Rural Resilience Evaluation and Risk Governance in the Middle Reaches of the Heihe River, Northwest China: An Empirical Analysis from Ganzhou District, a Typical Irrigated Agricultural Area

Abstract

Conducting research on the evaluation of rural resilience and risk governance strategies in the middle reaches of the Heihe River can provide a scientific basis for the sustainable development of rural areas in the inland river basins of arid regions. Affected by water resource constraints, the expansion of artificial oases, and excessive exploitation of groundwater, the rural areas in the middle reaches of the Heihe River Basin, the second largest inland river in the arid region of northwest China, are confronted with prominent contradictions in the human-land relationship and urgently need to enhance their ability to cope with risks. Based on the remote sensing data of land use and major socio-economic data, this study draws on the theory of landscape ecology to construct a disturbance-resistance-adaptability evaluation system. Taking Ganzhou District, a typical irrigated agricultural area, as a case study, the study uses the entropy weight method, resilience change rate, and obstacle degree model to analyze the rural resilience level and its changing characteristics from 1990 to 2020, identifies the key obstacle factors affecting the development of rural resilience, and proposes risk governance strategies accordingly. Main conclusions: (1) The overall rural resilience index is relatively low, showing significant spatial disparities. Towns with well-developed multifunctional agriculture, nature reserves, and ecological-cultural control lines have higher resilience indices. (2) The change rate of the rural resilience index demonstrates phase heterogeneity, generally undergoing a “relative stability-increase-decrease” process, and forming a differentiation pattern of “decrease in the north and increase in the south”. (3) Internal risks to rural resilience development in the Ganzhou District mainly stem from low economic efficiency, fragile ecological environment, and unstable landscape patterns, among which efficiency-dominant and landscape-stability obstacle factors have a broader impact scope, while habitat resistance-type obstacle factors are mainly concentrated in the western part and suburban areas. Enhancing the benefits of water and soil resource utilization, strengthening habitat resistance, and stabilizing landscape patterns are key strategies for current-stage rural resilience governance in the middle reaches of the Heihe River. This study aims to optimize the human-land relationship in the rural areas of the middle reaches of the Heihe River.

1. Introduction

As a typical human-land composite system, the development of rural areas has always been influenced by changes in both internal and external environments. In the face of various disturbances and impacts, why do some rural systems decline or even disappear, while others manage to develop continuously? The most important reason lies in the different levels of resilience of rural areas in adapting to and coping with environmental changes [1]. Resilience refers to the response and coping ability demonstrated by a system when facing risks and challenges. The theory regards the countryside as a dynamic social-ecological system, which not only emphasizes the system’s ability to withstand shocks in the face of internal and external disturbances, but also attaches great importance to the system’s ability to cope with changes through self-adjustment, learning, and innovation [2]. In such a complex and changeable environment, the research on rural territorial systems guided by the resilience theory breaks through the limitations of traditional static analysis thinking and provides a more adaptable methodological system, opening up new ideas for exploring and practicing rural sustainable development [3]. Given the important role of the resilience concept in rural sustainable development, rural resilience has become an emerging hot topic and frontier research area in the current academic community.

Scholars mainly conduct research on the evolution and connotations of rural resilience, assessment frameworks, measurement systems, influencing factors, and enhancement pathways. The specific contents are as follows: Firstly, the evolution and connotations of rural resilience. The concept of resilience, which has been translated as elasticity or recovery force by some scholars, originated from material mechanics, referring to the ability of a material or organism to return to its initial state after being subjected to pressure [4]. In 1973, Holling appropriated this concept to gauge the capacity of ecosystems to withstand changes and disturbances while preserving the continuity of their populations [5]. Subsequently, Walker et al. extended its application to social-ecological systems, highlighting its self-organizing, learning, and adaptive capacities in the face of stress [6]. As an active way and strategy of exploration, adaptation, and transformation, resilience research has received extensive attention in the field of rural sustainable development [7]. With the expansion of its application, the conceptual connotation of rural resilience has been continuously developed. It has evolved from the initial ability of a rural area to withstand disturbances and maintain its structure and function without undergoing qualitative changes [8], to the ability of a rural area to generate new factors to adapt to disturbances and thus achieve sustainable development [9], and then to the ability of a rural area to resist risks, make adaptive adjustments, and innovate and transform when facing disturbances [2]. Overall, rural resilience refers to the ability of the core system of a rural area to resist, absorb, adapt to, and transform to the greatest extent possible, avoiding unstable operation or even collapse of the system, and thus promoting the renewal, transformation, and entry of the system into a more stable and advanced state [10]. Secondly, the assessment frameworks and measurement systems. Currently, there is no unified evaluation paradigm for rural resilience assessment, which mainly includes the following aspects: Constructing an index system from the perspective of element composition, such as from the dimensions of rural industry, economy, society, ecology, culture, etc. [11,12,13] or from a comprehensive perspective of the human settlement environment territorial system [14,15,16] to measure rural resilience; In response to specific disturbances such as climate change [17,18,19], natural disasters [20,21], food security [22], and land use policies [23], obtaining data through case surveys from the perspectives of emergency management and community resources, and constructing an evaluation framework for rural resilience under specific pressure disturbances; In view of the continuous adaptation, learning, and transformation process demonstrated by the rural territorial system when facing external disturbances, an adaptive resilience framework based on grounded theory and adaptive cycles has been formed [24,25], as well as response resilience frameworks such as the models of pressure-state-response, resistance-adaptive-renewal, element support-structural resistance-functional adaptive [26,27,28], and community/household livelihood resilience analysis frameworks such as buffer-self-organization-learning capacity [29,30,31]. Commonly used evaluation methods for rural resilience include the structural dynamics and system dynamics method, comprehensive index evaluation method, function model method, scenario analysis method, etc. [32]. Finally, the influencing factors and enhancement pathways. The evolution of rural resilience in China is the result of the joint driving of factors such as economic, social, human, and ecological capital [33,34], policy, information, and the degree of locational advantage [35,36], community cohesion/sense of belonging [37], and the participation rate of rural households [38,39]. In developed countries, with strong government leadership, the ability of rural areas to respond to disturbances is comprehensively strengthened. They mainly enhance rural resilience through such paths as rural digital technology [40] and creative industries [41], and promote the long-term sustainable operation of rural areas through the diversified development of rural areas [42] and the trend of globalization [43]. In summary, the research on rural resilience has formed a relatively complete research system from theory to empirical evidence, and it is characterized by diverse evaluation frameworks and measurement methods, as well as a wide range of research scopes and scales. Although there are numerous achievements in the study of rural resilience, most of them are based on macro-level statistical data at the county level and above or micro-level survey data of communities and rural households. The research on the long-term time-series resilience evolution of towns and villages at the micro level is relatively weak.

The Heihe River is the second-largest inland river in the arid region of Northwest China. Starting from its source in the Qilian Mountains of Gansu Province and flowing to the Juyan Lake, its terminal lake in Inner Mongolia, it is divided into the upper, middle, and lower reaches with Yingluoxia and Zhengyixia as the boundaries. The upper reaches are the main runoff-producing area of the entire basin, the middle reaches have well-developed irrigated agriculture, and the lower reaches are in an extremely arid area [44]. Due to the disharmony between water and soil resources in the Heihe River Basin, there has always been a water use conflict in the middle and lower reaches. In order to alleviate this situation, in 1997, the State Planning Commission approved the water allocation plan for the Heihe River under different wet and dry year conditions, which was officially implemented in 2000 [45]. Although the water diversion plan has alleviated the water use conflict between the middle and lower reaches, the middle reaches concentrate 95% of the cultivated land, 91% of the population, and 89% of the gross domestic product of the basin. With the reduction of available water resources and the continuous expansion of artificial oases, agricultural irrigation water cannot be guaranteed, and it can only be maintained by over-exploiting groundwater, which has triggered new problems such as the decline of the groundwater level, soil degradation, and the intensification of desertification [46]. The continuous development of the unreasonable water use structure will endanger regional ecological stability and food security, posing a serious threat to the sustainable development of agriculture and rural areas.

The irrigated agriculture in the inland river basins of arid regions is highly dependent on external water conveyance, and it has the characteristics of being fragile and variable. It is extremely sensitive to the disturbances of environmental changes and human activities. Under the disturbance of multiple factors, it is very likely to experience phenomena such as system degradation and even disappearance [47]. As a typical representative of the inland rivers in the arid region of Northwest China, the Heihe River has prominent contradictions in the human-land relationship in its middle reaches under the interference of multiple factors, and the sustainable development of rural areas is seriously threatened. There is an urgent need to improve the resilience of rural areas to cope with risks. Therefore, taking Ganzhou District as an example, this study explores the rural resilience level from 1990 to 2020 and identifies the main obstacle factors affecting the development of resilience. The specific contributions are as follows: Firstly, a rural resilience evaluation system is constructed based on the concept of landscape ecology. Secondly, the changing characteristics of rural resilience in Ganzhou District from 1990 to 2020 are analyzed from the grid and village-town scales. Finally, the key obstacle factors affecting the development of rural resilience are identified and spatially clustered. Most importantly, risk governance strategies suitable for the development of rural resilience in the middle reaches of the Heihe River at present are proposed.

2. Material and Methods

2.1. Study Area

The middle reaches of the Heihe River encompass specific districts and counties within Zhangye City, namely Ganzhou District, Linze County, Gaotai County, Shandan County, and Minle County. The case study area, Ganzhou District (38°32′~39°24′ N, 100°06′~100°52′ E), is characterized by an elevation that varies from 1412 to 3591 m. Spanning 65 km from east to west and 98 km from north to south, Ganzhou District covers a total area of 3661 km². The district’s geography is defined by the Qilian Mountains to the south, the Heli and Longshou Mountains to the north, and the Hexi Corridor Plain in between, with a topographical feature of higher elevations in the north and south and a lower central region (Figure 1). Based on the seventh national population census data, in 2020, Ganzhou District had a permanent population of 519,100, with 241,000 being rural residents, representing 46.43% of the total. The district administers 18 townships and 245 administrative villages.

Ganzhou District has been selected as a case study locale for a multitude of compelling reasons: Firstly, the region is characterized by its scant rainfall, arid climate, and delicate ecological system. The water sources are predominantly sustained through a synergistic blend of atmospheric precipitation, snowmelt from the Qilian Mountains, and underground reserves, with the Heihe River’s water distribution exerting a significant influence on the area’s development. Secondly, the district’s rural expanses boast a robust irrigation agriculture sector, marked by an ongoing expansion of arable land. Agricultural water consumption has consistently accounted for approximately 90% of the total water usage, with the excessive demand leading to the critical overexploitation of groundwater resources. Thirdly, in response to these challenges, the government has introduced a suite of measures, including agricultural restructuring, water pricing reforms, management of overexploited groundwater zones, and ecological conservation initiatives. Yet, the oasis’s inherent fragility and susceptibility to change, alongside the sluggish shift from extensive to water-conserving agricultural practices, have positioned the rural areas at the crossroads of resource and environmental strain and underwhelming socio-economic progress. There is an urgent need to recalibrate the human-land dynamic and bolster the system’s resilience to risks; failure to do so could result in a pernicious cycle of maladaptation within the socio-ecological framework. In sum, Ganzhou District encapsulates the prevalent challenges confronting agricultural and rural sectors in the inland river basins of China’s arid northwest, making it an exemplary locale for examining the dynamics of rural resilience amidst multiple stressors. Furthermore, given its pronounced vulnerability to human activities, such as the proliferation of artificial oases in the Heihe River Basin and the intensive exploitation of groundwater, it serves as an ideal setting for delving into internal risk governance and the formulation of adaptive strategies.

2.2. Research Framework

In the arid northwest of China, water resources, despite their scarcity, are pivotal to socio-economic progress. However, the challenge of acquiring long-term sequential data on village water resources and socio-economics necessitates the development of a rural resilience evolution assessment framework that simplifies indicators while capturing the unique regional traits of the Heihe River’s middle reaches. In these arid zones, water-scarce irrigated agriculture predominates, with the availability of water resources dictating the patterns of agricultural development and the dynamics of land use. Consequently, the configuration of land use patterns can serve as an indicator of water resource management, encompassing aspects such as distribution, utilization efficiency, and the adaptive response of water resources to shifts in land use [48]. Land resources, as the foundation of agricultural activities in arid irrigated regions, are subject to continuous perturbations from socio-economic human endeavors. The resilience of rural areas is primarily manifested in the capacity of the habitat system to withstand disturbances under the pressures of socio-economic development and in the stability of land use patterns.

Hence, this research focuses on the quintessential irrigated agricultural region of Ganzhou District within the middle reaches of the Heihe River as a case study. Utilizing land use data and key socio-economic indicators, the study establishes a rural resilience evolution assessment system within the context of the disturbance-resistance-adaptability (DRA) framework (Figure 2). Within this framework, ‘disturbance’ pertains to the extent of disruption caused by socio-economic development, ‘resistance’ denotes the capacity of the habitat system to withstand such disruptions, and ‘adaptability’ refers to the stability attributes of the functional structure and form of the land use landscape pattern amidst socio-economic developmental pressures. Acknowledging the significant impact of the 2000 water allocation plan on regional development in the middle reaches of the Heihe River, the study delineates the period from 1990 to 2020 to capture the transformations resulting from the water allocation plan. By evaluating the characteristics of rural resilience, identifying internal risk factors, and formulating adaptive strategies through a classification and zoning approach, the study endeavors to enhance the human-land relationship within the rural areas of the Heihe River.

2.3. Indicator Construction and Data Processing

2.3.1. Construction of Measurement Indicators

Drawing on relevant studies in landscape ecology [49], a rural resilience evaluation indicator system based on disturbance, resistance, and adaptability is constructed from three aspects: the impact of socio-economic development disturbances, the resilience of the habitat system, and the stability of land use patterns encompassing a total of 14 indicators. As depicted in

Table 1. Rural resilience evaluation indicator system based on disturbance-resistance-adaptability.

Target Layer	Criterion Layer		Indicator Layer		Weighted Mean
	Indicator	Effects	Specific Indicators	Effects
Disturbance	Socio-economic Development	−	X1-Population Density (people/km2)	+	0.010
			X2-Per Capita Grain Output (kg/person)	+	0.186
			X3-Land Income per Unit Area (yuan/km2)	−	0.519
			X4-Per Capita Income (yuan/person)	−	0.285
Resistance	Ecosystem Resistance to Disturbances	+	X5-Elevation (m)	−	0.123
			X6-Slope (°)	−	0.130
			X7-Normalized Difference Vegetation Index (NDVI)	+	0.085
			X8-Distance to Ecological Protection Areas (m)	−	0.330
			X9-Distance to Water Bodies (m)	−	0.184
			X10-Distance to Roads (m)	+	0.149
Adaptability	Stability of Land Use Landscape Pattern	+	X11-Shannon Diversity Index (SHDI)	+	0.267
			X12-Largest Patch Index (LPI)	+	0.173
			X13-Landscape Shape Index (LSI)	−	0.224
			X14-Landscape Contagion Index (CONTAG)	+	0.335

Note: The “+” indicates a positive indicator, that is, the larger the value, the better the situation; the “−“ indicates a negative indicator, that is, the larger the value, the worse the situation.

: (1) Disturbance refers to the magnitude of disruption stemming from socio-economic development, shaped by the interplay of population dynamics, agricultural practices, and economic growth. Specifically, population density and per capita grain output, land income, and per capita income are used to reflect the degree of human activity disturbance that the water and soil resources of the arid area endure. The higher the population density and per capita grain output, the greater the impact of human activities, leading to greater resource and environmental pressure. Economic benefits are the driving force and support for sustainable ecological development, and land income and per capita income are used to reflect this. The higher the benefits, the more they contribute to the high-quality development of the ecosystem, and the less pressure is generated. (2) Resistance mainly reflects the resistance capacity of the habitat system, represented by elevation, slope, normalized vegetation index, and distances to ecological protection areas, water bodies, and roads. Where: the worse the topographical conditions, i.e., the higher the elevation and the greater the slope, the more susceptible the habitat system is to natural disasters and other threats, and the worse the resistance capacity; the higher the vegetation coverage, the stronger the resistance capacity of the habitat system; the closer to ecological protection areas, the less likely to be disturbed, and the stronger the resistance capacity of the habitat system; for arid areas, the closer to water bodies, the stronger the resistance capacity of the habitat system; the farther from roads, the less likely to be disturbed by the outside world, and the stronger the resistance capacity of habitat quality. (3) Adaptability is mainly expressed by the stability of the land use landscape pattern, which comprehensively reflects the changes in water resources. The Shannon Diversity Index (SHDI) is selected to represent landscape richness, the Largest Patch Index (LPI) to reflect landscape dominance, the Landscape Shape Index (LSI) to reflect the degree of regularity and fragmentation of the landscape after disturbance, and the Landscape Contagion Index (CONTAG) to reflect the connectivity of the landscape in space. Where: the larger the SHDI, LPI, and CONTAG, the more uniform and stable they indicate; while the larger the LSI, the higher the fragmentation and the worse the stability.

2.3.2. Data Sources and Processing

The 30 m land use data for Ganzhou District from 1990 to 2020 were obtained from the Resource and Environment Science Data Center, Chinese Academy of Sciences. Vector data for village boundaries, ecological zones, roads, and water systems were sourced from Ganzhou District’s Natural Resources Bureau. Elevation and NDVI data were acquired from the Geospatial Data Cloud, while socio-economic data were extracted from the “Ganzhou District Statistical Yearbook” and “Ganzhou Agricultural Statistical Report”.

The indicator data processing steps are as follows: (1) Socio-economic data for townships are integrated with village boundary attribute tables in GIS10.8, followed by empirical Bayesian kriging to create grid datasets aligned with land use data. (2) Elevation is derived through map sheet assembly, masking, projection, and resampling, while slope is extracted from DEM analysis. (3) The distances to ecological protection areas, roads, and water bodies are calculated using the Euclidean distance analysis tool in GIS10.8, and different levels of transportation lines (railways, national highways, provincial highways) and water systems (perennial rivers, seasonal rivers) are obtained through weighted summation. (4) Landscape pattern indices (SHDI, LPI, LSI, CONTAG) are analyzed using Fragstats3.3 [49] and resampled to match the scale of Ganzhou District and village boundaries [50], resulting in a raster map of these indices for the study area.

2.4. Research Methodology

2.4.1. Entropy Weighting Method

The entropy weight method is an objective weighting method based on the degree of data dispersion. It measures the degree of variation of data by calculating the information entropy of indicators and is often applied to the comprehensive evaluation of complex systems with multiple indicators [51].

Firstly, to eliminate the impact of the dimensions of the original data of the evaluation index system on the evaluation, the original data is standardized using the extreme value method. The specific process is as follows:

$$\mathrm{For} \mathrm{indicators} \mathrm{with} \mathrm{positive} \mathrm{effects}: x_{ij}^{\prime }={\displaystyle \frac{x_{ij}-\min (x_{ij})}{\max (x_{ij})-\min (x_{ij})}}$$

(1)

$$\mathrm{For} \mathrm{indicators} \mathrm{with} \mathrm{negative} \mathrm{effects}: x_{ij}^{\prime }={\displaystyle \frac{\max (x_{ij})-x_{ij}}{\max (x_{ij})-\min (x_{ij})}}$$

(2)

In the formula, $x_{ij}^{\prime }$ represents the standardized value of the indicator; $x_{ij}$ represents the original value of the indicator; $\max (x_{ij})$ and $\min (x_{ij})$ represents the maximum and minimum values of the indicator, respectively.

Secondly, the entropy weight method is used to calculate the weight values of the indicators. The specific process is as follows: Theorem-type environments (including propositions, lemmas, corollaries, etc.) can be formatted as follows:

$$\mathrm{Indicator} \mathrm{information} \mathrm{entropy}: H_j=-{\displaystyle \frac{1}{\ln m}}{\displaystyle \sum _{i=1}^m(p_{ij}\ln p_{ij})}$$

(3)

$$\mathrm{Indicator} \mathrm{weights}: w_j=\left(1-H_j\right)/{\displaystyle \sum _{j=1}^n(1-H_j)}$$

(4)

The formula, $H_j$ represents the information entropy value; $p_{ij}$ represents the proportion of indicator i in the j-th indicator; $w_j$ represents the weight of the j-th indicator.

Then, calculate the grid-scale disturbance, resistance, and adaptability indices. The specifics are as follows:

$$D_i={\displaystyle \sum _{j=1}^n{x}_{ij}^{\prime }}\times w_j$$

(5)

In the formula, $D_i$ is the disturbance/resistance/adaptability index value for grid i; $x_{ij}^{\prime }$ is the standardized value of indicator j for grid i; $w_j$ is the weight value of indicator j derived from the entropy weight method; and n is the total number of indicators.

Finally, calculate the rural resilience index R. After calculating the disturbance, resistance, and adaptability indices at the grid scale, the resilience value for each grid is obtained by summing them up with equal weights (1/3 for each). Then, the average value within the administrative boundaries of the township is extracted to represent the rural resilience index R.

2.4.2. Index Change Rate

To gain a clearer understanding of the spatial heterogeneity of grid and rural resilience changes, a model for calculating the change rate of grid/rural resilience indices at different stages is constructed [46]. As follows:

$$R^{\prime }=\left[\left(R_{t2}-R_{t1}\right)/R_{t1}\right]\times \left[1/\left(t_2-t_1\right)\right]\times 100\%$$

(6)

In the formula, $R_{t1}$ and $R_{t2}$ represent the grid/rural resilience indices for periods t₁ and t₂ respectively; $R^{\prime }$ represents the change rate of the grid/rural resilience index from period t₁ to t₂.

2.4.3. Barrier Degree Model

The Barrier Degree Model can quantify the barriers of various indicators of rural resilience, providing an analytical tool for identifying the main factors affecting resilience. To further analyze the obstacles to resilience development, this paper introduces the Barrier Degree Model for the identification of specific influencing factors.

The Barrier Degree Model is based on three indices: factor contribution, indicator deviation, and barrier degree [52]. The larger the barrier degree, the greater the hindrance the indicator poses to the development of rural resilience. The calculation formula is as follows:

$$V_j=1-X_j$$

(7)

$$M_j=(W_j\times V_j)/{\displaystyle \sum (W_j\times V_j)}\times 100\%$$

(8)

In the formula, $V_j$ represents the indicator deviation, which indicates the gap between a single indicator and the optimal target value; $X_j$ is the standardized value of a single indicator; $M_j$ represents the barrier degree, which indicates the extent to which a single indicator hinders rural resilience; $W_j$ represents the factor contribution, which is the weight of a single indicator.

3. Results

3.1. Rural Resilience Spatiotemporal Evolution Characteristics

3.1.1. Disturbance-Resistance-Adaptability Pattern

Using the entropy weighting method, the social and economic development disturbance, habitat system resistance, and land use landscape adaptability index values at the grid scale for Ganzhou District in the years 1990, 2000, 2010, and 2020 were calculated. Figure 3 illustrates the following patterns: (1) The disturbance index peaks in densely populated urban fringe areas and regions with high per capita grain yield in the central plains. Southern and northern areas, such as Pingshanhu Township, exhibit lower disturbance indices, presenting a “high central, low peripheral” spatial distribution. (2) The resistance index increases from southwest to northeast. Nature reserves and areas under ecological and cultural protection display the highest resistance to habitat system disturbances, followed by regions with dense water networks. Conversely, the southwestern mountainous areas, limited by topography, show the weakest resistance. (3) The adaptability index is spatially scattered, creating a complex and heterogeneous mosaic. Areas aligned with water systems have higher adaptability indices, while those with land use fragmentation, irregular patch shapes, small patches, and poor connectivity have lower indices.

3.1.2. Spatio-Temporal Pattern of Rural Resilience Index

Using the equal-weight summation method, the resilience index at the grid scale for Ganzhou District was obtained for the years 1990, 2000, 2010, and 2020, and then the GIS10.8 grading visualization was applied. Subsequently, the average value within the administrative boundaries of the township was extracted to represent the town and village resilience index, which was used to analyze the spatio-temporal pattern of the rural resilience index (Figure 4). Overall, the average rural resilience index for Ganzhou District in the years 1990, 2000, 2010, and 2020 were 0.4398, 0.4403, 0.4531, and 0.4401, respectively, indicating a pattern of initial increase followed by a decrease, with a small range of variation and an overall low level of resilience index. Spatially, regions with low socio-economic stress, diverse landscapes, and rich biodiversity, particularly those with nature reserves, exhibit higher resilience. Conversely, areas with high population density, intensive agriculture, significant topographical constraints, and inflexible land use patterns show lower resilience. There is a distinct east-west gradient in resilience, with the east showing higher indices than the west. At the township level, those with advanced multifunctional agriculture in suburban regions, like Liangjiadun and Shangqin, have maintained a high resilience index over time. Townships with nature reserves and strong leisure agriculture, such as Pingshanhu, Dangzhai, Wujiang, Sanzha, and Jiantan, also show high resilience. Next are those with significant state-owned forestry and agriculture farms, including Xindun and Chang’an Town. Conversely, townships along the Heihe River’s main stream, despite their water wealth, have lower resilience due to the lingering effects of intensive human activities, as seen in Jing’an, Mingyong, Shajing, Xiaoman, and Daman. The western mountainous areas, particularly the Ganjun-Longqu-Huazhai-Anyang region, exhibit the lowest resilience, reflecting their vulnerability to disturbances.

3.1.3. Spatio-Temporal Change Rate of Rural Resilience Index

To better understand the spatial heterogeneity of changes in the rural resilience index in Ganzhou District, the change rate of the grid-scale resilience index was calculated in stages and the changes in each township were statistically analyzed, with positive values indicating an increase and negative values indicating a decrease. As depicted in Figure 5, the rural resilience index in Ganzhou District underwent distinct changes. (1) From 1990 to 2000, the rural resilience index declined in areas near urban centers and the northern edge due to persistent arid climate disturbances. Yet, it rose in townships like Dangzhai, Anyang, and Ganjun due to new arable land. Notably, Wujiang and Jing’an townships, with their proximity to the Heihe River’s main stream and extensive paddy fields, saw a significant increase in resilience. (2) From 2000 to 2010, the Heihe River Basin’s water allocation plan rejuvenated water resources in peripheral regions, fostering a swift growth in arable land and the scale of agricultural and livestock production, thereby generally boosting the rural resilience index. However, some areas, such as Wujiang Town, saw a resilience decline due to the expansion of construction and living spaces, which compromised the ecological environment and led to a notable reduction in farmland, thus decreasing the resilience index. (3) From 2010 to 2020, the relentless expansion of arable land and agricultural production spaces led to overconsumption of water resources, with groundwater being exploited at the cost of ecological water, resulting in serious ecological and landscape degradation. Concurrently, the intense development pressures from homestead land, facility agriculture, rural revitalization zones, and rural tourism contributed to a general decline in the rural resilience index. (4) From 1990 to 2020, Ganzhou District’s rural resilience index exhibited a distinct north-south gradient, with a decline in the north and an increase in the south. Suburban and northern areas, characterized by high human activity and forest, grass, and water ecosystems, showed greater vulnerability and a decrease in resilience. In contrast, the less disturbed southern periphery, bolstered by water resource replenishment, saw a rise in resilience due to diversified agricultural activities, albeit with a modest amplitude. On average, the rural resilience index in Ganzhou District saw minimal changes of 0.019% from 1990 to 2000, a slight increase of 0.318% from 2000 to 2010, and a decline of 0.289% from 2010 to 2020. Amidst persistent arid conditions, water resource management, and significant developmental pressures, the district’s rural resilience has traced a pattern of initial stability, followed by an uptick, and then a downturn, suggesting that rural resilience development is now at an impasse.

3.2. Identification of Internal Risks in Resilience Development

3.2.1. Average Obstacle Degree of Indicator Factors

To identify the main risk factors affecting the development of rural resilience in Ganzhou District, the obstacle degree model was used to calculate the obstacle degree of resilience indicator factors for each region in the years 1990, 2000, 2010, and 2020. The average values were then calculated and ranked to grasp the internal risks to rural resilience development as a whole. As per

Table 2. Average degree of barriers for rural resilience indicator factors in Ganzhou District from 1990 to 2020.

Years	Average Indicator Factor Barrier Degree (%)
1990	X3	X14	X4	X11	X8	X10	X12	X13	X9	X2	X7	X6	X5	X1
	25.31	13.31	12.12	9.59	8.71	7.54	5.16	5.06	4.43	4.21	2.76	0.96	0.66	0.17
2000	X3	X14	X4	X11	X8	X10	X13	X12	X9	X2	X7	X6	X5	X1
	26.74	12.26	11.30	10.33	8.67	7.45	5.21	4.48	4.41	4.30	3.08	0.95	0.66	0.17
2010	X3	X4	X14	X11	X8	X10	X2	X13	X9	X12	X7	X6	X5	X1
	25.52	11.81	11.38	11.17	8.61	7.44	6.45	5.34	4.40	3.86	2.21	0.95	0.65	0.18
2020	X3	X4	X14	X11	X8	X10	X13	X2	X9	X12	X7	X6	X5	X1
	26.81	11.48	11.48	10.22	8.57	7.31	6.07	5.86	4.37	4.13	1.93	0.94	0.65	0.18

, factor X₃ consistently had the highest average obstacle degree across all years, trailed by X₁₄ and X₄. This suggests that land income per unit area is the primary constraint on rural resilience development in Ganzhou District, with landscape sprawl and per capita income also posing significant challenges, the latter increasingly so. Factors X₁₁, X₈, and X₁₀, which pertain to the Shannon diversity index, distance from ecological protection areas, and distance from roads, respectively, consistently ranked between 4th and 6th, highlighting their stable impact as secondary obstacles. Factors X₁₂, X₁₃, X₉, and X₂ had lower average obstacle degrees with fluctuating rankings, indicating a less stable and minor obstructive effect on resilience development. Lastly, factors X₇, X₆, X₅, and X₁ consistently ranked low, implying a persistently smaller hindrance to rural resilience. While the rankings of specific factors have varied, the key risk factors, particularly high obstacle degree factors like X₃, X₁₄, X₄, and X₁₁, have remained dominant with only minor fluctuations in their obstacle degree values.

3.2.2. Spatial Clustering of Major Barrier Factors

To further identify the main risk factors affecting the resilience of rural towns in Ganzhou District, the average degree of barriers for resilience indicator factors across all regions for each year was calculated and ranked. Then, based on the indicator factors with a high average barrier ranking and a cumulative sum greater than 50%, the main barrier factors were identified. Finally, the GIS symbol system was used for spatial visualization. Figure 6 illustrates that the key barriers to rural resilience development in Ganzhou District are land income (X₃), per capita income (X₄), distance from ecological protection areas (X₈), distance from roads (X₁₀), Shannon diversity index (X₁₁), landscape morphological index (X₁₃), and landscape sprawl index (X₁₄). The spatial distribution of these dominant barriers reflects the diversity and complexity stemming from varying water and soil resource use, ecological protection, and economic development across rural towns. However, notable spatial clustering trends emerge: the southern Huazhai and Anyang townships are characterized by X₃-X₄-X₈, northern rural towns by X₃-X₄-X₁₄, the southern suburbs by X₃-X₁₁-X₁₄, and central rural towns exhibit a more diverse and complex combination of factors.

Based on the connotations of the indicator factors, land income and per capita income (X₃, X₄), as efficiency-leading barriers, indicate the comprehensive utilization efficiency of water and soil resources in arid irrigation agricultural areas. The distance from ecological protection areas and roads (X₈, X₁₀), known as habitat resistance barriers, reflects the habitat system’s disturbance resistance capacity. The Shannon diversity index, landscape morphological index, and landscape sprawl index (X₁₁, X₁₃, X₁₄), classified as landscape stability barriers, measure the stability of land use patterns in these areas. Overall, the efficiency-leading barriers, notably land income and per capita income, are widespread across Ganzhou District, excluding the multifunctional agricultural area of Liangjiadun Town, highlighting the inefficiency in water and soil resource use as a primary resilience obstacle. Habitat resistance barriers are distinctly clustered in the west and suburban areas, with the west characterized by extensive farming and breeding that heavily impact the environment and lack ecological protections, resulting in poor disturbance resistance. Suburban areas suffer from urban interference, leading to weak habitat resilience. Landscape stability barriers are evenly spread, suggesting that various land use activities, from construction to agriculture, have pervasively affected the landscape pattern, rendering it generally fragile across the district.

4. Discussion

4.1. Theoretical Contributions

The Heihe River, as a typical representative of inland river basins in arid regions, faces prominent contradictions between human activities and land use in its midstream rural areas under water resource constraints [46]. The sustainability of agriculture and rural development is severely threatened, necessitating an enhanced capacity of the rural territorial system to cope with risks [25]. However, due to the difficulty in obtaining long-time series socio-economic data at the micro-township scale, current research on inland river basins in arid regions predominantly focuses on natural resource systems such as water [53], soil [54], and ecology [47,48]. Resilience studies that reflect risk governance are rarely addressed, and the few existing studies are based on static household survey data [55,56], or fail to fully demonstrate the changing characteristics [57,58].

In fact, resilience is not an ability that a system temporarily generates in the face of current changes or disturbances in the external environment, but is deeply rooted in the system’s historical memory of various shock fluctuations it has experienced in the past [25]. Through its self-organizing functions such as memory, regulation, adaptation, and evolution, the rural system has formed an ability to absorb, defuse, and cope with shocks and disturbances that come from both inside and outside the system, are periodic or aperiodic, and are either short-term and drastic or long-term and gradual [2,11]. It continuously accumulates and reserves this ability and behavior, making it a part of the self-organizing function of the rural system [7]. This property accumulated in the historical formation of the rural system enables it to instinctively and habitually demonstrate resistance, adjustment, and adaptation to shocks when faced with disturbances. It is precisely this ability of the system that maintains and ensures the sustainable existence of the countryside in the past and present.

However, due to the dilemma of the difficulty in obtaining water resources and socio-economic data of villages and towns in the historical sequence, it is necessary to construct a rural resilience assessment framework that can not only simplify the indicators but also reflect the regional characteristics of the middle reaches of the Heihe River. Therefore, based on landscape ecology and complex adaptive systems theory, this research examines resilience from three dimensions: disturbance (emanating from social and economic development pressures), resistance (manifested as the stability of the habitat system), and adaptability (reflected in the dynamic response of the landscape pattern). This framework not only uncovers the characteristics of resilience changes in the historical sequence at both the grid and township scales but also furnishes a cross-scale analytical tool for the adaptive management of the human—land system amidst global change. Moreover, it propels the theoretical integration of landscape ecology and rural geography. The empirical analysis of Ganzhou District, a quintessential irrigated agricultural area, indicates that achieving rural resilience necessitates striking a balance among ecological protection, economic efficiency, and landscape sustainability. This provides theoretical underpinnings for tracing the origin of vulnerabilities within rural systems in arid regions.

4.2. Policy Implications

Resilience is the underlying mechanism for the sustainable development of rural areas, and its development is influenced by many factors [59]. An empirical study of Ganzhou District, a typical irrigated agricultural area, shows that the low utilization efficiency of water and soil resources is the primary obstacle to the development of rural resilience in the midstream region of the Heihe River. Enhancing the resistance of habitats and the stability of the landscape are key strategies for resilience governance in this area. Spatially, the dominant obstacle factors in different townships exhibit characteristics of diverse and complex combinations, but there is also a clear clustering trend. For example, the townships in the south (such as Huazhai and Anyang) are mainly constrained by economic efficiency, while the townships in the north are more affected by landscape sprawl. This spatial differentiation provides a clear direction for local governments to implement precise management and formulate differentiated measures.

The enhancement of rural resilience requires the coordinated promotion of multi-dimensional strategies [60]. For example, the theory of ecosystem services should be deeply integrated into spatial planning, an ecological network should be constructed through natural solutions, and the ecological regulation function should be enhanced with green infrastructure [61]. At the same time, multi-criteria decision analysis tools should be adopted to coordinate economic benefits, social equity, and ecological values, so as to achieve the organic integration of rural production, living, and ecological spaces and systematically improve the capacity for sustainable development [62]. In order to improve the utilization efficiency of water and soil resources and enhance the stability of the ecological environment in the midstream region of the Heihe River, the following measures can be effectively implemented:

(1) Water resources. Upgrade traditional irrigation infrastructure with advanced water-saving technologies to improve water conveyance efficiency. This includes adopting drip or sprinkler irrigation systems that can minimize water waste. At the same time, enhance farmers’ awareness of water conservation through financial incentives publicity, and educational activities, encourage them to adopt water-efficient agricultural production methods, and implement a strict water extraction permit system to ensure the sustainable management and protection of groundwater resources and prevent overexploitation.

(2) Land use. Optimize the land use pattern according to the carrying capacity of resources and the environment, adjust the crop planting layout, reduce the planting area of high water-consuming crops, and promote crop varieties that are more suitable for local climatic conditions and have lower water requirements. Meanwhile, formulate reasonable land use policies to prevent unreasonable reclamation activities. In particular, initiate the project of returning farmland to forests and grasslands in ecologically sensitive areas, and protect and restore wetland ecosystems to support biodiversity and ecological balance.

(3) Ecological agriculture. Encourage the development of ecological agriculture, reduce the dependence on chemical inputs such as fertilizers and pesticides, and foster a more environmentally friendly agricultural production mode. Implement strict environmental supervision over large-scale planting and breeding activities, and impose penalties on any activities that cause pollution. In key areas, especially in the western and suburban areas, establish ecological buffer zones and wetland reserves to mitigate the environmental impact of intensive agriculture.

4.3. Research Limitations

Rural resilience is shaped by a multitude of factors, such as demographics, economy, society, resources, ecology, environment, and governance, necessitating high-quality, long-term data for a thorough analysis of its evolution and risk management strategies. However, micro-level rural data is scarce due to constraints in local records and statistical resources. This study employs the “disturbance-resistance-adaptability” framework to streamline the indicator system and evaluate the resilience characteristics of Ganzhou District, a typical irrigated agricultural area in the Heihe River’s middle reaches, across grid and township scales. While this methodology meets the research goals, the indicator limitations have impacted the precision of internal risk identification within rural resilience.

5. Conclusions

Based on the case study of Ganzhou District, a typical irrigated agricultural area in the middle reaches of the Heihe River, this research analyzes the characteristics of rural resilience from 1990 to 2020 and identifies key obstacles affecting the development of rural resilience. The results show that:

(1) From 1990 to 2020, the overall rural resilience index in Ganzhou District was relatively low with significant spatial differences. Over these 30 years, the social and economic development disturbance in the region consistently showed a pattern of “high in the middle, low in the north and south”, the habitat system resistance exhibited a “low in the southwest, high in the northeast” pattern, and the landscape pattern adaptability was complex and discrete, leading to a resilience differentiation pattern characterized by “low in the west, high in the east”. Townships with developed multifunctional agriculture, nature reserves, and ecological-cultural control lines had higher resilience.

(2) From 1990 to 2020, the rate of change in the rural resilience index in Ganzhou District varied over different stages. Over the 30-year period, the overall rural resilience index experienced a process of “relative stability-increase-decrease”, forming a differentiation pattern of “decrease in the north, increase in the south”. Among them, during 1990–2000, affected by chronic disturbances of arid climate, only the resilience index of individual townships in urban areas and the northern fringe decreased; from 2000–2010, there was a general increase, especially in areas influenced by water allocation plans; from 2010–2020, due to resource overconsumption and developmental impacts brought about by the expansion of agricultural production space, there was a general decline.

(3) The internal risks to the development of rural resilience in Ganzhou District mainly stemmed from low economic efficiency, fragile ecological environments, and unstable landscape patterns. Per capita, land income, landscape sprawl, and per capita income were the primary limiting factors for rural resilience development in Ganzhou District, followed by the Shannon diversity index, distance from ecological protection zones, and distance from roads. Efficiency-dominant and landscape-stability obstacle factors had a broader impact range, while habitat-resistance obstacle factors were mainly concentrated in the western and suburban areas.

The disturbance-resistance-adaptability rural resilience assessment framework constructed based on the theory of landscape ecology can, to a certain extent, reveal the formation mechanism of resilience in the inland river basins of the arid regions in Northwest China. However, since rural resilience in arid regions is greatly affected by human factors such as water-saving irrigation technologies, resource utilization efficiency, and local policy interventions, and it is difficult to obtain relevant quantitative data, the simplified indicator system will lead to certain limitations in the evaluation results. In the future, it is planned to conduct in-depth research after supplementing and improving the above data through channels such as questionnaires and in-depth interviews.