-scale units that are essential for the refined management and zoned regulation of cultivated land [49]. Moreover, there is a shortage of comprehensive studies that integrate resistance, adaptability, and reconstruction as internal dimensions of farmland system resilience. In regions subject to complex agricultural pressures—such as soil erosion, fertility loss, and socio-economic constraints—analyzing the resilience of the system is crucial for the formulation of effective land management strategies. To date, resilience research in fields such as engineering [50,51], urban systems [52,53], and ecosystems [54,55] has achieved substantial progress, offering valuable theoretical support for understanding and enhancing the resilience of farmland systems.

2.2. The Connotation and Mechanisms of Cultivated Land System Resilience

The cultivated land system is a semi-natural, human-modified system formed by human intervention in natural ecosystems to meet societal needs. It relies on land resources and utilizes crop growth and reproduction to produce agricultural products [56]. The use of cultivated land has evolved from traditional extensive agriculture, which heavily depended on natural resources, to a more technology-driven stage characterized by increased yields and environmental sustainability during the Green Revolution. This evolution reflects humanity’s ongoing exploration and improvement of agricultural production methods and land management concepts. Throughout this dynamic process, the cultivated land system itself has undergone structural adjustments to continually adapt to pressures arising from environmental, social, and economic changes, demonstrating its resilience.

By integrating theories from various disciplines, the resilience of a cultivated land system can be understood as its capacity to maintain core functions, recover quickly, and adapt to changes when facing natural, economic, or social disturbances. Applying resilience theory to cultivated land systems facilitates an analysis of their resilience evolution under external disturbances (Figure 1). When external disturbances such as climate change, human activities, and soil erosion occur, they directly or indirectly affect both the natural ecological and socio-economic subsystems. The core of the cultivated land system lies in the tight interconnection between its ecological and socio-economic components. The ecological system provides resources and ecosystem services that sustain food production and social stability, while external shocks disrupt these resources and functions, threatening system stability. To maintain stability and functionality, the system relies on three core resilience mechanisms: resistance, adaptability, and reconstruction.

Resistance refers to the system’s ability to withstand initial disturbances or pressures, preventing structural disorder and functional degradation [57]. This inherent capacity forms the foundation of cultivated land resilience. When the system faces long-term environmental or socio-economic pressures, adaptability comes into play. Through adjustments in management practices or production methods, the system can restore or enhance its functions. Adaptability thus supports and drives resilience; in the face of significant changes, the system may undergo deeper structural and functional transformations, leading to a new stable state in response to altered environmental conditions. This process, known as reconstruction, ensures the system’s resilience can be renewed and upgraded. These three mechanisms—resistance, adaptability, and reconstruction—interact within the system (Figure 2). When resistance is insufficient to withstand prolonged or severe impacts, the adaptability mechanism is triggered. By optimizing resource management or adjusting production methods, the system adapts to the changing environment, preserving functionality and even enhancing future resistance. If adaptability proves inadequate in response to strong disturbances, the system must undergo deeper reconstruction. Through this process, the system acquires new structures and functions, enhancing future resilience through positive feedback loops. However, frequent reconstructions may lead to increased vulnerability, creating negative feedback effects.

3. Materials and Methods

3.1. Overview of the Study Area

The Songnen Plain (119°52′–132°31′ E, 41°42′–51°38′ N) is located in Northeast China (see Figure 3) and comprises 64 counties (cities/districts), making it one of the largest plains in China. Renowned for its black soil resources, it is also a major production area for staple crops such as corn, rice, and soybeans. According to 2023 grain production data released by the National Bureau of Statistics, the Songnen Plain accounts for about 22.9% of China’s total grain output. Characterized by a temperate continental monsoon climate, the region has an annual average temperature ranging from 0 °C to 4 °C and an average annual precipitation of about 400–600 mm. Its terrain is generally flat, and the soil is highly fertile. The study area covers approximately 1.57 × 10⁴ km² in total, with cultivated land amounting to 8.31 × 10⁴ km²—about 52.83% of the total area—encompassing the western hilly region, central lowlands, and eastern piedmont plain.

3.2. Data Sources

The socio-economic data utilized in this study were derived from statistical yearbooks and bulletins of Heilongjiang Province, Jilin Province, Harbin City, Daqing City, Qiqihar City, Suihua City, Heihe City, Changchun City, Baicheng City, Jilin City, and Songyuan City for the years 2001, 2011, and 2021. The land use data were obtained from the Resource and Environment Science Data Center of the Chinese Academy of Sciences (https://www.resdc.cn/DOI/ (accessed on 18 April 2025)). The net primary productivity (NPP) data were acquired from the MOD17A3HGF.006 dataset and aggregated spatially by county, based on the 2021 administrative division of China, using the mean value for statistical purposes. A 30-m Digital Elevation Model (DEM) was sourced from the ASTER Global Digital Elevation Model V003 dataset. The meteorological data were collected from the National Earth System Science Data Center, specifically the China Temperature and Precipitation Dataset with a 1 km spatial resolution covering 2000–2020 (http://www.geodata.cn/ (accessed on 18 April 2025)) (

Table 1. Data sources.

Data	Type	Source
Land Use Data	Raster Data	Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences
NPP	Raster Data	MOD17A3HGF.006 Data
DEM	Raster Data	ASTER GDEM Digital Elevation Model
Meteorological Data	Raster Data	National Earth System Science Data Center
Administrative Boundaries	Vector Data	Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences
Socio-economic Data	Statistical Data	Statistical Yearbooks of Various Provinces and Cities

). To address adjustments made to certain county or district names, this study standardized the county/district names from 2010 to match those of 2021 based on administrative boundaries and name-change references, facilitating a comparative analysis. To ensure data integrity, district-level administrative units within each prefecture-level city were merged. Missing values for specific years or regions were filled using adjacent values or linear interpolation methods.

3.3. Research Methods

3.3.1. Development of the Cultivated Land System Resilience Evaluation Framework

To effectively measure the resilience of cultivated land systems, this study constructs a comprehensive indicator system that integrates multiple dimensions, including economic, technological, productive, and ecological factors. Resistance is evaluated through environmental indicators such as precipitation and net primary productivity (NPP), adaptability is measured using socio-economic variables such as agricultural output, and reconstruction is quantified by technical and structural indicators such as vegetation cover and mechanization level. This multi-dimensional approach reflects the inherent complexity of resilience and provides a framework to better understand how cultivated land systems respond to dynamic pressures (

Table 2. Evaluation index system for cultivated land system resilience in the study area.

Criteria Layer		Indicator	Indicator Definition	Attribute	Weight
Primary Level	Secondary Level
Resistance	Cultivated Land Production Sustainability	Cultivated Land Area Change Rate	(Calculated as cultivated land area in the previous year/current cultivated land area/spatial unit) × 100; higher values indicate greater land fluctuation, negatively affecting system stability and resilience	−	0.0166
		Multiple Cropping Index	Measurement of land use efficiency; reflects the degree to which cultivated land in the study area is effectively used within a year, calculated as follows: Total Sown Area/Total Cultivated Land Area × 100%	+	0.1597
		Net Primary Productivity	Indicates productivity fluctuation; higher NPP reflects stronger ecosystem function and agricultural potential	+	0.2774
	Climate Conditions	Annual Precipitation	Considers the impact of natural baseline factors on the resistance of cultivated land	+	0.3522
		Sunshine Hours	Indicates solar radiation intensity; more sunlight supports photosynthesis and longer growing periods, enhancing productivity and climate adaptability	+	0.1941
Adaptability	Socio-Economic Conditions	Primary Industry Output Value	Measurement of social state changes; reflects the economic benefits of cultivated land	+	0.5125
		Proportion of Rural Population	Higher proportion of rural population; cultivated land with a higher rural population proportion has a stronger response to various changes	+	0.1700
		Patch Cohesion Index	Higher cohesion index; indicates a stronger ability to adapt to disturbances	+	0.0272
		Per Capita Cultivated Land Area	More cultivated land per capita indicates lower land pressure and better resource availability, supporting sustainable management	+	0.2903
Reconstruction	Agricultural Management Capacity	Total Agricultural Machinery Power	The level of agricultural machinery reflects the agricultural technological level within the region	+	0.3207
		Amount of Fertilizer Applied	The capacity to respond to cultivated land that has been subjected to human disturbances exceeding a threshold	−	0.0382
	Ecological Restoration Capacity	Fractional Vegetation Cover	Reflects the ecological condition of cultivated land; higher coverage indicates better ecological restoration, soil, and water conservation and stronger risk resistance	+	0.0545
		Landscape Evenness	Higher landscape evenness indicates a more uniform distribution of landscape types, which enhances the cultivated land system’s ability to reconstruct itself	+	0.1668
		Patch Density	Greater patch density suggests a stronger ability to learn and reconstruct in response to external disturbances	+	0.4197

Note: “+” indicates a positive indicator attribute, while “−” indicates a negative indicator attribute.

).

In addition, the Stata 15.0 software platform was used to standardize the raw data to eliminate differences in measurement units and ensure that all indicators could be compared and weighted on a common scale. Since the traditional entropy weight method may produce biased results when the entropy value approaches 1—leading to disproportionately small weights for certain indicators with low variation—this study applies an improved entropy weighting method (see Equations (1)–(7)) to calculate the weight coefficients shown in Table 1, resulting in a more reasonable and scientifically grounded distribution of weights. This modification makes the weight distribution more reasonable and avoids distortion under specific conditions, thereby ensuring that the weight coefficients are more accurate and scientifically grounded. The final indicator weights used in the evaluation are presented in Table 1. This framework facilitates a robust and integrated assessment of the overall resilience of the cultivated land system.

3.3.2. Improved Entropy Weight Method

To scientifically and objectively determine the weights of each indicator, this study employs the entropy weight method as a weighting tool. By standardizing the raw data, the information entropy and redundancy for each indicator can be calculated, thus reflecting the indicator’s importance in the comprehensive assessment. Additionally, to address the distortion in weight allocation that arises when the entropy value is close to 1 in the conventional entropy weight method, this study introduces an improved formula to optimize weight allocation, making the results more reasonable and accurate [58].

$$Q_{\mathit{ij}}={\displaystyle \frac{x_{ij}-\min (x_{ij})}{\max (x_{ij})-\mathit{min}(x_{ij})}}$$

(1)

$$Q_{\mathrm{ij}}={\displaystyle \frac{\max (x_{ij})-x_{ij}}{\max (x_{ij})-\min (x_{ij})}}$$

(2)

$$P_{ij}={\displaystyle \frac{x_{ij}}{\sum _{i=1}^n{x^{\prime }}_{ij}}}$$

(3)

$$e_i=-{\displaystyle \frac{l}{\ln n}}{\sum }_{i=1}^n{P}_{ij}\ln P_{ij}$$

(4)

$$f_i=1-e_i$$

(5)

$$W_{ij}={\displaystyle \frac{f_i}{{\sum }_{j=1}^m{f}_i}}$$

(6)

where Q_ij represents the standardized values derived from the original data; the maximum value of the original data is denoted by max(x_1j, x_2j, x_3j, …, x_nj); t represents the number of years; n indicates the number of counties (or districts); m is the number of evaluation indicators; x_ij represents the numerical value of the j^th evaluation indicator for the i^th county (or district) (i = 1, 2, …, n; j = 1, 2, …, m); and W_ij denotes the weights of indicators.

However, determining indicator weights using Equation (6) presents certain limitations: when the entropy e_i approaches 1, even slight differences can significantly alter the indicator weight W_i [59]. Examples illustrating the limitations of the traditional entropy weight method are shown in

Table 3. Examples showing limitations of the traditional entropy weight method.

Entropy Value			Entropy Weight
e1	e2	e3	W1	W2	W3
0.9999	0.9998	0.9997	0.1667	0.3333	0.5

.

$$W_{ij}=\left\{\begin{array}{ll}(1-{\overline{e}}^{35.35}){\displaystyle \frac{f_i}{{\sum }_{i=1}^m{f}_i}}+{\overline{e}}^{35.35}{\displaystyle \frac{1+\overline{e}-e_j}{{\sum }_{k=1,e_k\ne 1}^m(1+\overline{e}-e_k)}}& e_j<1\\ 0& e_j=1\end{array}\right.$$

(7)

where $\overline{e}$ denotes the average of all entropy values that are not equal to 1.

3.3.3. Three-Dimensional Euclidean Distance

Three-dimensional Euclidean distance is a commonly used spatial distance measure in multivariate analysis [60]. It requires converting observations into standardized three-dimensional coordinates. By combining the values of the “resistance–adaptability–reconstruction” dimensions into a single distance measure, it comprehensively assesses and quantifies differences in resilience among regions, reflecting the temporal and spatial changes across all dimensions and helping reveal the overall dynamic evolution of resilience. The strength of cultivated land system resilience is therefore characterized using three-dimensional Euclidean distance. The cultivated land system resilience measurement model is expressed as follows:

$$l=W*Q$$

(8)

$$V=\sqrt{({l_1}^2+{l_2}^2+{l_3}^2)}$$

(9)

where V represents the resilience value of the cultivated land system, and l₁, l₂, and l₃ represent resistance, adaptability, and reconstruction values, respectively.

3.3.4. Kernel Density Estimation

Kernel density estimation is a commonly used non-parametric method for estimating the density function of data points [61]. In this study, kernel density estimation is applied to identify temporal distribution patterns of cultivated land system resilience in the Songnen Plain. By calculating the density of points in different regions, the temporal evolution of resilience can be visualized more intuitively. The kernel density function is expressed as follows:

$$f(x)={\displaystyle \frac{1}{nh}}\sum _{i=1}^{n}K\left({\displaystyle \frac{x-x_i}{h}}\right)$$

(10)

where f(x) represents the kernel density estimate at point x; n represents the number of samples; h is the bandwidth (smoothing parameter), determining the width of the kernel function; K denotes the kernel function; and x_i is the i^th sample point.

3.3.5. K-Means Clustering

K-means clustering is a widely used unsupervised learning algorithm that divides datasets into k distinct clusters [62]. Data points within each cluster are highly similar to each other and exhibit notable differences from points in other clusters. The algorithm classifies data by minimizing the distance between the data points and their corresponding cluster centroids. By grouping counties based on their resilience characteristics, the algorithm helps identify regions with similar resilience patterns and highlight key areas for targeted management and intervention. This approach enables a more nuanced understanding of regional variations in cultivated land resilience and supports the development of tailored strategies for sustainable land use.

$$s(c,m_1,m_2,m_3..m_p)={\sum _{z=1}^p‖x_z-m_{c\left(p\right)}‖}^2$$

(11)

$${c_i={\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{m}\mathrm{i}\mathrm{n}}_{je}}_{(1,2\dots j)}‖x_i-m_i‖,i\in N_n$$

(12)

$$m_p={\displaystyle \frac{1}{n_p}}\sum c\left(i\right)=p^{x_i},p\in N_p$$

(13)

$$‖x-y‖=\sqrt{\sum _{i=1}^q(x_i-y_i{)}^2}$$

(14)

where z denotes the index of a data point; p represents the total number of clusters (the K-value); X_z indicates the z^th data point; m_c(p) denotes the centroid of the p^th cluster; x and y represent two data points; q refers to the dimension of the data; x_i and y_i are the coordinates of the i^th dimension; C_i denotes the cluster number to which the i^th data point belongs; “argmin” represents finding the index that minimizes the objective function; specifically, each data point x_i is assigned to its nearest centroid m_j, where j is the cluster number (from 1 to j); N denotes the set of all data points within the dataset; n_p indicates the total number of data points belonging to the p^th cluster (cluster size); c(i) refers to data points belonging to the p^th cluster; p_i represents the coordinate values of each data point in the p^th cluster; and N_p denotes the set of all clusters.

4. Results

4.1. Spatiotemporal Differentiation Characteristics of Overall Cultivated Land System Resilience in the Study Area

Overall, the cultivated land system resilience in the study area exhibited a significant upward trend from 2001 to 2021. The resilience improved by 0.027 more notably during the period from 2011 to 2021 compared with from 2001 to 2011. However, the overall resilience level was moderate to high, indicating room for further improvement. A comparison between the resilience indices from 2001 to 2011 and from 2011 to 2021 reveals that resilience enhancement was more pronounced in the latter decade (Figure 4).

Analyzed using different dimensions, the resistance, adaptability, and reconstruction of the cultivated land system in the Songnen Plain all demonstrated varying degrees of enhancement between 2001 and 2021. Ranked by their intensity of contribution from lowest to highest, the dimensions were adaptability (0.1525–0.2427), reconstruction (0.2627–0.3297), and resistance (0.2968–0.3828). Ranked by their magnitude of increase from highest to lowest, they were resistance (0.2652), adaptability (0.1331), and reconstruction (0.1112).

These findings indicate that the study area’s cultivated land possesses a relatively strong natural baseline and favorable external conditions, enabling the system to remain stable and effectively resist environmental pressures and external shocks. The adaptability indicator exhibited a steady upward trend but with a relatively small increase, suggesting weaker socio-economic influence, insufficient agricultural management capacity, and limited ecological recovery ability, resulting in lower resilience to environmental dynamics and slower recovery after disturbances. Notably, enhanced resistance was identified as the primary contributor to improvements in cultivated land system resilience, followed by adaptability and reconstruction.

Using the natural breaks method in ArcGIS 10.8, cultivated land resilience was classified into five levels: low, relatively low, medium, relatively high, and high. To clearly illustrate temporal changes in cultivated land system resilience across the Songnen Plain, kernel density curves were employed. The cultivated land system in the Songnen Plain exhibited significant changes over time: In 2001, most regions had medium resilience or below, with low and relatively low resilience concentrated primarily in the central and southern regions. By 2011, the overall resilience level improved notably, with the proportion of areas classified as low and relatively low resilience, decreasing from 92% to 35%, while areas with medium and relatively high resilience expanded from 8% to 65%, indicating substantial improvements in the central and northern regions. By 2021, low and relatively low resilience levels were observed only in a few counties, signifying a significant improvement in cultivated land resilience across most of the study area. Medium and relatively high resilience became predominant, especially in the central and northern regions, reflecting sustained progress in agricultural management, technological application, and ecological restoration. Additionally, some regions attained high resilience levels.

As shown in Figure 5, the kernel density distribution for the entire Songnen Plain shifted rightward, suggesting an overall slow upward trend in resilience. The kernel density curve’s width expanded from the range of 0.2–0.6 in 2001 to 0.5–0.9 in 2021, indicating an increasingly concentrated distribution of resilience levels and reflecting the overall enhancement in resilience. However, differences in resilience levels between regions also intensified, highlighting increasing unevenness and spatial differentiation. Peaks continually shifted to the right, regional differences became more pronounced, and resilience distributions became more uniform.

Specifically, in 2001, the resilience index distribution was relatively concentrated within a lower resilience range (0.3–0.5), indicating weak resistance, adaptability, and reconstruction capacities. By 2011, the region with higher kernel density values expanded toward higher resilience values (0.5–0.7), peaking around 0.6. This improvement signifies that agricultural management practices and policies effectively enhanced the adaptability and reconstruction capabilities of the cultivated land system. In 2021, high-density values further shifted toward the range of 0.6–0.8, although the peak height decreased slightly, suggesting a slowdown in resilience improvement. This slowdown aligns with new challenges and increased environmental pressures associated with recent economic development in the Songnen Plain.

4.2. Spatiotemporal Differentiation of Cultivated Land System Resilience at the County Level

By comparing the dynamic changes in resistance, adaptability, reconstruction, and overall cultivated land system resilience across counties in the study area during the three periods (2001, 2011, and 2021), the resilience levels and characteristics of each county can be clearly identified (Figure 6).

From the perspective of resilience levels, Wuchang City consistently maintained high cultivated land system resilience from 2001 to 2021. Specifically, Wuchang ranked fourth (0.6405) in 2001, first (0.7582) in 2011, and third (0.9389) in 2021, making it the county-level city with the highest overall resilience. This superior resilience primarily stems from its advantageous geographic location, characterized by a mild climate and fertile black soil, which significantly promotes agricultural productivity. Additionally, during the study period, Wuchang City actively integrated agriculture with technological innovation, promoted technological advancements in agriculture, expanded green and organic crop cultivation, implemented black soil conservation practices, encouraged eco-friendly and sustainable circular agriculture, and developed demonstration bases for standardized agricultural production. These modern practices greatly enhanced the city’s ability to resist environmental disturbances, adapt effectively to external changes, and reconstruct its cultivated land system after experiencing disturbances.

In contrast, Ningjiang District exhibited consistently low resilience throughout the study period, ranking at the lowest resilience level in all evaluated years. Ningjiang District is located in the transitional zone between semi-arid and semi-humid climates, marked by unique climatic conditions and pronounced topographical differences caused by neotectonic movements and erosion from the Songhua River. Additionally, due to atmospheric pressure patterns and wind-tunnel effects from the terrain of the Songliao Plain, the region frequently experiences strong southwest winds, resulting in severe soil erosion. These natural factors weaken the district’s resistance, making its cultivated land system particularly vulnerable to external disturbances. Furthermore, the adverse environmental conditions restrict socio-economic development, thereby limiting Ningjiang District’s adaptability to disturbances. Problems such as ecological vulnerability and economic underdevelopment have further impeded the district’s transition toward modern, sustainable agriculture.

From the perspective of different dimensions, cultivated land system resilience across counties in the study area generally exhibited an increasing trend. Among them, Nenjiang City and Harbin urban districts showed the most significant fluctuations, particularly notable improvements in adaptability and reconstruction. With the exceptions of Jiutai District (−0.2292), Ningjiang District (−0.0359), Jiaohe City (−0.0038), and Bei’an City (−0.01569) during 2001–2011, which experienced reductions in resilience primarily due to declining resistance, the resilience of cultivated land systems in other areas showed widespread improvement.

Specifically, resistance in most counties demonstrated an overall upward trend (Figure 7), except for Changchun urban districts, which experienced a decline (−0.0363). The most significant improvements in resistance occurred in Changjiang County, Nianzishan District, and Fulaerji District, while adaptability consistently increased during the 20-year period, with Nenjiang City exhibiting the most pronounced enhancement. Conversely, Changchun urban districts experienced the largest decline in adaptability (−0.1700) (Figure 8). Regarding reconstruction, all counties showed an increasing trend, except for Honggang District, which saw a decline (−0.0816). Harbin urban districts, Nong’an County, and Jiutai District experienced particularly notable increases in reconstruction capacity (Figure 9).

Notably, accelerated urbanization in the Changchun urban area led to gradual degradation in multiple ecological functions of cultivated land, characterized by reduced farmland area, declining soil fertility, and weakened agricultural ecological services, presenting substantial challenges to regional ecological conditions and sustainable agricultural development. Meanwhile, between 2001 and 2021, agricultural subsidies, government-supported projects, and technical training facilitated increased agricultural investment, enhanced production efficiency, the promotion of efficient water-saving irrigation technologies, and agricultural mechanization. These measures, coupled with recent initiatives like black soil conservation projects and high-standard farmland construction, steadily strengthened the resilience of cultivated land systems in the study area. Overall, the region’s cultivated land systems demonstrated progressively enhanced capacity to cope with external disturbances, reflecting substantial achievements in farmland protection and sustainability.

4.3. Analysis of Zoning Results for Cultivated Land System Resilience

Using the MATLAB R2022b software program, K-means clustering was applied to classify cultivated land system resilience across counties in the Songnen Plain. The study area was divided into four zones based on three resilience dimensions (resistance, adaptability, and reconstruction): the core advantage zone, the stable maintenance zone, the transition adjustment zone, and the vulnerable risk zone (Figure 10). The distinct characteristics of these zones highlight significant differences in agricultural production, ecosystem stability, and the ability to respond to external disturbances, providing a scientific basis for the precise management and sustainable development of cultivated land systems in the Songnen Plain (

Table 4. Characteristics of each zone in the study area.

Type	Quantity	Characteristics
Core Advantage Zones	2	High agricultural productivity, ecological stability, diversified economy, and strong risk resilience.
Stable Maintenance Zones	23	Moderate production and ecological recovery capacity, and agriculture dependence, with limited pressure resistance.
Transition Adjustment Zones	20	Traditional production methods, strong resource dependence, and weaker adaptability.
Vulnerable Risk Zones	18	Low productivity, ecological fragility, economic monoculture, and high susceptibility to external shocks.

).

(1)

Core Advantage Zone. This zone exhibits the highest resilience level, primarily ecompassing Wuchang City and Nenjiang City, accounting for a relatively small proportion of the total area. It has advantageous natural resources, well-developed agricultural infrastructure, and advanced modern agricultural technologies, resulting in outstanding agricultural productivity and ecosystem stability. Wuchang City is renowned for its high-quality rice production, supported by abundant water resources, fertile black soil, and advanced agricultural mechanization and technological innovation. Nenjiang City similarly features high agricultural mechanization, high productivity, strong ecosystem recovery capacity, and diversified economic structures that ensure regional economic stability. Due to their strong resource management and adaptability to external changes, these areas serve as core demonstration regions for agricultural and resource management in the Songnen Plain. Policies should therefore consolidate their leading roles in modern agriculture, resource management, and ecological protection.

(2)

Stable Maintenance Zone. This zone is located in the central and northern regions of the Songnen Plain, comprising 23 counties and accounting for 36.51% of the study area. Agricultural production here remains relatively stable, characterized by limited changes in cultivated land area, moderate levels of agricultural mechanization, and medium values of cropping intensity and net primary productivity. Ecosystems in this zone recover gradually after disturbances but less efficiently than those in the core advantage zone. The stable maintenance zone demonstrates moderate adaptability and resistance but has limited resilience against extreme climate events or economic fluctuations. The primary sector dominates local economies, and while rural populations are sufficient, low skill levels and insufficient agricultural innovation limit agricultural productivity improvements. Counties such as Hulan District and Yi’an County exemplify this zone, having solid agricultural foundations but significant potential for enhancing mechanization and agricultural modernization. Policy recommendations include improving agricultural technology, optimizing resource management, and promoting economic diversification to strengthen regional adaptability and stability.

(3)

Transition Adjustment Zone. Primarily distributed in the northeastern fringe and southern areas of the Songnen Plain, this zone includes 20 counties, covering 30.16% of the study area. Agricultural practices remain traditional, with low mechanization, low cropping intensity, and limited ecosystem recovery capacity. Production efficiency is significantly constrained by natural conditions. This region suffers from moderate vegetation cover, fragmented landscapes, weak soil and water conservation, and high vulnerability to climate change and extreme weather events. The rural economy depends heavily on agriculture with a single-industry structure, lacking technological support and innovation. Typical counties include Wudalianchi City and Jiaohe City, characterized by low agricultural productivity, fragile ecosystems, and outdated water and soil management practices. Enhancing productivity and ecological stability in this zone thus requires strengthening water and soil resource management, developing irrigation infrastructure, and introducing modern agricultural technologies to enhance ecosystem restoration capacity.

(4)

Vulnerable Risk Zone. This zone is located in the central–western areas of the Songnen Plain, comprising 18 counties and accounting for 28.57% of the study area. It exhibits low agricultural productivity and weak ecosystem resilience, with significantly lower cropping intensity and production efficiency. Agricultural production is highly dependent on natural conditions, yet the region demonstrates limited resilience against extreme weather events. Declining soil fertility, severe soil erosion, and weak ecosystem recovery capacity exacerbate vulnerability. The rural economy is simplistic, with limited economic activities beyond agriculture, and suffers from severe labor migration and aging populations, further constraining agricultural productivity. Counties like Mingshui County and areas surrounding Harbin illustrate traditional agricultural practices with low mechanization and limited technology adoption, facing ecosystem degradation pressures. Priority policy support is necessary, particularly through ecological restoration, infrastructure development, and agricultural technology dissemination, enhancing the adaptability and stability of agricultural systems.

Overall, the cluster analysis highlights significant regional differentiation in agricultural productivity, ecological resilience, and resource management within the Songnen Plain. The core advantage zone should reinforce its demonstration role in advancing modern agricultural technology and optimized resource management. The stable maintenance zone should focus on technological improvement and economic restructuring, while the transition adjustment and vulnerable risk zones require targeted policy support, technology promotion, and ecological conservation measures to enhance their agricultural and ecological resilience.

5. Discussion

5.1. Spatiotemporal Differentiation and Zoning of Cultivated Land System Resilience

The results indicate that cultivated land system resilience in the Songnen Plain showed an upward trend from 2001 to 2021, accompanied by significant spatial heterogeneity among different regions. Human intervention positively influenced resilience enhancement. Although previous studies specifically on cultivated land system resilience in the Songnen Plain are limited, similar research supports these findings. Song et al. [11] analyzed the spatiotemporal changes and driving mechanisms of ecological vulnerability in the Songnen Plain from 1980 to 2020, demonstrating an overall decreasing vulnerability, indirectly indicating an upward trend in resilience over the past 40 years. Wu et al. [12] observed significant spatial differentiation and agglomeration of agricultural regional functions, characterized by higher resilience in eastern regions and lower resilience in western regions, aligning with the results of this study. Qin et al. [63] reported notable improvements in the Songnen Plain cultivated land system over the past two decades, reflected in improved soil quality, ecosystem service restoration, and increased agricultural productivity, thus indirectly supporting the observed resilience enhancement.

While the cultivated land system resilience in the Songnen Plain has shown an overall upward trend, this improvement is also reflected in the simultaneous enhancement of its three functional dimensions. The increase in resistance can be attributed primarily to rising net primary productivity (NPP), precipitation, and accumulated temperature. Li et al.’s [64] study indicated that from 2001 to 2020, the NPP of cultivated land in the study area demonstrated a consistent growth trend, accompanied by increased precipitation levels. Similarly, research by Lv et al. [65] revealed a rise in active accumulated temperature between 1993 and 2017, with projections suggesting continued growth across most of the region from 2017 to 2027. These climatic improvements lend credibility to the observed enhancement in the resistance dimension of cultivated land resilience. The enhancement of adaptability is closely linked to the increase in per capita cultivated land area and the growth of primary industry output. Statistical data show a negative population growth trend in the region, which has resulted in greater per capita land availability. As a major grain-producing area dominated by crop farming, the Songnen Plain has experienced increases in planting income, which contributed significantly to the growth of the primary industry output. These increases can be attributed to adjustments in cropping structure, increased agricultural inputs, and policy interventions. According to Du et al. [66], driven by economic incentives, the region witnessed an expansion of maize and rice cultivation and a decline in soybean production between 2000 and 2020. This shift in cropping pattern led to higher yields and, consequently, increased farmer income. Additionally, research by Zhang et al. [67] further corroborates that increased inputs, government subsidies, and farmland protection policies have enhanced land use efficiency and planting income in the region. Regarding reconstruction, total agricultural machinery power, vegetation cover, and landscape evenness have played pivotal roles in strengthening system resilience. The rise in total machinery power indicates a transition in agricultural operations from small-scale, labor-intensive practices to large-scale, mechanized production. Meanwhile, ecological spatial indicators such as vegetation cover, landscape evenness, and patch density have all shown positive trends, indicating an enhanced capacity of the cultivated land system to respond to environmental and socio-economic changes. The synergistic enhancement of resistance, adaptability, and reconstruction dimensions has collectively facilitated a regional-scale transformation of cultivated land system resilience in the Songnen Plain over the past two decades. Changes in climatic conditions and cropping regimes have improved land productivity, while the continued implementation of policies such as the “High-Standard Farmland Construction” and the “Black Soil Protection Project” have significantly improved land fertility and infrastructure conditions. Simultaneously, ecological policies like “Returning Farmland to Wetland” and the delineation of “Ecological Protection Red Lines for Wetlands” have contributed to the restoration of ecological functions. These multi-dimensional driving forces—including national farmland protection policies, agricultural modernization investments, and ecological restoration initiatives—have jointly propelled the system from a low-resilience, vulnerable state toward one with a medium-to-high resilience, robust status.

In addition, changes in major crop types provide insights into how adjustments in cropping structure may influence the evolution of cultivated land system resilience. According to the findings of Chen et al. [68] and Ma et al. [66], the maize-based cropping system in the Songnen Plain has undergone significant expansion, while the rice system has remained relatively stable, with a slight increase. In contrast, soybean cultivation has gradually become marginalized [69]. The expansion of maize cultivation has contributed to the growth of primary industry output and increased agricultural mechanization, which in turn positively affects resilience in terms of adaptability and reconstruction. However, the intensification of maize production is often accompanied by higher fertilizer input, which may impose ecological pressure on the system. In comparison, the relatively stable rice areas benefit from sufficient precipitation and strong ecological foundations, with favorable performance in indicators such as vegetation cover and landscape evenness, which support the ecological regulation capacity of the system. On the other hand, crops such as soybeans—characterized by lower economic returns and higher ecological dependency—have declined in proportion, reflecting a trade-off between economic efficiency and ecological function under current agricultural development pathways. Overall, changes in the spatial distribution of cropping zones indirectly contribute to the dynamic evolution of cultivated land system resilience by affecting agricultural output, input structures, and ecological restoration capacity.

The analysis of resilience indicators identified northern and central regions, exemplified by Wuchang City and Nenjiang City, as areas with higher resilience. Similar regional studies support these findings, with Zhang et al. [70] and Lei et al. [71] noting that Wuchang City’s sustained high resilience is primarily attributable to innovative agricultural management systems, enabling efficient utilization of production factors and improved resistance to external disturbances. Dai et al. [72] and Wang et al. [73] found that Nenjiang City’s strong resilience stems from its abundant natural resources, high agricultural mechanization, and effective policy support, enhancing agricultural productivity and ecological restoration capacity. These core advantage zones, such as Wuchang and Nenjiang, provide models for agricultural innovation and resource management that can further promote resilience and sustainability.

In contrast, transitional adjustment zones and vulnerable risk zones, mainly located in peripheral and central–western areas of the Songnen Plain, exhibit lower adaptability and recovery capacities due to constraints from natural conditions (e.g., insufficient rainfall and soil degradation) and socio-economic factors (e.g., inadequate infrastructure and limited technology dissemination).

5.2. Innovations

This study explores the spatiotemporal differentiation and zoning patterns of cultivated land resilience in the Songnen Plain, providing significant insights for sustainable land use, food security assurance, and social development needs. Similarly to previous resilience studies, this research interprets resilience as an interaction among the “resistance–adaptability–reconstruction” dimensions, positing that systems undergo adaptive cycles when faced with disturbances, leading to internal reorganization and upgrading. While international research predominantly focuses on theoretical frameworks, there are few studies explicitly adopting a “resistance–adaptability–reconstruction” perspective. Instead, many emphasize socio-economic, socio-ecological, or structural scales for evaluation. For instance, Huang et al. [74] developed evaluation indicators from natural suitability and socio-economic benefits for land consolidation, while Lv et al. [75] assessed resilience through dimensions such as production, ecology, form, and quality resilience or from perspectives of green agricultural transformation. This study introduces a novel resilience evaluation framework based on “resistance–adaptability–reconstruction”, providing a comprehensive perspective to assess cultivated land system responses to complex environmental and socio-economic challenges, thus informing targeted regional agricultural policies and land management strategies.

This research highlights the complexity and multi-dimensional nature of cultivated land systems by integrating climate conditions, socio-economic contexts, agricultural management capacity, and ecological restoration abilities. Temporal and spatial analyses, exemplified by Wuchang City, reveal regional differentiation and provide targeted recommendations for enhancing resilience in specific zones.

In summary, this study employs resilience theory to develop an innovative cultivated land resilience evaluation framework at the county scale. By adopting the “resistance–adaptability–reconstruction” framework, it enriches the existing literature on cultivated land resilience, facilitating effective regional spatiotemporal analysis and zoning. The research outcomes are crucial for ensuring sustainable land use, food security, adaptive transformation of farmland systems, and sustainable development.

5.3. Policy Recommendations

From the perspective of farmland protection policies and implementation, the pressures on cultivated land systems have been partially alleviated. However, given the region’s unique geographic, climatic, and spatial characteristics, further efforts are required to improve resilience. Based on these factors, this study proposes the following recommendations:

(1)

Enhance agricultural technology, and establish a robust disaster early warning system. Given the temperate continental monsoon climate of the Songnen Plain, characterized by warm, humid summers and cold, dry winters, it is essential to select and breed crop varieties that are drought-tolerant, cold-resistant, and pest-resistant, all suited to local climatic conditions and geared toward improving crop resilience. Additionally, adopting precision agriculture technologies—such as drone-based remote sensing, smart irrigation systems, and precise fertilization methods—can further strengthen agricultural resilience. This study found that natural factors, including precipitation and sunlight duration, influence farmland system resilience, so improving disaster early warning systems and enhancing meteorological monitoring to issue timely warnings will therefore help farmers prepare better and mitigate potential income losses from natural disasters. Furthermore, providing subsidized agricultural insurance can encourage greater participation from farmers. In core advantage regions of the southern and central areas (e.g., Wuchang City and Shuangcheng District), government-funded smart agriculture initiatives can introduce technologies like drone monitoring, soil sensors, and precision irrigation for more targeted management. For example, installing soil moisture sensors can enable real-time monitoring for precise irrigation and fertilization. As these technologies evolve, their expansion into areas with moderate adaptability can serve as a model to enhance production efficiency and resource conservation.

(2)

Strengthen land management and protection; complement ecological conservation and restoration. Preventing the conversion of farmland into non-agricultural or non-grain uses is crucial. Strict enforcement of farmland protection policies can curb illegal occupation and excessive development, thereby stabilizing farmland areas. Equally important is preserving the natural environment of the Songnen Plain to maintain ecological balance and prevent over-exploitation or ecological degradation. As a major production base for crops such as corn, soybeans, and rice, the black soil region of the Songnen Plain has increasingly been prioritized by China’s farmland protection initiatives in recent years. With programs such as high-standard farmland construction, land consolidation, and ecological restoration, the resilience of farmland systems has improved. As China transitions to high-quality development, refining farmland management mechanisms becomes even more critical. Ensuring no net reduction in farmland area, guaranteeing quality and quantity, optimizing land use structures, and enhancing land use efficiency are all necessary measures. These actions will not only promote economic development and sustainable agriculture but also protect ecosystems and reinforce the Songnen Plain’s role as a critical grain production base for China.

(3)

Adopt region-specific strategies based on zoning outcomes. Based on the zoning results, distinct strategies should be implemented for each zone according to its specific characteristics. In core advantage zones, focus should be placed on high-efficiency agriculture, the promotion of precision farming technologies, and green agricultural development, while establishing demonstration bases for the coordinated advancement of agriculture and ecology. In stable maintenance zones, optimizing crop structures, diversifying operational models, enhancing soil and water conservation, and improving soil quality through policy support and technology adoption can increase regional adaptability. In low-resilience zones, priorities should include soil improvement and ecological restoration projects, optimizing irrigation and drainage systems, developing climate-resilient specialty crops, and minimizing overdevelopment. In ecologically sensitive zones, strict ecological protection lines should be established, promoting ecological farming practices and intensifying efforts to restore wetlands and forests, ensuring the recovery and stability of regional ecosystems. Ultimately, these zoned management practices will support the sustainable use of farmland resources while promoting coordinated regional economic development.

5.4. Limitations and Outlook

5.4.1. Study Limitations

Despite the relatively comprehensive analytical framework constructed in this study, several limitations still warrant acknowledgment: First, this study adopts county-level administrative units as the basic scale of analysis, which facilitates overall comparative analysis at the regional scale. However, due to the relatively large spatial scale, it may obscure fine-grained spatial heterogeneity within cultivated land systems at the plot, village, or township levels. This limitation may lead to insufficient capture of important characteristics such as localized variations in farmland resilience, thereby affecting the precision of spatial differentiation results. Second, although this study integrates multiple datasets, due to the vast spatial extent of the study area and the relatively small size of research units, certain key socio-economic and ecological indicators (e.g., rural infrastructure levels and soil quality indicators) were unavailable during the data collection process, thereby limiting the depth and comprehensiveness of the analysis. Third, the selected study period (2001–2021) facilitates the understanding of historical trends in farmland resilience. However, with the intensification of global climate change, the acceleration of agricultural restructuring, and dramatic changes in rural socio-economic environments, it is difficult to comprehensively predict the evolution of farmland system resilience under different future scenarios based solely on past data, thus restricting reliable future trajectory predictions. Fourth, the four-zone classification method adopted in this study effectively reflects the spatial differences in farmland resilience across the Songnen Plain. However, compared to the internationally adopted hierarchical protection models, such as the UNESCO Biosphere Reserve zoning framework (core areas, buffer zones, and transition zones) [76], it remains relatively coarse. This discrepancy may limit the applicability and scalability of the study results in multi-scale land management practices.

5.4.2. Outlook and Future Work

Moving forward, advancing the research on farmland system resilience requires interdisciplinary collaboration across fields such as ecology, socio-economics, and agricultural sciences. The introduction of higher-resolution spatial data will contribute to more precise identification of local differences in farmland resilience, thereby improving the granularity and accuracy of spatial analysis. Future research should also focus on the spatiotemporal evolution of specific cropping systems and their dynamic coupling with resilience mechanisms. In addition, the establishment of long-term monitoring networks is crucial for systematically tracking changes in farmland resilience and evaluating the actual effectiveness of technological innovations and policy interventions. Finally, continuous evaluation and refinement of farmland policy mechanisms will be critical to further improving the resilience, sustainability, and productivity of farmland systems. Through the construction of a comprehensive and integrated research framework, the Songnen Plain is expected to achieve sustainable agricultural development while contributing to national food security and ecological integrity.

6. Conclusions

Drawing upon resilience theory, this study evaluates the cultivated land system resilience in the Songnen Plain by developing an indicator system for resilience assessment. Utilizing multi-source data—including remote sensing, meteorological records, land use information, and socio-economic statistics—and employing three-dimensional Euclidean distance, GIS spatial analysis, and other methods, we quantitatively characterize the spatiotemporal evolution of the cultivated land system in the Songnen Plain over three time points (2001, 2011, and 2021). We further reveal how both natural and anthropogenic factors contribute to spatial heterogeneity in cultivated land resilience. The key findings are as follows:

(1)

Overall Upward Trend in Resilience

Between 2001 and 2021, the cultivated land system resilience of the Songnen Plain showed a clear upward trend. This improvement is closely related to agricultural technological advancements and the optimization of land use structures, which collectively enhance the system’s resistance, adaptability, and reconstruction capacity. Spatially, resilience was higher in the eastern regions and relatively lower in the western and central regions. Areas such as Wuchang City and Nenjiang City consistently maintained higher resilience over the past two decades, whereas places like Ningjiang District and Jiaohe City saw comparatively modest improvements. Additionally, the finding that improvements in resistance contribute most to resilience enhancement—followed by adaptability and reconstruction—offers quantitative guidance for balancing agricultural productivity with ecological sustainability. By establishing a three-dimensional resilience framework that integrates resistance, adaptability, and reconstruction, we provide a unified basis for identifying vulnerability and assessing recovery capacity under complex external pressures.

(2)

Kernel Density Analysis

Kernel density estimation effectively mitigates noise in the data and reveals authentic distribution trends, identifying hotspots and distribution patterns of cultivated land system resilience. The results indicate that resilience values have grown increasingly concentrated over time, shifting steadily to the right on the value axis. Peak values initially increased and then decreased slightly, while the differences among regions have gradually intensified. Additionally, it is important to note that a Kernel Density Analysis (KDA), while effective in identifying the spatial distribution trends of resilience, does not independently explain the underlying mechanisms driving resilience changes. Therefore, this study complemented KDA with three-dimensional Euclidean distance measurements and K-means clustering analyses to provide a more comprehensive interpretation of cultivated land system resilience dynamics.

(3)

Zoning Results

According to the clustering-based zoning, transition adjustment areas and vulnerable risk areas face substantial ecological and economic challenges, characterized by lower adaptability and resource-use efficiency. These areas urgently require policy support, resource allocation, and technological intervention to bolster resilience. Stable maintenance areas display a certain capacity for adaptation and resistance to external pressures, albeit with a limited ability to respond to abrupt environmental changes. In comparison, core advantage areas exhibit strong agricultural productivity and resource management capabilities, effectively coping with external changes while sustaining ecosystem health and productivity. The zoning approach thus provides a powerful tool for evaluating farmland resilience and yields operational tools for differentiated land consolidation and ecological restoration. It enables more precise, region-specific strategies for farmland management in the Songnen Plain, thereby supporting the achievement of sustainable development goals.

In summary, to address the increasing vulnerability of cultivated land systems under climate change, land degradation, and socio-economic pressures, this study proposes targeted solutions through the construction of a comprehensive resilience evaluation framework. By quantifying resilience characteristics across resistance, adaptability, and reconstruction dimensions, revealing spatial heterogeneity using three-dimensional Euclidean distance, and implementing regional zoning through K-means clustering, this study provides practical tools for enhancing farmland stability, promoting adaptive agricultural practices, and guiding ecological restoration efforts. These findings offer a scientific basis for formulating integrated, multi-dimensional land management strategies, thereby facilitating the sustainable development of agricultural production and ecosystem services under dynamic environmental conditions.