Spatiotemporal Characteristics of Land Ecological Security and Its Obstacle Factors in the Yangtze River Basin

Abstract

The Yangtze River Basin serves as the socioeconomic core of China, and rapid development in recent years has intensified the conflict in the area between economic growth and ecological conservation. This study evaluated the spatiotemporal evolution of the land ecological security (LES) across 11 provinces and municipalities in the Yangtze River Basin from 2008 to 2023 by using the framework of the drivers, pressures, state, impact, and response model of intervention. We forecasted the trends of LES (2024–2033) by using a grey prediction model and identified the key obstacles to it through an obstacle degree model. The findings revealed the following: (1) Economic density (D3) and per capita water resources (S4) had significantly high weights, disproportionately impacting LES. Shanghai scored highest for Drivers, Impact, and Response subsystems, while Tibet led in Pressures and State. (2) Basin-wide LES scores improved from “less safe” to “critical safe” but saw no fundamental breakthrough. LES exhibited a three-tier spatial pattern: higher in the middle-lower reaches (e.g., Shanghai, Jiangsu) and lower in the upper reaches (e.g., Qinghai). Tibet remained “critical safe” with minor fluctuations; other regions improved gradually yet mostly remained “less safe” or “critical safe”. (3) Forecasts (2024–2033) indicate continued overall LES improvement. Shanghai and Jiangsu are projected to reach “safe” status, Qinghai will remain “unsafe”, while most others persist as “critical safe”. Basin LES remains fragile, requiring intervention. (4) The Drivers (D) and State (S) subsystems were the primary constraints on LES. Critical obstacle indicators included economic pressure (per capita GDP (D2), D3), resource availability (S4, ratio of effectively irrigated area (I1)), land productivity (agricultural/forestry output per unit area (I3)), and forest coverage rate (R6). Enhancing LES necessitates implementing regionally tailored policies addressing spatial variations, prioritizing urban economic optimization, strengthening water resource management, and ensuring effective cross-regional governance.

1. Introduction

The land ecological system forms the foundation of human survival by delivering essential ecosystem services, such as food and habitat. However, it currently faces daunting threats in the Yangtze River Basin. Rapid urbanization, unplanned expansion, and unreasonable land use have severely disrupted the system and triggered a cascade of issues, including soil pollution and declining quality of soil. Yu et al. [1] found that cadmium (Cd) was the primary contaminant in the intensive vegetable production system used in the Beijing-Tianjin-Hebei region, with 35% of the monitored vegetable soils exceeding the national risk-screening value (0.6 mg/kg) for agricultural soils specified in GB 15618-2018. Moreover, the indices of soil quality based on its physical, chemical, and biological properties were considerably low in degraded red soils in a hilly region in southern China, with a value that was 21.2–28.8% lower than that of undegraded reference soils [2].

Consequently, the land ecological security (LES) has lately emerged as a paramount public concern and a core component of China’s aim to foster an ecological civilization. It embodies the fundamental balance among ecological health, human well-being, and sustainable development, and emphasizes the capacity of the land system to maintain biodiversity and ecological processes, resist threats (e.g., pollution, erosion, and desertification), and sustain long-term productivity and service provision [3,4,5,6].

In recent decades, the Chinese government has implemented a series of measures to protect soil, ranging from legislative frameworks to large-scale projects for ecosystem restoration. Since the 14th Five-Year Plan, 27 integrated projects to protect and restore mountains, rivers, forests, farmlands, lakes, grasslands, and deserts have been launched nationwide (Bulletin on the Status of China’s Land Greening in 2024). The central government has allocated CNY 53 billion (USD 7.3 billion) to the effort in subsidies with cumulative ecological restoration covering over 5 million hectares. Safeguarding the LES requires the sustained integration of scientific restoration, policy-driven governance, and societal participation to reverse the trends of degradation, and to maintain the long-term balance between socioeconomic development and ecological health. This shows the critical need to analyze the status and health of the ecosystem as a prerequisite for mitigating the adverse impacts of human activities on it.

The land ecological security (LES) is a multi-disciplinary concept that integrates natural ecological attributes with socioeconomic factors [7]. These diverse environmental drivers [8] and socioeconomic factors [9,10] influencing LES across different regions. The LES assessment continues to be challenged by methodological inconsistencies. In response to the need for systematic LES evaluation, two broad methodological approaches have emerged: the characteristic indicator method and the indicator system method. The former, including ecological footprint and energy analysis, offers simplified assessments but often fails to identify specific risk drivers due to its reliance on highly aggregated metrics [7]. In contrast, the indicator system method encompassing models such as Pressures–State–Response (PSR), Drivers–Pressures–State–Impact–Response–Management (DPSIRM), Driver–Pressure–State–Welfare–Response (DPSWR), and Drivers–Pressures–State–Impact–Response (DPSIR) models enables multidimensional assessment and facilitates the diagnosis of underlying obstacles [11,12,13,14].

Among these, the DPSIR framework [15] links socioeconomic Drivers with environmental Pressures, changes in ecosystem State, associated Impacts, and societal Responses. By integrating human activities with ecological outcomes, DPSIR effectively captures cross-scale interactions within land systems [6,16]. Moreover, its dynamic structure supports scenario simulation under different policy interventions, making it a powerful tool for informing sustainable land management decisions [6,14,17]. As such, DPSIR provides a robust and adaptive analytical foundation essential for addressing the complex and evolving challenges in LES assessment. However, the choice of indicators used to assess the LES varies considerably across studies. This variation stems from regional characteristics, the availability of data, and differing conceptual interpretations of the LES [13,14,18,19]. Researchers often resort to proxy indicators that may not directly capture the target variables, such that this obscures the relevant causal relationships. Crucially, the prevalent use of proxy indicators and the absence of standardized frameworks hinder cross-study comparability [14,16,20]. Therefore, establishing a unified evaluation system is essential to ensure consistent LES characterization and support effective cross-regional policy-making.

The LES evaluation primarily involves determination of weights of the variables, classification of levels of ecological security, predictions in scenarios of early warning, and analyses of obstacles and spatial distribution [6,16]. For instance, Li et al. [6] identified the area subjected to soil erosion, volume of sewage discharge, area of control over waterlogging, area of sown grains, rate of urban green coverage, and effective area of irrigated farmland as the main obstacles to the LES. Li et al. [6] demonstrated that a combination of the DPSIR model and the GM can be used to predict the trends of ecological security, which in turn provides scientific support for ensuring the ecological protection of national nature reserves. Meanwhile, the thresholds of classification remain inconsistent across studies, and researchers often categorize the indices of LES into five discrete levels: safe, relatively safe, critically safe, relatively unsafe, and unsafe. However, the values of the critical threshold have not been standardized [21]. This inconsistency exacerbates uncertainty in determining LES levels, and undermines the comparability of the findings.

The Yangtze River is the longest river in China and Asia, and spans 11 provinces and regions of the country as a natural geographical boundary. Originating on the Tibetan Plateau and emptying into the East China Sea, it serves as the country’s vital waterway. Referred to as China’ s “great granary,” the river’s basin supports nearly a third of China’s population, and forms the economic and agricultural backbone of the nation. The assessment of LES serves as the foundation of research on ecological security, and is conducive to maintaining the functions of the ecosystem. Although some research has reported the LES of specific regions of the Yangtze River Basin (e.g., the middle-to-lower reaches of the Yangtze River Economic Belt) [22], the ecological security of the entire basin and the factors influencing it remain underexplored. The state of the LES across the Yangtze River Basin likely varies considerably owing to significant regional discrepancies in levels of economic development, population density, industrial structure, and natural geographical characteristics [23] have reported that levels of the LES in the eastern and western regions, encompassing 126 cities within the Yangtze River Economic Belt, surpassed those in the central regions. It is necessary to ensure the objectivity of these assessments of the dynamic spatial differentiation in the LES across the basin, and to accurately identify the key obstacles to it. This is important for formulating differentiated strategies for the ecological protection and restoration of watersheds and for achieving China’s stated goals of sustainable development.

The objectives of this study were to assess and analyze the spatiotemporal distribution of the LES in the Yangtze River Basin in 11 provinces of China from 2008 to 2023, predict the LES in 2024–2033 to provide early warnings of significant issues, identify the obstacles to LES, and propose targeted countermeasures to optimize the pattern of ecological security. We provided theoretical support and practical guidance for the sustainable use of land resources, maintenance of the health of the ecosystem and advancement of China’s goal of an ecological civilization within the Yangtze River Basin.

2. Materials and Methods

2.1. Study Area

We considered 11 provinces and municipalities within the Yangtze River Basin, as illustrated in Figure 1. The upper reaches of the river extend from its source to Yichang, and encompass mainstream sections like the Tuotuo River and Jinsha River along with their tributaries in Qinghai Province, Sichuan Province, Tibet Autonomous Region, Yunnan Province, and Chongqing Municipality. The middle reaches span from Yichang to Hukou, which are characterized by plains and lakes, and feature the Jingjiang River section and such water systems as Dongting Lake and Poyang Lake. They cover Hubei, Hunan, and Jiangxi Provinces. The lower reaches appear to the area east of Hukou, where the terrain flattens and the river network becomes dense to form the Yangtze River Delta. Its estuary is located east of Chongming Island in Shanghai, and includes Anhui Province, Jiangsu Province, and Shanghai Municipality.

2.2. Evaluation Index Based on DPSIR Framework

The DPSIR model has been widely applied to assessments of the LES [11,12], and contains economic, social, and environmental dimensions. However, no unified standard of evaluation has yet been established owing to significant ecological heterogeneity across the study area as well as the complex interplay of natural, economic, and social factors in the LES. Based on the characteristics of land use and ecological features of the Yangtze River Basin in conjunction with currently available systems of indicators based on the DPSIR model [13,14,18,19], we selected 28 representative indicators based on their comprehensiveness, representativeness, and operationalizability, as shown in

Table 1. System of indices of the LES for the Yangtze River Basin.

Criterion Layer	Weight	Index Layer	Symbol	Measuring Unit	Weight
Driving force (D)	0.2916	Natural population growth rate (D1)	−	‰	0.023
		GDP per capita (D2)	+	ten thousand yuan/person	0.046
		economic density (D3)	+	ten thousand yuan/km2	0.21
		Urbanization rate (D4)	−	%	0.021
		GDP growth rate (D5)	+	%	0.0051
Pressure (P)	0.0764	Population density (P1)	−	people/km2	0.013
		food production per capita (P2)	+	kg/person	0.020
		Fertilizer application per unit area (P3)	−	kg/hm2	0.021
		Unit industrial sulfur dioxide emissions (P4)	−	ton/km2	0.0027
		The usage amount of agricultural plastic films by the unit (P5)	−	ton/km2	0.0091
		Pesticide application rate per unit area (P6)	−	ton/km2	0.016
State (S)	0.2875	Per capita planted area of crops (S1)	+	hm2/person	0.016
		Per capita green space (S2)	+	km2/person	0.018
		Greenery coverage of urban area (S3)	+	%	0.0043
		per capita water resource (S4)	+	m3/person	0.24
		Per capita cultivated land area (S5)	+	hm2/person	0.021
		Construction Land area per capita (S6)	−	m2/person	0.0076
Impact (I)	0.1715	Effective irrigated area ratio (I1)	+	hm2/person	0.050
		Grain yield per hectare (I2)	+	km2/person	0.023
		Output value of agriculture, forestry, animal husbandry, and fishery per unit area (I3)	+	yuan/hm2	0.057
		Per capita disposable income of rural residents (I4)	+	yuan/person	0.033
		Multiple cropping index (I5)	+	%	0.020
Response (R)	0.1730	The proportion of the tertiary industry in GDP (R1)	+	%	0.026
		Centralized treatment rate of sewage treatment plant (R2)	+	%	0.0085
		Comprehensive utilization rate of industrial solid waste (R3)	+	%	0.019
		Harmless disposal rate of domestic garbage (R4)	+	%	0.0073
		Ratio of planted area (R5)	+	%	0.025
		forest coverage rate (R6)	+	%	0.038

Note: “+” denotes indicator positively related with LES and “−” denotes a negative relationship.

.

The driving forces (D) within the DPSIR framework are the social and economic factors that increase or reduce pressure on the land ecosystem of the basin. They cover such factors as socioeconomic development, population growth, policy orientation, and technological changes. These factors have direct or indirect impacts on the land ecosystem by changing the patterns of land use, production, and consumption. Pressures (P) describe the specific role of or load on the land ecosystem. They are mainly manifested as the unreasonable use of land that results in the loss of its functions. This includes the generation of pollutants, such as industrial emissions and agricultural waste. Once these substances enter the ecosystem, they can damage living organisms to affect the structural integrity and functional performance of the land ecosystem. The state (S) reflects the actual status of the land ecosystem, including its health, the status of use of cultivated land and land for construction, and the area of greenery. The impact (I) reflects the influence of changes in the state of land ecosystems on human social and economic activities, and considers such issues as the degradation of the ecosystem functions of the land and soil pollution, and their influence on the quality of agriculture and forestry, and use of water resources. It also determines how these impacts further feed back into the economy, culture, and quality of life of society. The response (R) reflects the policy-related responses of society to issues with the LES, with a focus on the industrial structure, treatment and disposal of pollutants, and measures to improve the environment of land. The conceptual framework and analysis process of this study is illustrated in Figure 2.

2.3. Standardization of Data and Determination of Weight of Evaluation Indicators

To ensure the comparability and accuracy of the results of evaluation, we standardized the data to eliminate the influence of differences in the dimensions, units and orders of magnitude of different indicators. This allowed us to compare them on the same scale, such that their relative importance in the overall system of evaluation could be accurately determined [24,25]. Data standardization was carried out by using the maximum–minimum normalization method. The formulae are as follows:

Positive indicator:

$$X_{ij}^{\prime }={\displaystyle \frac{X_{ij}-X_{\min }}{X_{\max }-X_{\min }}}$$

(1)

Negative indicator:

$$X_{ij}^{\prime }={\displaystyle \frac{X_{\max }-X_{ij}}{X_{\max }-X_{\min }}}$$

(2)

In the above formulae, $X_{\max }$ is the maximum value of a given variable, $X_{\min }$is its minimum value, and $X^{‘}$_ij is the standardized data item.

The weight is a quantitative measure of the relative importance of indicators of evaluation within an assessment system and provides a direct reflection of the influence of each indicator and its contribution to the overall evaluation. We used the entropy weight method to determine the weights of the indicators [26,27]. This method has several advantages, including operational simplicity, strong objectivity, high reliability, and the ability to minimize subjective bias when weighing the factors to yield representative results [11]. The computational procedure is as follows [18]:

(1)

Ratio of each indicator:

$$P_{ij}={\displaystyle \frac{X_{ij}^{\prime }}{{\displaystyle {\sum }_{i=1}^n{X}_{ij}^{\prime }}}}, i=1,2,\mathrm{3\dots \dots }n$$

(3)

(2)

Entropy of each indicator:

$$e_j=-k{\displaystyle {\sum }_{i=1}^n{P}_{ij}\ln P_{ij}}, k={\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$\ln m$}\right.}, and k>0$$

(4)

(3)

Difference coefficient of each indicator:

$$g_j=1-e_j$$

(5)

(4)

Weight of each indicator:

$$W_j={\displaystyle \frac{g_j}{{\displaystyle {\sum }_{j=1}^m{g}_j}}}, j=1,2,\mathrm{3\dots \dots }m$$

(6)

In the above, i = 1, 2, …, n, where n is the year in which the data were collected, n = 16; j = 1, 2, …, m, and m is the number of indicators of evaluation, m = 28. A large weight represents a strong impact on the system of evaluation. The calculated weights of the criterion layer of the DPSIR framework for the LES were the sum of weights of all indicators of the subsystem.

2.4. Evaluation and Classification of LES

The comprehensive index of the LES was obtained by the following formula [28]:

$$T_{ij}=X_{ij}^{\prime }\times W_j$$

(7)

$$E_i={\displaystyle {\sum }_{j=1}^m{T}_{ij}}$$

(8)

$X^{‘}$_ijis the standardized value of the j-th indicator in the i-th year, $W_j$ is the weight of the j-th indicator, T_ij is the LES of the j-th indicator in the i-th year, m is the number of indicators, and E_i is the comprehensive LES in the i-th year. Higher values indicate better LES conditions.

No standardized system of classification is available for assessing the LES. We used a method of classification of unequal intervals [3,29] to evaluate provincial-level LES values in the Yangtze River Basin. We categorized the results into five tiers, as shown in

Table 2. Classification of LES.

Safety Class	Safe Level	Safety Value Interval	Ecological Characteristics
I	Unsafe	[0.00, 0.20)	Land ecological security is very insecure with the damaged system structure and seriously degraded function, reduced resources, and severe pollution.
II	Less safe	(0.20, 0.30]	Land ecological security is insecure with fragile ecological structure, partially lost function, and increased pollution risk.
III	Critical safe	(0.30, 0.50]	Land ecological security is in a critical security state with damaged ecosystem structure and, basically maintained function easily disturbed by the out system.
IV	Relatively safe	(0.50, 0.80]	Land ecological security is relatively safe, with a stable ecological structure, pleasing function, sustainable resources, and anti-risk solid ability.
V	Safe	(0.80, 1.00]	Land ecological security is safe with complete ecological structure, vital function, sufficient resources, stable environment, and controllable risk.

. This approach helped integrates insights from prior studies [13,14] while accounting for regional ecological characteristics.

2.5. Grey Prediction Model GM(1,1)

We used data on the study area from 2008 to 2023 with the grey prediction model GM(1,1) to forecast the trends of the LES in the Yangtze River Basin from 2024 to 2033. The GM(1,1) model has been widely used to predict ecological security owing to its computational efficiency and reliable accuracy [21]. $x^{\left(0\right)}\left(n\right)$ in the model denotes the original time-series data. The cumulative sequence $x^{\left(1\right)}\left(n\right)$ was derived through cumulative generation, and was then modeled by using a first-order linear differential equation [30].

$${\displaystyle \frac{d{x}^{(1)}}{dt}}+a{x}^{(1)}=u$$

(9)

where a represents the coefficient of development and u is the grey action. The grey prediction equation can then be written as follows:

$${\widehat{x}}^{(1)}(t+1)=[{\widehat{x}}^{(0)}(1)-{\displaystyle \frac{u}{a}}]{\exp }^{-at}+{\displaystyle \frac{u}{a}}$$

(10)

where t represents a time series (t = 1, 2, 3, …, n). The accuracy of the model was validated by using two metrics: the posterior difference ratio (C) and the probability of a small error (p). These criteria are defined as follows:

$$C=S_1/S_2$$

(11)

$$p=\left\{\left.\left|q^{(0)}(t)-{\overline{q}}^{(0)}\right.\right|<0.6745S_1\right\}$$

(12)

where S₁ represents the deviation in the data and S₂ the deviation in the residuals. The accuracy of prediction was classified into four grades: good (p ≥ 0.95, C ≤ 0.35), qualified (0.80 ≤ p < 0.95, 0.35 < C ≤ 0.50), barely qualified (0.70 ≤ p < 0.80, 0.50 < C ≤ 0.65), and unqualified (p < 0.70, C > 0.65).

We now validate the GM(1,1) model. Let S₁ denote the standard deviation of the original time-series data and S₂ that of the residual errors. The predictive accuracy of the model was then assessed by using the following standardized criteria of classification [31]: excellent (p ≥ 0.95, C ≤ 0.35), qualified (0.80 ≤ p < 0.95, 0.35 < C ≤ 0.50), barely qualified (0.70 ≤ p < 0.80, 0.50 < C ≤ 0.65), and unacceptable (p < 0.70, C > 0.65).

2.6. Obstacle Degree Model

We formulated an obstacle degree model to intuitively analyze the factors hindering the achievement of high levels of LES. By introducing the degree of deviation of the indicators and the degree of obstacles, we conducted an obstacle factor analysis to comprehensively assess the impact of the factor and indicator layers on the LES. The calculation was as follows [18]:

$$I_{ij}=1-X_{ij}^{\prime }$$

(13)

$$O_{ij}={\displaystyle \frac{W_j{I}_{ij}}{{\displaystyle {\sum }_{j=1}^m{w}_j{I}_{ij}}}}\times 100\%$$

(14)

$$U_{ij}={\displaystyle \sum O_{ij}}$$

(15)

The standardized value of the index is denoted by $X^{‘}$_ij, and $I_{ij}$ represents the degree of deviation of the j-th index in the i-th year. The $w_j$ denotes the weight of the j-th index, and can be used to derive the degrees of obstacles in both the index layer ($O_{ij}$) and the criterion layer ($O_j$).

2.7. Sources of Data and Their Processing

Data on the indicators of the LES were obtained from the statistical yearbooks, statistical bulletins on economic and social development, and bulletins on the water resources of all considered provinces as well as the China Urban Statistical Yearbook and China Forestry Statistical Yearbook. Certain economic, social, and ecological data were obtained from provincial statistics bureaus, the China Economic and Social Big Data Research Platform, and the relevant literature. The dataset encompassed 4928 data points (11 provinces × 28 indicators × 16 years, 2008–2023). Within this dataset, missing values were confined to 10 instances, specifically the per capita disposable income of farmers in Tibet and Qinghai for the years 2008–2012, accounting for a negligible proportion of approximately 0.20%. Given the limited extent and localized nature of the gaps, the missing values were addressed using linear interpolation to preserve the dataset’s completeness in a systematic manner.

The data were processed by using Microsoft Excel and SPSS 27, while ARCGIS 10.8.1 and Origin 2022 were used for mapping.

3. Results and Analysis

3.1. Weights of Indicators of LES

As is shown in Figure 3, Pearson’s correlation analysis was applied to assess the relationships among 28 indicators. The findings revealed moderate to strong correlations between several variables. Of the forces driving economic development, a robust positive correlation (r = 0.84) was obtained between the per capita GDP (D2) and the rate of urbanization (D4), which mirrored the synergistic bond between economic growth and urban expansion. The population density (P1) was highly positively correlated with both the use of plastic film for agriculture per unit area (P5) (r = 0.95) and the per capita area of land for urban construction (S6) (r = 0.81), where this highlighted intensified human activities on land. The correlation between the per capita grain output (P2) and per capita area of sown crop (S1) (r = 0.84) illustrated the close nexus between the use of land for agriculture and the capacity for food production. Notably, while these correlations were statistically significant, each indicator encapsulated distinct facets of regional development within the DPSIR framework.

Principal component analysis (PCA) was also applied to the indicators in the DPSIR framework to identify potential relationships among them based on the levels of significance of Bartlett’s test (p < 0.01;

Table A1. KMO and Bartlett’s test.

KMO Measure of Sampling Adequacy		0.776
Bartlett’s Test of Sphericity	Approximate Chi-Square	9556.219
	Degrees of Freedom	378
	Significance	<0.001

in the Appendix A). The cumulative percentage of total variance of the first six principal components was 87.366% (

Table A2. Matrix eigenvalues and cumulative contribution rate.

Component	Initial Eigenvalues			Extraction Sums of Squared Loadings			Rotation Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	10.731	38.324	38.324	10.731	38.324	38.324	8.522	30.436	30.436
2	5.837	20.846	59.170	5.837	20.846	59.170	4.795	17.125	47.561
3	3.925	14.016	73.186	3.925	14.016	73.186	3.611	12.897	60.459
4	1.830	6.535	79.721	1.830	6.535	79.721	3.539	12.640	73.098
5	1.134	4.051	83.772	1.134	4.051	83.772	2.186	7.808	80.907
6	1.006	3.595	87.366	1.006	3.595	87.366	1.809	6.460	87.366
7	0.763	2.726	90.093
8	0.563	2.010	92.102
9	0.444	1.586	93.688
10	0.379	1.353	95.042
11	0.329	1.174	96.216
12	0.214	0.765	96.980
13	0.188	0.670	97.650
14	0.135	0.481	98.131
15	0.115	0.410	98.542
16	0.100	0.357	98.899
17	0.074	0.265	99.163
18	0.062	0.221	99.384
19	0.054	0.192	99.577
20	0.032	0.114	99.691
21	0.027	0.098	99.788
22	0.017	0.060	99.849
23	0.015	0.054	99.903
24	0.010	0.036	99.939
25	0.009	0.032	99.971
26	0.004	0.014	99.985
27	0.002	0.008	99.993
28	0.002	0.007	100.00

), indicating that these components could explain most of the information in the DPSIR framework. A detailed analysis has been provided in the Appendix A. This multi-dimensional perspective enriched the theoretical basis for analyzing the LES, and showed how interdependent socioeconomic and environmental factors collectively molded ecosystem dynamics.

Figure 4a, revealed significant disparities among them (Figure 4b). Two key indicators, the economic density (D3) and per capita water resources (S4), had substantially larger weights (>0.20) than the other factors which reflected their dominant role in determining the LES. The remaining indicators had relatively smaller weights (ranging from 0.0027 to 0.057), but still contributed to the LES to some extent. This pronounced pattern of weights directly influenced the performance of the subsystem with the drivers and state subsystems achieving notably higher scores, while the pressure subsystem had lower scores (Figure 4b).

3.2. Spatiotemporal Discrepancy in LES Subsystems

(1)

Drivers subsystem

The results of evaluation of the five subsystems (drivers, pressures, state, impact, and response) in 11 provinces/municipalities are presented in Figure 5. From 2008 to 2023, Shanghai consistently led in terms of scores of the driving forces, with Jiangsu Province ranking a distant second. Both maintained a significant advantage over the other regions, and their scores were clustered closely and exhibited modest growth. As China’s premier economic hub, Shanghai represented highly intense development. Its per capita GDP (190,321 CNY) and economic density (810.79 million CNY/km²) in 2023 far exceeded the regional average [32]. This reflected its successful transition to a high-value-added economy (e.g., digital/financial sectors), which reduced pollution from traditional industries. Similarly, Jiangsu, a manufacturing powerhouse, exhibited sustained growth in terms of scores of the drivers. Its per capita GDP of 150,487 CNY and economic density of 122.13 million CNY/km² in 2023 ranked among the nation’s highest, and were driven by industrial upgrade. This included the integration of advanced manufacturing with modern services and digital transformation. These initiatives spurred green transitions in enterprises and fostered strategic emerging industries that led to a direct increase in the LES in both regions.

(2)

Pressures subsystem

The scores of the pressures subsystem reveal an inverse relationship with the levels of ecological stress (higher scores indicate lower pressure). The regional rankings could be categorized into three tiers: low-pressure regions (Tibet Autonomous Region, and Qinghai, Sichuan, Yunnan, Chongqing, Anhui, Hunan, and Jiangxi Provinces), moderate-pressure regions (Hubei and Jiangsu Provinces), and high-pressure regions (Shanghai Municipality). As China’s premier global financial center, Shanghai had the most severe land ecological pressures that were evidenced by its significantly low subsystem score, mainly owing to demographic and spatial pressures, challenges to industrial legacy, and agricultural constraints. As a global financial hub, Shanghai has continually attracted talent and labor. Its population density was 3923 people per square kilometer in 2023, and has led to a burgeoning demand for land for construction that has reduced the area of ecological land. The excessive exploitation of resources has exerted significant pressure on the LES [33]. Over the past decade or so, such heavy industries as steel and chemical engineering have accounted for a high ratio of its economy and the base of emissions remains large despite industrial transformation in recent years [32]. Meanwhile, the resources for agricultural land in Shanghai are limited. The use of plastic films for agriculture in the area is higher than in other regions along the Yangtze River. The residue of agricultural films directly leads to a decline in the quality of farmland soil and poses a significant threat to the ecological security of land.

(3)

State subsystem

The Tibet Autonomous Region and Shanghai yielded the highest and lowest scores, respectively, for the state subsystem, with a pronounced disparity between them. The other regions recorded intermediate scores that were clustered closely together. The Tibet Autonomous Region’s exceptional score stemmed predominantly from its abundant water resources. Its per capita water resources were 121,462.28 m³ in 2023 and ranked among the nation’s highest [34]. This abundance was sustained by reliable replenishment from glacial meltwater from the plateau and precipitation, and helped maintain stable water supply for wetlands, meadows, and other critical ecosystems. Furthermore, the region had a low intensity of use of cultivated land. Despite having only 4.4525 million hectares of arable land in 2023, its low population density (3.03 persons/km²) ensured a high per capita redundancy of land resources. These natural advantages, combined with minimal anthropogenic disturbance, underpinned the region’s relatively favorable ecological security conditions. Conversely, Shanghai’s per capita water resources in 2023 stood at a mere 167.27 m³, dramatically lower than the other regions [33]. This situation has been compounded by urban expansion that has persistently encroached on ecological spaces, and has resulted in substandard metrics of the per capita area of sown crops, green space for urban parks, and green coverage in built-up areas [35]. The confluence of these factors has significantly degraded Shanghai’s LES.

(4)

Impact subsystem

From 2008 to 2023, the scores of the impact subsystem exhibited a consistent upward trajectory. Shanghai and Jiangsu Province emerged as top performers, while Yunnan and Qinghai Provinces were persistently ranked the lowest, with other regions occupying intermediate positions. Constrained by limited land resources, Shanghai has pioneered intensive land use practices by adopting facility agriculture, intelligent greenhouses, and precision farming technologies. This, aligned with its modern urban agricultural framework, has yielded a highly efficient output. For instance, its grain yield per unit area was 8070.15 kg/ha in 2023, far exceeding that of traditional agricultural regions. Moreover, such initiatives for rural revitalization as collective asset reforms and tourism development have substantially boosted the income of farmers. Shanghai’s rural disposable income peaked at 42,988.05 yuan in 2023, the highest in the study area. As a national grain-producing hub, Jiangsu has optimized its agricultural output through land transfers and cooperative farming models to ensure a stable growth in production. The province has expanded beyond staple crops to the cultivation of cash crops, and has diversified into agroforestry, livestock, and fisheries to create a multi-faceted industrial system that has increased rural income. Southern Jiangsu has capitalized on its industrial expertise to advance facility-based and ecological agriculture, while northern Jiangsu has focused on high-quality farmlands to enhance its grain production. These strategies collectively yielded 833.53 yuan/ha from agriculture, forestry, livestock, and fisheries in 2023, outperforming other regions. Yunnan’s progress has been hindered by the fragmentation of farmland, low rates of mechanization, and an inadequate rural infrastructure. Qinghai has faced compounded challenges: an ecologically fragile environment, recurrent natural disasters (e.g., droughts and low temperatures), and a monocultural cropping system. These factors have collectively led to relatively low values of the indicators of ecological security of land in both provinces.

(5)

Response subsystem

From 2008 to 2023, the scores of the regional response subsystem in the Yangtze River Basin exhibited pronounced fluctuations. The Tibet Autonomous Region and Qinghai Province consistently recorded the lowest scores, while Shanghai maintained a comparatively high score. The poor performance of Tibet and Qinghai primarily stemmed from their cold and arid climates, and extreme topographical conditions that severely limited forest coverage which is the most heavily weighted indicator in the response subsystem. With average elevations exceeding 3000 m, these regions record mean annual temperatures below 0 °C, receive less than 300 mm of annual precipitation and have soils with a minimal amount of organic matter. Such conditions are prohibitive for the growth and survival of trees, and led to rates of forest coverage that were drastically lower than those of other regions in the basin. Shanghai’s robust response capacity reflected its advanced economic restructuring. It is China’s leading economic hub, and the development of its tertiary sector significantly outpaced the national average. In 2023, the service industry contributed 75% to Shanghai’s GDP, a share that far exceeded other regions of the Yangtze River Basin. This service-dominated economy enhanced institutional agility to enable more effective responses to challenges to land security.

3.3. Spatiotemporal Characteristics of LES

As illustrated in Figure 6, the comprehensive LES scores across the Yangtze River Basin exhibited distinct trends from 2008 to 2023. The Tibet Autonomous Region maintained slight fluctuations in its score within the range of critical safe, while the other regions exhibited gradual improvements and predominantly persisted in less safe states or those of critical safe. The hierarchical classification of LES scores was as follows: Shanghai, Tibet Autonomous Region > Jiangsu Province > Yunnan, Sichuan, Jiangxi, Hubei, Hunan, Chongqing, Anhui > Qinghai Province (Figure 6a). The basin-wide LES score rose from 0.25 (2008) to 0.34 (2023), reflecting a transition from less safe to critical safe (Figure 6b). This upward trajectory shows systemic progress, yet significant disparities persisted among the regions.

As the “Asian Water Tower”, Tibet’s robust LES performance can be attributed to its abundant water resources and proactive conservation policies. Key measures include initiatives to protect the sources of rivers and a management infrastructure for them, implementation of protection for natural forests, wetland restoration, and horizontal mechanisms of ecological compensation. It has also engaged in traditional ecological practices that minimize anthropogenic disturbances to preserve near-pristine ecosystems. Shanghai had surpassed Tibet in terms of the LES score by 2019, and achieved a relatively safe state (0.54 in 2023). This was primarily due to its strong driving forces, influence, and response capacity as an economic center. To ensure a high LES, Shanghai has controlled its water sources, promoted garbage classification, reduced the use of chemical fertilizers and pesticides, implemented projects for the renovation of Suzhou River and the “Ten Thousand Rivers Action,” built shore-to-ship power systems, and optimized sewage treatment and the use of water resources. It has also promoted the upgrade of its industrial structure and increased its farmers’ incomes to form a model of coordinated development that was driven by policies, the economy, and technology. Qinghai faces a severe situation with regard to the LES, and its comprehensive score was consistently at the bottom in the rankings (Figure 6b). Its LES has transitioned from an “unsafe” state to a “less safe” state. Situated on the Qinghai–Tibet Plateau, Qinghai is plagued by severe glacial melting, degradation of grassland, and desertification that have led to a decline in its capacity for water conservation. Moreover, natural disasters such as earthquakes and mudslides have exacerbated the degradation of ecological land. Although Qinghai has implemented projects for ecological protection, their effect remains insignificant due to limitations of funding and technology. Furthermore, it emits a large amount of pollutants and its water system is severely polluted [34]. Lu et al. [36] have also claimed that the ecological security of the Qinghai–Tibet Plateau is poor, and their spatial analyses revealed that the northern Kunlun Mountains and the Huangshui River Valley had the lowest LES levels, while the Qaidam Desert and densely populated urban areas on the plateau were subject to risks as well.

To characterize the spatial distribution of the LES across the Yangtze River Basin, we analyzed four representative years (2008, 2013, 2021, and 2023) that captured key transitions in it (Figure 7). Our analysis revealed significant spatial heterogeneity in the pattern of the LES. Bolstered by its unique natural resource endowments, the Tibet Autonomous Region in the upper reaches maintained a critical state of security in 2008 while Qinghai Province remained in an unsafe state. All other regions were categorized as relatively unsafe. Downstream regions like Shanghai and Jiangsu Province yielded an improved LES in 2013 to reach the critical state of security. This improvement then spread from downstream to midstream. Provinces in the middle reaches of the basin (Anhui, Hubei, Hunan) attained critical security in 2021. The basin-wide pattern of security was consolidated in 2023, with only the provinces in the upper reaches (Qinghai, Sichuan, and Yunnan) remaining relatively unsafe. Jing et al. [37] have confirmed that the economically developed lower reaches, with substantial investment in the environment and effective implementation of the relevant policies, sustained high LES scores from 2012 to 2021 in a clear upward trend. The emergent spatial structure yielded the highest LES levels in the economically developed regions of the delta, gradual improvement following downstream precedents in the midstream area, and persistent vulnerabilities in the ecologically fragile upstream regions. This gradient structure of multi-level ecological security explicitly reflected the disparities in the ecological vulnerability of land in different geographical units within the Yangtze River Basin.

The Shanghai Specialized Ecological Space Planning (2021–2035) program was implemented in 2021. It is designed to control the expansion of land for construction, demarcate red lines for ecological protection, and systematically curb the erosion of the land ecology owing to unordered urban expansion. Shanghai has also driven ecological improvements through technological innovation. For example, the Shanghai Water Affairs Implementation Plan for Systematically Promoting Sponge City Construction states that more than 40% of the built-up area in Shanghai should meet the requirements for sponge city construction by the end of 2025, and this ratio should exceed 80% by the end of 2030. This has significantly alleviated the pressure of declining groundwater levels. Jiangsu Province’s Spatial Ecological Protection and Restoration Planning (2021–2035) defines the overall pattern of “Five Zones and Three Belts” for ecological protection and restoration, and establishes a target system for the protection and restoration of important ecosystems. The province has focused on innovative practices to promote collaborative regional governance within and outside its boundaries to explore diversified mechanisms of investment, the construction of an innovation experimental zone for the integrated protection and restoration of mountains, rivers, forests, fields, lakes, grasslands, and deserts, and the creation of a full-chain communication and cooperation platform that covers theoretical research and application. This has injected momentum into measures for ecological protection and restoration. In contrast to the downstream regions, the three provinces in the middle reaches have encountered the problem of increasingly fragmented farmland owing to urban expansion.

For instance, freshwater lakes such as Dongting Lake and Poyang Lake in the Yangtze River Basin have significantly decreased in area, and this has triggered chain reactions that have disrupted the connectivity of the ecological network and obstructed channels for biological migration to increase the difficulty of restoring ecological service functions. Overall, Shanghai and Jiangsu Province have relied on the model of high-intensity investment, technological innovation, and institutional innovation. They have subsidized the costs of ecological governance and formed a sustainable long-term mechanism for improving the ecology. In contrast, other regions are constrained by several factors, including ecologically fragile backgrounds, shortages in governance capacity, and financial and technological constraints, which have made it difficult to achieve phased improvements in LES levels in the short term. Hence, it is evident that the comparison between downstream and midstream regions offers valuable insights for formulating policies for sustainable development. In particular, focusing on spatial disparities in ecological security highlights the significance of collaborative governance, which, under the framework of the Yangtze River Protection Law, requires upstream–downstream provinces to establish ecological compensation mechanisms and joint monitoring systems to transcend administrative fragmentation.

3.4. Predictive Analysis of LES

We used the grey prediction model to analyze the trends of the LES across China from 2024 to 2033. The accuracy of the model varied significantly by region, as detailed in

Table 3. Predictive accuracy of levels of land ecological security.

Area	Variation Ration (C)	Small Error Probability (p)	Grade	Area	Variation Ration (C)	Small Error Probability (p)	Grade
Shanghai	0.0049	1.00	Good	Chongqing	0.034	1.00	Good
Jiangsu	0.019	1.00	Good	Yunnan	0.050	1.00	Good
Anhui	0.021	1.00	Good	Sichuan	0.026	1.00	Good
Jiangxi	0.083	1.00	Good	Tibet	0.74	0.69	unqualified
Hubei	0.017	1.00	Good	Qinghai	0.18	0.88	Qualified
Hunan	0.036	1.00	Good	Yangtze river basin	0.0073	1.00	Good

. It had a high accuracy for nine provinces (p ≥ 0.95, C ≤ 0.35), and captured the trends of the LES in most areas of the Yangtze River Basin. In contrast, the model achieved only a passing level of accuracy for Qinghai, which showed that it had limited predictive reliability for this area. Meanwhile, its performance on the Tibet Autonomous Region was classified as unqualified (p < 0.70, C > 0.65), suggesting substantial uncertainty in the forecasts. This regional discrepancy can be attributed to the distinct environmental conditions of the Qinghai–Tibet Plateau. Its ecosystems have evolved under extreme conditions, including a low temperature, low atmospheric oxygen levels, and high solar radiation that result in an inherently fragile land ecosystem with a weak self-regulatory capacity. Consequently, even minor climatic variations or anthropogenic disturbances can trigger cascading ecological responses, such as a degradation in permafrost and desertification of grasslands. This led to a high volatility in the indices of regional ecological security [38]. These findings highlighted the importance of region-specific approaches to modeling when assessing ecological security, particularly in sensitive and extreme environments like that of the Qinghai–Tibet Plateau.

The values of LES predicted by the grey prediction model in the Yangtze River Basin are illustrated in Figure 6. The LES scores of 10 provinces and cities were projected to continue an upward trend (Figure 6a), signifying an overall improvement in the land ecological environment. Specifically, Shanghai’s LES is expected to reach a safe state by 2032, while Jiangsu Province will enter a relatively safe stage in 2033. Qinghai Province demonstrated a sustained upward trend in its security score, and was projected to remain in a relatively unsafe state throughout the next decade. Other regions were expected to enter into or maintain a state of critical safe in the same period. The ecological security of all regions except for Shanghai and Jiangsu, which were expected to be the first to achieve relatively safe levels, require close monitoring over the next 10 years. The overall LES of the basin exhibited a fluctuating upward trend from 2024 to 2033 but remained within the range of critical safe (Figure 6b). This indicated that although the land ecological environment of the Yangtze River Basin was generally improving, the magnitude of improvement remained limited, and no substantive leap in ecological security was achieved. The LES thus continues to face severe challenges.

A thorough causal analysis revealed a number of systemic challenges, including the accumulation of pollution and the stasis of efforts for ecological restoration as well as unsustainable patterns of land use. Industrial emissions, sewage from irrigation, and agricultural inputs have created a massive baseline for soil contamination.

Pollution control takes a long time and is technically complex, such that it slows ecological restoration. Moreover, ecologically fragile zones in the upstream regions (Tibet and Qinghai) face a crisis owing to insufficient government funding and weak technical support. The midstream provinces (Hubei and Hunan) are subject to pressure from encirclement by the chemical industry along the banks of the Yangtze, while rapid urban expansion is encroaching on ecological zones. Despite notable achievements in pollution control in the downstream areas (Jiangsu and Shanghai), intensive land development continues to encroach on ecological habitats. A basin-wide modern framework of governance is thus urgently needed. The core of this framework should involve establishing cross-regional collaborative mechanisms to address structural challenges. A cornerstone could be the implementation of a basin-wide “River Chief System,” which would institutionalize leadership and accountability, integrating scientific monitoring with cross-regional collaborative mechanisms, which is essential for reversing ecological degradation in the basin.

3.5. Obstacles to LES

Figure 8 shows that the degrees of obstacles to the criterion layer of the LES in the Yangtze River Basin from 2008 to 2023 exhibited distinct dynamic characteristics. Overall, the distribution of the obstacles was dominated by a dual “drivers–state” mechanism, followed by the factors influencing them, while the obstacles to the response and pressure subsystems remained relatively weak. A critical turning point occurred in 2020, when the degree of obstacles to the state subsystem surpassed those to the driving forces subsystem for the first time and rose to the top. A detailed indicator-level analysis reveals that this shift was primarily driven by the following changes, as illustrated in Figure A2. The obstacle degree of S4 (per capita water resources) saw a substantial increase, rising from 5.23% in 2019 to 8.91% in 2020. Concurrently, key indicators of the Driving (D) system exhibited notable decreases: D3 (economic density) fell from 7.34% to 5.12%, while D2 (GDP per capita) declined from 6.85% to 4.97%. These fluctuations align remarkably well with regional observations. The sharp rise in S4’s obstacle degree directly corresponds to the severe floods that impacted multiple provinces along the main stream of the Yangtze River in 2020, particularly Hubei, Hunan, and Jiangxi provinces, where the availability and distribution of water resources were significantly disrupted (China Water Resources Bulletin, 2021). Meanwhile, the reduced obstacle degrees of D2 and D3 reflect the temporary economic adjustments and altered development patterns during this period, which is consistent with the concept of “ecological debt”: accumulated environmental pressures ultimately manifest as deterioration of the state system.

This shift indicated that the core contradiction concerning ecological security in the basin has gradually transitioned from human activity-driven pressures to the evolution of the intrinsic state of the land ecosystem’s inherent properties and functional integrity, particularly its capacity for self-regulation and resilience. This turning point stemmed from the cumulative ecological debt induced by long-term unsustainable development, irrational patterns of land use, and extreme climate events. From 2008 to 2019, the model of extensive development in the Yangtze River Basin resulted in the unregulated expansion of land for construction, continual shrinkage of natural wetland systems, a significant decline in ecological connectivity, and a gradual decrease in the basin’s overall resilience.

Peng et al. [39] found that the overall vulnerability index of the Yangtze River urban agglomeration was in a mild fragile state in 2005, 2011, and 2017, and its most fragile and slightly fragile cities were transitioning to moderate-to-severe vulnerability. Moreover, changes in precipitation are a key climatic factor influencing land ecosystem services. Xu et al. [40] have shown that the capacity of the basin for soil and water conservation did not improve, but instead declined from 2001 to 2010, possibly as a long-term ecological consequence of the 1998 flood. The once-in-a-century flood in 2020 also caused large-scale inundation of wetlands and soil erosion. Dynamic monitoring by the Ministry of Water Resources showed that 23,600 km² of land had been subjected to soil erosion in Jiangxi Province in 2020, accounting for 14.13% of its total area (Jiangxi Province Soil and Water Conservation Bulletin, 2020). Increased precipitation may reduce the height and coverage of vegetation at the source of the Yangtze River [41]. Liu et al. [42] have also noted that changes in land use in Wuhan over the past 20 years have significantly increased its risk of exposure to floods which is positively correlated with economic losses.

The analysis of indices of the degrees of obstacles revealed substantial regional disparities in constraints on the LES (

Table 4. Ranking of obstacles to the LES in the Yangtze River Basin.

Area		Obstacle Degree Ranking of Indicators
		1	2	3	4	5
Shanghai	Obstacle factors	S4	D3	R6	I3	I1
	Obstacle degree/%	39.30	14.32	5.29	4.52	4.15
Jiangsu	Obstacle factors	S4	D3	R6	I1	D2
	Obstacle degree/%	35.03	28.46	4.51	4.32	3.71
Anhui	Obstacle factors	S4	D3	D2	I3	I1
	Obstacle degree/%	32.43	28.66	5.20	4.94	4.45
Jiangxi	Obstacle factors	S4	D3	I3	I1	D2
	Obstacle degree/%	31.90	28.89	6.19	5.41	5.31
Hubei	Obstacle factors	S4	D3	I1	I3	D2
	Obstacle degree/%	31.97	28.39	5.40	4.91	4.72
Hunan	Obstacle factors	S4	D3	I1	I3	D2
	Obstacle degree/%	32.23	28.88	5.46	5.42	5.18
Chongqing	Obstacle factors	S4	D3	I1	I3	D2
	Obstacle degree/%	31.94	28.25	6.67	5.65	4.73
Yunnan	Obstacle factors	S4	D3	I3	I1	D2
	Obstacle degree/%	31.02	28.39	6.70	6.05	5.36
Sichuan	Obstacle factors	S4	D3	I3	I1	D2
	Obstacle degree/%	31.04	28.17	6.33	5.90	5.15
Tibet	Obstacle factors	D3	I3	D2	S4	R6
	Obstacle degree/%	38.16	10.07	7.22	6.35	5.82
Qinghai	Obstacle factors	S4	D3	I3	D2	I1
	Obstacle degree/%	27.39	26.88	7.01	4.99	4.65

). The five most significant obstacles in Shanghai were the per capita water resources (S4), economic density (D3), rate of forest coverage (R6), agricultural/forestry/husbandry output per unit area (I3), and effective area ratio of irrigation (I1). Notably, both the per capita water resources (S4) and economic density (D3) had obstacles exceeding 10%. The particularly high obstacles to the economic density (D3), which is a core indicator of drivers, underscored the tension between economic concentration and ecological carrying capacity in Shanghai. This was manifested in highly intense economic activities that drove the expansion of land for construction, energy consumption, and industrial emissions to create a self-reinforcing negative feedback cycle involving economic density and ecological pressure. Similarly, the high obstacles to the per capita water resources (S4) reflected the fundamental mismatch between Shanghai’s dense population and its natural water endowment. This pattern extended beyond Shanghai, with Jiangsu, Hunan, Sichuan, Yunnan, and Anhui Provinces, along with the Chongqing Municipality and Qinghai Province, facing comparable challenges. Both S4 and D3 had comparably high obstacles. This finding aligned with that by [43], who identified water resources, level of erosion control, and local fiscal expenditure on agriculture as the primary constraints on cultivated LES in China. Lu et al. [36] documented similar patterns that showed that such indicators as the population density, net primary productivity index for vegetation, and GDP per unit area were key regulators of ecological security on the Qinghai–Tibet Plateau.

The index of obstacles in the Tibet Autonomous Region exhibited a distinct distribution. The top five constraining factors there were the economic density (D3), output value per unit area of agriculture/forestry/fishery (I3), per capita GDP (D2), per capita water resources (S4), and rate of forest coverage (R6). Notably, the economic density (D3) had the highest obstacles. This could be attributed to Tibet’s unique geographical and socioeconomic conditions: The region’s terrain is dominated by plateaus and mountains, which increases the cost of infrastructure while leading to dispersed economic activities. Moreover, Tibet’s industrial structure remains underdeveloped, and relies primarily on agriculture, forestry, and tourism. Consequently, the region has a limited economic capacity to invest in critical infrastructure and ecological governance [44]. These findings aligned with those by Lai et al. [45], who identified the per capita GDP and level of regional development as the key obstacles to ecological security in Fuzhou City.

4. Discussion

4.1. Key Obstacles to LES and Policy Recommendations

Under the DPSIR framework and the model of obstacle degrees, the economic density (D3) and per capita water resources (S4) emerged as critical determinants of the LES in the Yangtze River Basin. As a core indicator of the driving forces layer of the DPSIR model, the economic density reflected the spatial intensity of economic activities, and exerted a significantly positive influence on ecological security. Regions with a high economic density leveraged technological innovations, such as clean production technologies and circular economic models, to reduce their resource consumption under the strategy of the high-quality development of the Yangtze River Basin in order to mitigate pressure on land development. Moreover, economically developed areas were supported by robust financial capacities, and were thus well positioned to implement institutional measures like mechanisms of ecological compensation and territorial spatial planning. This fostered a virtuous cycle of economic growth and ecological protection. A higher economic density also facilitated industrial upgrade. For instance, Shanghai and Jiangsu have transitioned toward service-oriented and high-tech industries, and this has substantially lowered pollution emissions from traditional sectors.

In contrast, the Tibet Autonomous Region was constrained by its fragile high-altitude ecosystem, and faced such challenges as a dispersed population, a mono-industrial structure, and acute conflicts between resource exploitation and ecological preservation. These factors have resulted in a low economic density, and have hindered the quality of its urbanization. Similarly, Yunnan Province has struggled with regional disparities owing to its mountainous terrain, and has suffered from a severe outmigration of labor and an overreliance on resource-based industries. This has stymied the growth of emerging sectors and the overall quality of economic development [44].

To address the above disparities, strategies should be formulated over three dimensions: the construction of an intensive industrial ecosystem, strengthening of innovation-driven development and upgrade, and an enhancement in regional coordination in ecological compensation mechanisms, environmental standard alignment, and data-sharing platforms. Regions to the north in the Yangtze River Delta should encourage the agglomeration of high-value-added industries to enhance the efficiency of land use, with a focus on developing strategic emerging sectors like integrated circuits and biomedicine. Moreover, the central and western provinces (e.g., Yunnan and Sichuan) must develop specialized industrial clusters (e.g., green energy and eco-tourism) to avoid homogeneous competition, and complement this with innovative land use policies that prioritize high-tech industries and curb inefficient urban sprawl. To strengthen innovation and upgrade, cross-regional innovation corridors should be established and anchored by such hubs as Shanghai’s Zhangjiang and Wuhan’s Optics Valley. This can help foster industry–academia collaboration and accelerate the adoption of clean production technologies. Furthermore, targeted industrial funds must be allocated to lagging regions (e.g., Tibet and Yunnan) to support the technological transformation of local enterprises and reduce their dependence on resource-based economies. Inter-provincial zones of industrial collaboration should also be created to incentivize labor retention and promote localized urbanization. Concurrently, upgrades in infrastructure (e.g., high-speed rail networks and 5G coverage) in the central and western regions are essential for reducing logistical costs for decentralized industrial layouts.

The per capita water resources (S4) serve as a vital state-level indicator for assessing the ecological security of the Yangtze River Basin, with their sustainability playing a decisive role in maintaining the stability of the land ecosystem [34]. Water resources dynamically regulate the moisture and biogeochemical cycles in soil, thereby preserving its physicochemical properties. The supplementation of glacial meltwater can enhance the retention of water in soil in water-abundant regions, such as high-altitude alpine meadows. This can in turn promote the accumulation of organic matter and increase vegetation coverage to reduce wind erosion and risks to land degradation. Conversely, the compaction of soil occurs to reduce the stability of the aggregate in areas subject to seasonal water scarcity, while intense rainfall during wet seasons exacerbates erosion to accelerate a decline in the productivity of land. Furthermore, the availability of water is the foundational element of wetlands, forests, and other critical ecosystems, and directly influences the functionality of the habitat and its capacity for carbon sequestration [46]. For example, large lakes can help expand riparian vegetation zones and stabilize shoreline soils through ecological water-level management, whereas declining groundwater tables lead to the degradation of wetlands and diminished carbon sinks that intensify imbalances in the land ecosystem. While the Yangtze River Basin has abundant water resources, their spatial distribution is highly uneven. The Tibet Autonomous Region (upper basin) recorded an annual volume of 121,462.28 m³ of water per capita in 2023 while Shanghai (lower basin) had merely 167.27 m³.

A coordinated basin-wide framework of governance is essential to address the above-mentioned disparity. Inter-basin water transfer projects should be developed to secure supply for water-stressed cities like Shanghai to optimize the infrastructure of water allocation. Concurrently, systems to store and regulate glacial meltwater should be constructed in upper-basin regions, such as Tibet, to stabilize water supply in the dry season. Water-saving technologies like industrial wastewater recycling and agricultural drip irrigation should be mandated to promote the conservation of water and the efficiency of its use, while incentives for cultivating drought-resistant crops must be implemented to reduce dependency on water. Projects to rehabilitate wetlands and floodplains (e.g., Poyang Lake’s farmland-to-wetland conversion) should be executed to enhance the capacity for water retention through vegetation-based soil stabilization, which can reinforce the resilience of the ecosystem. Differentiated performance metrics are essential to strengthen institutional coordination: Ecological indicators (e.g., water conservation) should supersede the GDP in evaluations of protected ecological areas like Tibet, while standards for the efficient use of industrial water should be strictly enforced in high-density economic zones of the lower basin to drive sustainable water use.

4.2. Strategies for Enhancing LES

The Yangtze River Basin traverses China’s three major topographic terraces and exhibits distinct characteristics of the gradient of the LES across its upper, middle, and lower reaches. These variations stem from disparities in geographical conditions, patterns of economic development, and policy responses. Given the basin’s pronounced imbalances in regional development, establishing a precisely targeted and differentiated system of ecological governance, one that accounts for regional specificities, is crucial for harmonizing ecological conservation with economic growth.

The upper Yangtze River region, particularly the Tibet Autonomous Region and Qinghai Province, faces significant challenges to the LES. Tibet’s LES persisted within the critical threshold from 2008 to 2023, while that of Qinghai’s was at the bottom of the rankings, and in a less safe state. Tibet and Qinghai exhibited pronounced obstacles to the driving forces, impact and response subsystems of the LES, which revealed deep issues for the government (Figure 9). As a frontier ethnic region, Tibet struggles with policy-related coordination between ecological protection and people’s livelihood, and this is compounded by delayed policy responses. The region spans 1.23 million square kilometers in the ecologically fragile Qinghai–Tibet Plateau, and its high altitude and low temperature have led to medium- and low-yield farmlands that severely limit agricultural productivity. Agriculture risks exacerbating the degradation of the plateau through the deterioration of its grasslands and compaction of its soil if the use of fertilizers and pesticides remains unregulated [24]. Further challenges include a low population density (3.0 persons/km²), poor transportation infrastructure, constraints imposed by a high altitude, insufficient public awareness, and limited technological resources and talent. They collectively hinder breakthroughs in efforts to improve the ecology. Despite enacting the “Qinghai 14th Five-Year Plan for Ecological Environmental Protection”, Qinghai Province struggles with an inadequate policy, a single path to technical industrial transformation, and a high reliance on traditional industries.

A dual-pronged approach is proposed to address these systemic challenges. First, an integrated iceberg-permafrost-grassland system of protection should be implemented in ecologically sensitive areas (particularly Tibet and Qinghai) by combining the preservation of the cryosphere with technologies to restore alpine grassland to mitigate risks of land degradation. Second, high-standard farmland projects must be prioritized in Yunnan’s and Sichuan’s mountainous and karst terrains, where the fragmentation of the farmland and low productivity persist. These initiatives should focus on the modernization of the infrastructure, improving the soil quality, and optimizing systems for water conservation. Together, these measures will strengthen upstream ecological barriers while ensuring sustainable water resources for the middle and lower reaches [33]. This can help create a synergistic framework for regional ecological security.

The middle region of the Yangtze River basin has maintained a relatively stable LES and transitioned to the state of critical safe over the past 2 years. As it is a vital hub for China’s agricultural and industrial production, intensive farming and industrial activities have imposed significant ecological pressures on the region. Such persistent issues as non-point agricultural source pollution and industrial pollution discharge continue to threaten the ecological stability [47]. Hubei, Hunan, and Jiangxi Provinces have dense populations and extensive agricultural outputs, and encounter pronounced obstacles to their driving forces and state subsystems for LES (Figure 9). Contributing factors include the per capita GDP, effective irrigation coverage, and agriculture/forestry/livestock output per unit area, all of which amplify the ecological risks to it. Despite abundant water resources, these provinces face dual pressures from population growth and the preservation of farmlands. For instance, Hubei’s population density reached 314 persons/km² in 2023, while the Wuhan urban agglomeration exceeded 550 persons/km². Rapid urbanization has led to encroachment by land for construction on farmland, while urban greening efforts lag behind population concentration to further compress the ecological space. Industrial pollution exacerbates these challenges. High-emission zones like Wuhan’s Qingshan Industrial Zone and Yichang Chemical Park substantially contribute to sulfur dioxide emissions, while traditional industries remain slow to modernize. Although current policies for ecological restoration have yielded partial improvements, progress in governance remains insufficient to meet the required demands. To address these issues, it is important to strengthen policies for green development, accelerate large-scale projects for ecological restoration, and enhance cross-regional mechanisms of collaborative governance.

The lower reaches of the Yangtze River Basin have long faced acute pressures to the LES owing to rapid urbanization and industrialization. Economically developed areas like Shanghai, which has the highest population density in China, have a high demand for land for construction that leads to shortages. Urban expansion has fragmented the ecological landscape, and encroached on farmland and forestland to lead to a sharp decline in biodiversity and the degradation of ecosystem service functions [29]. Although the region is rich in water resources, its high population density and demand for water as well as scarcity of good-quality water have led to insufficient resources. Expanding green spaces in highly dense built-up areas remains a challenge. Shanghai’s industrial sulfur dioxide emissions per unit area and use of agricultural films in 2023 were 5.46 and 3.84 times the average of upstream provinces, respectively. The rate of self-sufficiency in local grains was inadequate, and highlights the difficulty in balancing input resources with ecological sustainability. In Jiangsu, the pressure to transform the industrial structure toward high-end sectors, combined with a large population and highly intense industrial activities, has resulted in a heavy burden of pollutant treatment and slow growth in forest coverage. However, the lower region leveraged its strong economic might and technological edge to initiate systematic projects for ecological protection earlier than other parts of the study area. This has led to increased investment in ecological restoration and fostered steady improvements in the LES. Going forward, the region must implement strategies to conserve industrial water, reduce emissions, and replenish ecological water in zones with a high intensity of development while strengthening the supervision of groundwater extraction. It should also adopt innovative models of cross-regional ecological governance that are supported by financial resources and technology transfer to back upstream ecological protection.

4.3. Study Limitations and Future Research

While this study provides a systematic assessment of land ecological security in the Yangtze River Basin, several limitations should be acknowledged. First, the DPSIR-based indicator system, while comprehensive, may not fully capture all relevant aspects of ecological security, particularly some region-specific characteristics in the upper reaches such as grassland degradation and wetland conservation. Second, data quality constraints exist, particularly regarding the per capita disposable income of farmers in Tibet and Qinghai for 2008–2012, which required interpolation. Although the missing data proportion was minimal (0.20%), this may still introduce minor uncertainties. Third, methodological limitations include the entropy weight method’s sensitivity to extreme values and the grey prediction model’s constraints in long-term prediction accuracy. Additionally, the obstacle degree model provides primarily static diagnosis rather than dynamic analysis.

Future research could address these limitations by developing more nuanced indicator systems that better capture regional ecological particularities within the DPSIR framework, incorporating more robust data validation techniques, developing dynamic weighting approaches, and integrating machine learning methods for improved prediction accuracy. The establishment of a standardized land ecological security monitoring framework would significantly enhance comparative studies across different regions and temporal scales.

5. Conclusions

In this study, we developed a system of indicators of land ecological security and used it to evaluate the LES in the Yangtze River Basin based on the DPSIR model, grey prediction model, and obstacle degree model. The key findings can be summarized as follows:

(1)

We observed certain strong correlations among the 28 indicators used. Each indicator encapsulated distinct facets of regional development within the DPSIR, with the economic density (D3) and per capita water resources (S4) having the largest weights. The LES of the provinces along the mainstream of the Yangtze River was mainly influenced by the drivers and state subsystems, while the pressure subsystem had a minor impact. Shanghai had the highest scores for the drivers, impact, and response subsystems, while Tibet had the highest scores for the pressures and state subsystems.

(2)

From 2008 to 2023, most regions exhibited a fluctuating upward trend in terms of the LES, and were mostly in a relatively unsafe or critical safe state. Shanghai and Jiangsu entered the critical safe state in 2013, while the comprehensive score of the Tibet Autonomous Region fluctuated slightly within the range of critical safe. Qinghai had the lowest LES score. The overall score of the basin showed an upward trend, from a relatively unsafe state to a critical safe state. The LES exhibited a multi-level spatial distribution that was “higher in the middle and lower reaches, and lower in the upper reaches” of the basin.

(3)

We predicted that the overall LES score of the basin will continue to rise from 2024 to 2033. Shanghai and Jiangsu will reach a safe state, Qinghai will remain in an unsafe state, while the other regions will maintain a critical safe state. The LES of the basin will continue to improve on the whole, but its level of ecological safety will remain concerning.

(4)

The distribution of the degrees of obstacles was dominated by the driving forces and state subsystems, with the impact subsystem ranking second, while the obstacles posed by the response and pressure subsystems were relatively minor. The per capita GDP (D2), economic density (D3), per capita water resources (S4), ratio of effective irrigation area (I1), outputs per unit area of agriculture, forestry, animal husbandry, and fishery (I3), and rate of forest coverage (R6) influenced the LES of different areas to varying degrees. D3 and S4 were the most critical obstacles to the LES in the Yangtze River Basin.

(5)

The government should adopt a regionally differentiated approach to enhance the LES that addresses spatial heterogeneity across the Yangtze River Basin. The key priorities include strengthening water resource management, enhancing cross-regional collaborative governance, and optimizing economic and urban development within ecological carrying capacities.

In response to the problem of unbalanced development in the Yangtze River Basin, future research should consider the new requirements of the Yangtze River protection strategy, and should focus on building a differentiated system of governance for the land ecology that takes into account regional characteristics, to promote the coordinated optimization of ecological protection and economic development. It is necessary to systematically analyze the land ecosystem of the Yangtze River from multiple dimensions and at different levels: On the one hand, the framework of assessment needs to be improved, and requires the use of such techniques of quantitative analysis as structural equation models, geographic detector models, and geographic perception models to accurately identify the key factors driving the LES. On the other hand, the interactions between influential factors need to be analyzed, and a joint early warning mechanism for the LES needs to be established at the scale of the basin. A combination of theoretical innovation and practical exploration can not only provide a scientific basis for the sustainable development of the Yangtze River Basin but can also be used to maintain the health and stability of the land ecosystem. This can in turn help strategically transform ecological governance from that based on “passive response” to “active prevention and control.”