Rural Tourism and Ecosystem Recovery in the Yangtze River Delta: Spatial Coupling and Influencing Factors

Abstract

As a key driver of the rural revitalization strategy, the uneven development of rural tourism urgently requires resolution. To break through the limitations of traditional rural tourism research that focuses on a single economic dimension, this study innovatively constructs a comprehensive analytical framework integrating multi-dimensional evaluation, coupling measurement, and factor identification to examine the interaction between rural tourism development and ecosystem recovery capacity. Taking the Yangtze River Delta region of China as an empirical case, this paper analyzes the spatial coupling relationship and its associated factors between rural tourism development capacity and ecosystem recovery capacity. The results reveal that: (1) At the socio-economic level, the development of rural tourism in the Yangtze River Delta presents a spatial differentiation, with the southeastern region performing significantly better than the northwestern region, and 60.46% of the areas reaching a moderate level or above; (2) At the ecosystem level, high-value areas of ecosystem recovery capacity (50.28%) are mainly concentrated in the southern part; (3) The overall regional coordination level is relatively low, with 13 regions in the coordination stage (accounting for 35.3% of the total spatial area); (4) Technology and financial investment are the dominant factors associated with the coupling coordination degree, indicating a spatial pattern characterized by “innovation-driven” rather than “resource dependence”. Relying solely on natural background conditions is insufficient to build core advantages; we hypothesize that external interventions such as “capital injection” and “technological support” may serve as potential pathways to improve coordination and facilitate ecological value realization. The findings not only provide a new paradigm for evaluating the development quality of rural tourism, but also establish a complete research chain of “diagnosis-classification-optimization,” providing a scientific basis for formulating regionally differentiated development strategies. This study holds significant theoretical value and practical guiding significance for promoting the sustainable development of rural tourism.

1. Introduction

With the advancement of global urbanization, the imbalance between urban and rural development has become a universal challenge for countries worldwide [1]. In China, continuous urbanization and industrialization have catalyzed transformations in rural areas regarding agricultural production, land use, demographic structure, and industrial composition [2,3]. However, urban-biased development policies have also resulted in issues such as rural development lag and agricultural decline [4]. As a key driver of the rural revitalization strategy, rural tourism has emerged as a new engine for rural economic growth [5]. Statistics show that China has established over 130 national key tourism towns and more than 100,000 leisure agriculture and rural tourism characteristic villages, benefiting 55 million villagers and playing a vital role in boosting rural economies and improving livelihoods [6]. Nevertheless, rural tourism currently faces prominent challenges, including uneven development, ecological degradation, resource depletion, and infrastructure overload [7]. Therefore, scientifically identifying obstacles in the development of rural tourism and clarifying the complex interactions between socio-economic systems and eco-environmental systems [8] have become core issues for promoting high-quality and sustainable rural tourism development. Although this study focuses on the coupling coordination measurement rather than an overarching theoretical mapping, our conceptualization of these interactive dynamics is fundamentally inspired by foundational theories on ecological resilience [9] and the governance of complex interactive systems [10].

Existing research has primarily unfolded from two perspectives: single-dimensional element assessment and multi-dimensional system coupling. At the socio-economic level, scholars have confirmed that economic level, social costs, and infrastructure quality are core factors determining the sustainability of rural tourism [11,12,13]. Among these, the supporting role of the village collective economy [14], as well as factors such as employment, cultural heritage, and benefit distribution [15], are considered directly related to the environmental benefits and economic performance of rural tourism [16]. At the eco-environmental level, research focuses on the balance between “utilization” and “protection.” On one hand, emphasis is placed on transforming ecological processes into tourism experiences through landscape pattern construction and the utilization of carbon sink carriers [17,18]. On the other hand, the fragility and non-renewability of the ecological environment are highlighted [19,20], advocating for environmental carrying capacity as a rigid constraint to implement moderate development and ecological restoration [21,22]. Regarding the interaction between rural tourism systems and external elements, coupling coordination models have been widely applied to analyze synergistic states, covering basic support from public services and “Beautiful Countryside” construction [23,24], industrial integration of agriculture and the digital economy [25,26], and spatio-temporal differentiation strategies of green development and regional integration [27,28].

Despite the rich existing research, two main limitations remain. First, there is a lack of a holistic perspective. Most studies focus on single-system evaluations or bilateral relations between specific industries, rarely employing a coupled system perspective to comprehensively examine the misalignment and interaction/coercive mechanisms between rural tourism development and ecological conditions during development. Second, attribution methods are insufficient. Traditional methods often overlook spatial heterogeneity and struggle to overcome subjective interference caused by manual parameter settings when handling continuous variables, thereby affecting the accuracy of influencing factor identification. Consequently, this study innovatively constructs an “Evaluation-Coupling-Association” comprehensive framework: (1) establishing a dual-perspective evaluation system integrating rural tourism development capacity and ecosystem recovery capacity; (2) utilizing the Coupling Coordination Degree Model (CCDM) to measure the synergistic level between social and ecological systems; and (3) introducing the Optimal Parameters Geographic Detector (OPGD) to accurately identify key factors and their interactive effects influencing coupling coordination. By deeply examining the coupled interactions alongside attribution methods, this study aims to break through the limitations of single-system assessment and enrich the quantitative paradigm for rural tourism complex systems. Simultaneously, by precisely locating the key factors spatially associated with system coupling coordination, it provides scientific decision-making support for resolving the “economic growth-ecological protection” dichotomy and formulating differentiated regulation strategies.

Specifically, this study seeks to address the following three scientific questions: (1) How can we effectively measure the multidimensional development quality of rural tourism and the recovery capacity of the local ecosystem? (2) What is the current state of spatial coupling and coordination between these two subsystems in the Yangtze River Delta? (3) Which core factors and interaction mechanisms drive the regional differentiation of this coupling coordination? By addressing these questions, we aim to provide a more robust quantitative paradigm and scientific guidance for the sustainable transition of rural tourism.

2. Materials and Methods

2.1. Theoretical Model

Rural tourism destinations are not merely single industrial spaces, but rather complex interactive areas driven by human socio-economic activities and constrained by the natural eco-environment [29]. Complex interactive relationships exist between the two: the over-expansion of the social subsystem may coerce the stability of the ecosystem, while a sound eco-environment, in turn, enhances tourism attractiveness and economic value [30]. Therefore, realizing the high-quality development of rural tourism essentially involves seeking the optimal dynamic balance between socio-economic benefit growth and eco-environmental carrying capacity [31].

However, traditional evaluation systems for rural tourism often suffer from the limitation of fragmented perspectives. On one hand, existing models tend to focus on limited single-dimensional evaluations, concentrating solely on the development potential of tourism resources or the unidirectional monitoring of environmental quality. They lack a systematic consideration that treats the two as an organic whole, making it difficult to reveal the interactive coercive mechanisms between “development and protection.” On the other hand, when exploring the driving forces affecting the coordinated development of the system, traditional frameworks often stop at status descriptions and lack precise quantitative attribution methods. This makes it difficult to fundamentally identify the key obstacles restricting the synergistic evolution of the coupled system.

Breaking through the limitations of traditional single perspectives, this study innovatively constructs a theoretical analytical framework for the high-quality development of rural tourism, integrating “Comprehensive Evaluation—Coupling Measurement—Attribution Analysis” based on the interactive logic of human-environment coupling (Figure 1). This framework regards rural tourism as a complex system intertwined with human activities and the natural environment, logically encompassing three progressive levels: (1) System Deconstruction: The model divides rural tourism destinations into a social subsystem (representing rural tourism development capacity) and an ecological subsystem (representing ecosystem recovery capacity), clarifying the system’s background conditions through a multi-dimensional indicator system. (2) Relationship Measurement: The Coupling Coordination Degree Model (CCDM) is introduced to quantitatively measure the interaction intensity and coordination level between the social and ecological systems, thus judging their developmental stages. (3) Association Analysis: Taking the coupling coordination degree as the dependent variable, an influencing factor system is constructed, encompassing four dimensions: material production foundation, human resources and investment support, innovative technology drive, and surface background conditions. The Optimal Parameters Geographic Detector (OPGD) is applied to precisely identify the key factors spatially associated with system coordination. This model achieves a logical closed-loop from “state evaluation” to “mechanism diagnosis,” aiming to provide scientific decision-making support for the sustainable management of rural tourism.

2.2. Study Area

This study takes the Yangtze River Delta (YRD) region in eastern China as the study area (Figure 2). This region encompasses the municipality of Shanghai and the three provinces of Jiangsu, Zhejiang, and Anhui, covering a total area of 358,000 square kilometers [32]. As a pioneering demonstration zone for rural tourism development in China, the YRD region exhibits three typical characteristics: (1) Superior resource endowment. Over 3100 villages are distributed across the region (accounting for 9.4% of the national total) [33], boasting rich historical and cultural heritage as well as picturesque natural landscapes [34]; (2) Well-developed infrastructure. A highly developed tourism transportation network and a mature tourist source market system have been established [35]; (3) Remarkable development outcomes. In recent years, the tourism industry has shown a rapid growth trend, making prominent contributions to regional economic development [36]. As a pilot zone for the implementation of the national rural revitalization strategy [37], the development level of rural tourism in the YRD region ranks among the top nationwide [23], and its developmental experience holds significant demonstration value for other regions [38]. It should be explicitly acknowledged that while this study is conducted at the prefecture-level city scale to ensure data consistency and macro-policy alignment across the 41 cities, this spatial aggregation inherently involves the risk of the Modifiable Areal Unit Problem (MAUP). Rural tourism activities and ecosystem processes often operate at finer scales, such as townships or villages. However, due to the current lack of uniform and high-precision socio-economic statistical data at the county level for the entire YRD region, the prefecture-level scale remains the most viable unit for revealing regional differentiation patterns in this trans-provincial study.

2.3. Data Sources

This study employs multi-source data to construct the evaluation system, with data sources categorized into two main groups: the social subsystem and the ecological subsystem. Regarding the social subsystem, rural tourism popularity data integrates rural ratings (https://m.dianping.com/) and review volumes (https://m.dianping.com/) from the Dianping App, as well as rural popularity data from the Ctrip App (https://m.ctrip.com/). Data on the richness of tourism resources encompasses the spatial distribution of 5 batches of national intangible cultural heritage (https://www.geodoi.ac.cn/WebCn/Default.aspx, http://www.geodoi.ac.cn/WebCn/doi.aspx?Id=1925, accessed on 9 April 2026), the distribution of national key cultural relics (http://gl.ncha.gov.cn/#/Home, accessed on 9 April 2026), 4 batches of the national key rural tourism villages list (https://www.mct.gov.cn//wlbphone/, accessed on 9 April 2026), 6 batches of the traditional villages list (https://www.mohurd.gov.cn/xinwen/, accessed on 9 April 2026), and POI (Point of Interest) data for tourism-related commercial facilities from Amap (https://m.amap.com). Public service development data includes POI data for life, transportation, and medical facilities from Amap (https://m.amap.com) and road network data from OpenStreetMap (https://www.openstreetmap.org/). The data on the scale of the tourism industry is derived from the annual work reports of local tourism bureaus. The data collection for the ecological subsystem takes into account both scientific validity and timeliness: the ecological resources dimension utilizes the 2023 China 30 m resolution land cover dataset (https://zenodo.org/records/12779975, accessed on 9 April 2026); the ecological quality dimension integrates data from the Global Biodiversity Information Facility (GBIF) network (https://www.gbif.org/), the high-resolution near-surface air pollutant dataset from the National Tibetan Plateau Data Center (https://data.tpdc.ac.cn/zh-hans/data/6168e75d-93ab-4e4a-b7ff-33152e49d0bf, accessed on 9 April 2026), and the Azure Water Index report from the Institute of Public and Environmental Affairs (https://www.ipe.org.cn/); the ecological resilience dimension combines the aforementioned land cover dataset (https://zenodo.org/records/12779975, accessed on 9 April 2026) with construction land data from the China Urban Construction Statistical Yearbook (https://www.stats.gov.cn/sj/ndsj/).

Furthermore, the study constructs an attribution indicator system. The theoretical basis for selecting indicators in each dimension and their specific data sources is explained as follows: First, in the dimension of material production foundation, according to the sustainable livelihood theory, stable capital accumulation is the endogenous driving force for enhancing system resilience [19]. Therefore, this study selects the per capita disposable income of rural residents, comprehensive grain production capacity, and per capita total power of agricultural machinery for characterization, with relevant data compiled from the China Rural Statistical Yearbook (https://www.zgtjcbs.com/) and the China Labour Statistical Yearbook (https://www.zgtjcbs.com/). Second, in the dimension of human resources and investment support, considering that high-quality human capital and the supply of public goods can significantly reduce transaction costs and promote positive feedback [37], the study selects four indicators: demographic structure, local public budget expenditure, rural greening rate, and the average years of schooling for rural residents. Among them, demographic structure data originates from the Seventh National Population Census, while data for the remaining indicators come from the China Rural Statistical Yearbook. Third, in the dimension of innovative technology drive, based on the technology empowerment theory, digital technology is regarded as a new dynamic momentum to break the urban-rural dual barrier [39]. Accordingly, this study selects the level of rural informatization, cable TV coverage rate, and government investment in science and technology to depict technology penetration, with data mainly obtained from the China Rural Statistical Yearbook, China Labour Statistical Yearbook, and China Statistical Yearbook (https://www.stats.gov.cn/sj/ndsj/, accessed on 9 April 2026). Finally, in the dimension of surface background conditions, according to the ecological carrying capacity theory, high-quality matrix conditions determine the physical threshold for system coordination [40]. This study selects the area of soil erosion, soil organic matter content, and green space development rate for quantification, with relevant raster data sourced from the Resource and Environment Science and Data Center of the Chinese Academy of Sciences (CAS) (https://www.resdc.cn/Default.aspx) and the National Tibetan Plateau Data Center (https://data.tpdc.ac.cn), respectively.

2.4. Social-Ecological” Evaluation Framework for Rural Tourism

2.4.1. Evaluation Method for the Social Subsystem of Rural Tourism

Based on the goals of the rural revitalization strategy and existing research outcomes [41], this study constructs a four-dimensional evaluation framework comprising 13 indicators (

Table 1. Evaluation Index System for Rural Tourism Development Capacity.

Criterion	Indicator	Principles of Computing	Description of Variables	Reference
Rural Tourism Popularity	X1: Rating of key tourist villages	$K_i=\left({\displaystyle \frac{1}{n_i}}{\displaystyle \sum _{j=1}^{n_i}}{V}_{ij}\right)\times {\displaystyle \frac{n_i}{N}}$	Ki: Comprehensive rating of key tourist villages in the i-th city; Vij: Individual rating of the j-th key tourist village in the i-th city; ni: Number of key tourist villages in the i-th city; N: Total number of key tourist villages in the entire YRD region.	[44]
	X2: Review volume of key tourist villages	$M_i=\left({\displaystyle \frac{1}{n_i}}{\displaystyle \sum _{j=1}^{n_i}}{C}_{ij}\right)\times {\displaystyle \frac{n_i}{N}}$	Mi: Comprehensive review volume of key tourist villages in the i-th city; Cij: Individual review count of the j-th key tourist village in the i-th city; ni: Number of key tourist villages in the i-th city; N: Total number of key tourist villages in the entire YRD region.	[45]
	X3: Popularity index of key tourist villages	$H_i=\left({\displaystyle \frac{1}{n_i}}{\displaystyle \sum _{j=1}^{n_i}}{h}_{ij}\right)\times {\displaystyle \frac{n_i}{N}}$	Hi: Comprehensive popularity index of key tourist villages in the i-th city; Hij: Individual popularity value of the j-th key tourist village in the i-th city; ni: Number of key tourist villages in the i-th city; N: Total number of key tourist villages in the entire YRD region.	[45]
Richness of Tourism Resources	X4: Distribution of intangible cultural heritage	${Fy}_i={\displaystyle \frac{F_i}{S_i}}$	Fyi: Spatial density of intangible cultural heritage in the i-th city; Fi: Total number of intangible cultural heritage projects in the i-th city; Si: Administrative area of the i-th city.	[46]
	X5: Distribution of key cultural relics	${Wu}_i={\displaystyle \frac{W_i}{S_i}}$	Wui: Spatial density of key cultural relics in the i-th city; Wi: Total number of key cultural relics in the i-th city; Si: Administrative area of the i-th city.	[47]
	X6: Distribution of key villages	${Cz}_i={\displaystyle \frac{Z_i}{S_i}}$	Czi: Spatial density of key villages in the i-th city; Zi: Total number of key villages in the i-th city; Si: Administrative area of the i-th city.	[48]
	X7: Distribution of traditional villages	${Ct}_i={\displaystyle \frac{T_i}{S_i}}$	Cti: Spatial density of traditional villages in the i-th city; Ti: Total number of traditional villages in the i-th city; Si: Administrative area of the i-th city.	[48]
	X8: Distribution of tourism-related commercial and amusement facilities	${Ls}_i={\displaystyle \frac{B_i}{S_i}}$	Lsi: Spatial density of tourism-related commercial and amusement facilities in the i-th city; Bi: Total number of tourism-related commercial and amusement facilities in the i-th city; Si: Administrative area of the i-th city.	[49]
Public Service Development	X9: Distribution of public service facilities (living, transportation, medical)	${Gf}_i={\displaystyle \frac{G_i}{S_i}}$	Gfi: Spatial density of public service facilities in the i-th city; Gi: Total number of public service facilities in the i-th city; Si: Administrative area of the i-th city.	[49]
	X10: Road network accessibility	${Kd}_i={\displaystyle \frac{1}{S_i}}{\displaystyle \sum _{p=1}^4}\left(l_{ip}\times {\omega }_p\right)$	Kdi: Road network accessibility index of the i-th city; Si: Administrative area of the i-th city; Lip: Total length of the p-th type of road in the i-th city; ωp: Weight coefficient of the p-th type of road; p: Iteration variable for road types (p = 1, 2, 3, 4), corresponding to: p = 1: Elevated and expressways; p = 2: Main roads; p = 3: Secondary roads; p = 4: Branch roads.	[50]
Scale of the Tourism Industry	X11: Tourism revenue	$E_i={\displaystyle \frac{1}{O}}{\displaystyle \sum _{\gamma =1}^O}{Sr}_{i\gamma }$	Ei: Comprehensive level of tourism revenue in the i-th city; Sriγ: Total tourism revenue of the i-th city in the γ-th year; γ: Year iteration variable, corresponding to 2021, 2022, 2023; O: Statistical year span, where O = 3.	[51]
	X12: Proportion of tourism in the tertiary industry	$P_i={\displaystyle \frac{\sum _{\gamma =1}^O{Sr}_{i\gamma }}{\sum _{\gamma =1}^O{Bz}_{i\gamma }}}\times 100\%$	Pi: Proportion of tourism in the tertiary industry of the i-th city; Sriγ: Total tourism revenue of the i-th city in the γ-th year; Bziγ: Gross Domestic Product (GDP) of the tertiary industry of the i-th city in the γ-th year; γ: Year iteration variable, corresponding to 2021, 2022, 2023; O: Statistical year span, where O = 3.	[51]
	X13: Number of tourists	$A_i={\displaystyle \frac{1}{O}}{\displaystyle \sum _{\gamma =1}^O}{\alpha }_{i\gamma }$	Ai: Average total number of received tourists in the i-th city; αiγ: Total number of tourist arrivals in the i-th city in the γ-th year; γ: Year iteration variable, corresponding to 2021, 2022, 2023; O: Statistical year span, where O = 3.	[51]

) to systematically measure rural tourism development capacity. Simultaneously, the Projection Pursuit Model (PPM) is employed to identify indicator weights and calculate the rural tourism development index. This method is a data-driven multi-factor analysis tool [42]. The core principle of PPM is to reduce high-dimensional data to a low-dimensional subspace by finding the optimal projection direction. During this process, the projection function is optimized by setting constraint conditions [43]. The optimal projection vector obtained from solving the model serves a dual function: on the one hand, its component values objectively reflect the relative importance weights of each evaluation indicator (

Table 2. Direction Vectors of Each Evaluation Indicator.

Indicator	X1	X2	X3	X4	X5	X6	X7	X8	X9	X10	X11	X12	X13
Vector Value	0.698	0.161	0.062	0.357	0.231	0.107	0.174	0.076	0.012	0.333	0.333	0.006	0.184

); on the other hand, the projected values can serve as a quantitative representation of rural tourism development capacity. This method achieves effective dimensionality reduction and comprehensive evaluation of high-dimensional data while maintaining the maximum inter-class variance of the original data. Compared to traditional weighting methods, PPM is calculated entirely based on the intrinsic characteristics of the data, thereby avoiding the subjectivity of manual weighting. For the specific calculation process, please refer to [4].

The parameters of the genetic algorithm were determined based on extensive pre-experiments and established empirical practices in spatial multi-criteria evaluation [52]. Specifically, the initial population size was set to 300 to ensure sufficient diversity in the solution space without imposing excessive computational burden. The crossover probability (0.9) and mutation probability (0.05) were carefully chosen to strike a balance between global exploration and local exploitation, preventing the algorithm from falling into local optima. The maximum number of iterations was set to 500. Furthermore, a parameter sensitivity analysis was performed by reasonably varying the crossover probability (ranging from 0.8 to 0.95) and mutation probability (ranging from 0.01 to 0.1). The results indicated that the fluctuation of the derived indicator weights remained within a negligible margin (less than 2%), demonstrating the high robustness and stability of the model’s outputs under the selected parameter configuration.

2.4.2. Evaluation Method for the Ecological Subsystem of Rural Tourism

This study employs a multi-dimensional indicator system (

Table 3. Evaluation Index System for Rural Tourism Ecological Resilience.

Criterion	Indicator	Principles of Computing	Description of Variables	Reference
Ecological Resources	Proportion of cultivated land	${Gd}_i={\displaystyle \frac{G_i}{S_i}}$	Gdi: Proportion of cultivated land in the i-th region; Gi: Area of cultivated land in the i-th region; Si: Administrative area of the i-th region.	[4]
	Proportion of forest land	${Ld}_i={\displaystyle \frac{L_i}{S_i}}$	Ldi: Proportion of forest land in the i-th region; Li: Area of forest land in the i-th region; Si: Administrative area of the i-th region.	[4]
	Proportion of grassland	${Cd}_i={\displaystyle \frac{C_i}{S_i}}$	Cdi: Proportion of grassland in the i-th region; Ci: Area of grassland in the i-th region; Si: Administrative area of the i-th region.	[4]
	Proportion of water bodies	${Sd}_i={\displaystyle \frac{Q_i}{S_i}}$	Sdi: Proportion of water bodies in the i-th region; Qi: Area of water bodies in the i-th region; Si: Administrative area of the i-th region.	[4]
Ecological quality	Biodiversity	$B_i={\displaystyle \frac{1}{{\mu }_i}}{\displaystyle \sum _{j=1}^{{\mu }_i}}{\beta }_{ij}$	Bi: Biodiversity index of the i-th city; βij: Biodiversity value of the j-th raster cell within the administrative boundary of the i-th city; μi: Total number of valid raster cells within the i-th city.	[54]
	PM2.5	${PM}_i={\displaystyle \frac{1}{{\mu }_i}}{\displaystyle \sum _{j=1}^{{\mu }_i}}{H}_{ij}$	PMi: PM2.5 index of the i-th city; Hij: PM2.5 value of the j-th raster cell within the administrative boundary of the i-th city; μi: Total number of valid raster cells within the i-th city.	[55]
	Water quality monitoring index	${SC}_i={\displaystyle \frac{1}{O}}{\displaystyle \sum _{\gamma =1}^O}{JC}_{i\gamma }$	SCi: Comprehensive water quality monitoring index of the i-th city; JCiγ: Water quality monitoring index of the i-th city in the γ-th year; γ: Year iteration variable, corresponding to 2021, 2022, 2023; O: Statistical year span, where O = 3.	[56]
Ecological Resilience	Ecosystem adaptability	$\mathrm{Sy}=0.25\times \mathrm{PD}+0.25\times \mathrm{LPI}+0.25\times \mathrm{SHDI}+0.25\times \mathrm{SHEI}$	Sy: Calculation result of ecosystem adaptability for each research unit; PD: Patch Density; LPI: Largest Patch Index; SHDI: Shannon’s Diversity Index; SHEI: Shannon’s Evenness Index.	[4]
	Ecosystem Recoverability	$\mathit{Sh}={\displaystyle \sum _{\mathit{i}=1}^{\mathit{n}}}{A}_i\times {\mathit{SX}}_{\mathit{i}}$	Sh: Ecosystem recovery capacity; Ai: Area of different land use types; SXi: Relative coefficient of different land use types.	[4]

) to comprehensively evaluate the ecosystem recovery capacity of rural tourism, a concept defined as the capacity of an ecosystem to return to its original state or achieve a higher level of vitality after experiencing disturbances [53].

As a fundamental constituent element of the ecosystem, the sustainable utilization of ecological resources is a prerequisite for ensuring the sustainable development of the tourism industry [57]. This study conducts a spatial quantitative assessment based on the ArcGIS10.8 platform. Specifically, by extracting four key elements—forest land, grassland, water bodies, and cultivated land—and calculating their proportions relative to the total area of the research unit, the land resource coverage index is constructed. The principle is as follows:

$$\mathrm{Sz} = \mathrm{Gd} \times 0.25 + \mathrm{Ld} \times 0.25 + \mathrm{Cd} \times 0.25 + \mathrm{Sd} \times 0.25$$

(1)

where Gd is the proportion of cultivated land; Ld is the proportion of forest land; Cd is the proportion of grassland; Wd is the proportion of water bodies; and Sz represents the calculation result of the ecological resources dimension for each research unit.

While Equation (1) initially assigns equal weights (0.25) to the four land cover types to assess baseline spatial coverage without imposing subjective prior bias, we acknowledge the substantial ecological differences in their recovery capacities. To address this, a sensitivity analysis was conducted by applying alternative variance-based weights and expert-derived ecological weights (e.g., assigning higher weights of 0.4 to forest and water bodies, and 0.1 to cultivated land and grassland). A Pearson correlation analysis between the comprehensive indices calculated under the equal-weighting scheme and the alternative weighted schemes yielded correlation coefficients exceeding 0.95 (p < 0.01). This confirms that the equal-weighting structure does not materially alter the spatial patterns and final conclusions of the ecological resource evaluation. Therefore, the baseline weighting is retained for structural simplicity, supported by its demonstrated analytical robustness.

The development of evaluation indicators for ecological quality is challenging and complex, and it currently remains in the exploratory stage [31]. The development of evaluation indicators for ecological quality is challenging and complex, and it currently remains in the exploratory stage:

$$\mathrm{Sl} = {\displaystyle \frac{\mathrm{B}+\mathrm{Pm}+\mathrm{SC}}{3}}$$

(2)

where Sl represents the calculation result of the ecological quality dimension for each research unit. Among them, B is the biodiversity index; Pm is the PM2.5 index; and SC is the water quality monitoring index. Given the different units involved, all three indices undergo a normalization process.

As a key element for the sustainable development of rural tourism, ecological resilience is primarily evaluated from two dimensions: ecosystem adaptability and ecosystem recoverability [58]. Among these, ecosystem adaptability reflects the system’s potential to maintain structural stability under external disturbances [59]. Based on the FRAGSTATS 4.2 platform, we select four key landscape indices (PD, LPI, SHDI, and SHEI) that characterize fragmentation, stability, diversity, and evenness for the evaluation and measurement. Ecosystem recoverability, on the other hand, characterizes the system’s self-repair level after being damaged [60,61]. Drawing on the methods of Colding et al. (2007) [62] and Peng et al. (2015), we calculate this using an area-weighting method based on the differences in ecological elasticity among different land-use types (e.g., forest land, water bodies, and construction land) (

Table 4. Evaluation Index System for Rural Tourism Ecological Resilience.

LUT	Cultivated Land	Forest Land	Grassland	Water Bodies	Construction Land
Relative Coefficient	0.3	0.6	0.8	0.8	0.2

) [63]. Ultimately, the comprehensive evaluation value of ecological resilience is obtained by integrating the two aforementioned dimensions.

$$\mathrm{Sr}={\displaystyle \frac{\mathrm{Sy}+\mathrm{Sh}}{2}}$$

(3)

where Sr represents the calculation result of the ecological resilience dimension for each research unit; Sy represents the calculation result of ecosystem adaptability; and Sh represents ecosystem recoverability. Based on the above, the calculation principle for ecosystem recovery capacity can be expressed as:

$$\mathrm{STF} = \sqrt[3]{\mathrm{Sz}\times \mathrm{Sl}\times \mathrm{Sr}}$$

(4)

2.5. Coupling Coordination Degree Model

The Coupling Coordination Degree Model (CCDM) is utilized to evaluate the coordinated development between different systems or among elements within a system. Building upon normative methods such as the Projection Pursuit Model (PPM), this model assesses the degree of coordinated development by calculating the coupling coordination degree between the systems, namely the Coupled System of rural tourism. For the specific calculation procedure, please refer to [64]. The resulting values of the model fall within the range of (0, 1), where a value closer to 1 indicates superior coupling coordination. In accordance with the actual model results and referring to established classification standards from existing research, the coupling coordination degree is categorized into 10 intervals and 3 developmental stages (

Table 5. Classification Criteria for Coupling Coordination Degree.

Developmental Stage	Incoordination Stage			Transition Stage			Coordination Stage
Coordination Level	Extreme Incoordination	Serious Incoordination	Moderate Incoordination	Mild Incoordination	On the Verge of Incoordination	Barely Coordination	Primary Coordination	Intermediate Coordination	Good Coordination	Excellent Coordination
CCD Range	(0.0~ 0.1)	[0.1~ 0.2)	[0.2~ 0.3)	[0.3~ 0.4)	[0.4~ 0.5)	[0.5~ 0.6)	[0.6~ 0.7)	[0.7~ 0.8)	[0.8~ 0.9)	[0.9~ 1.0)

) [65].

2.6. Optimal Parameters Geographic Detector (OPGD) Model

The Geographic Detector (Geodetector) is a statistical method capable of revealing the impact intensity of influencing factors on the formation process of spatial heterogeneity, and it is widely utilized for driving force and factor analyses [66]. However, the process of discretizing spatial data is susceptible to the influence of subjective factors, which may lead to insufficiently significant detection results [67]. Consequently, this study selects the Optimal Parameters Geographic Detector (OPGD) as the analytical tool. This model incorporates a parameter optimization module to automatically select the optimal parameter combination through algorithms. Given the large geographical span of the Yangtze River Delta (YRD) region, multiple grid scales are employed during sampling to conduct significance testing on the q-statistic. The scale with the most significant q-statistic is subsequently selected as the optimal scale to explore the driving mechanism. For the specific calculation procedure, please refer to [68].

3. Results and Analysis

3.1. Evaluation Results of the Coupled System of Rural Tourism

As shown in Figure 3a, the study finds that the rural tourism development capacity in the Yangtze River Delta (YRD) region presents a spatial pattern of “high in the southeast and low in the northwest.” Southeastern cities such as Shanghai, Suzhou, and Hangzhou rank in the first (1.173–2.424) and second tiers (0.845–1.173) by virtue of their outstanding rural tourism development capacity, accounting for 13.4% of the total area. A group of 19 cities, including Jinhua and Jiaxing, constitutes the third to sixth tiers (0.346–0.845), exhibiting significant characteristics of unbalanced development among internal dimensions. Conversely, 17 cities in the northwestern part are at a distinct disadvantage in the two key dimensions of resource richness and market popularity due to insufficient resource excavation or a lack of brand construction, resulting in a relatively lagging overall development level (0.103–0.346).

Figure 3b shows that the spatial differentiation pattern of the ecosystem recovery capacity of rural tourism exhibits an obvious “high in the south and low in the north” characteristic. Areas with recovery capacity in the top four tiers are concentrated in the southern YRD, accounting for 50.28% of the total area, demonstrating significant agglomeration characteristics. Typical representatives, such as Hangzhou and Chizhou, demonstrate robust ecosystem stability. Moderate recovery capacity areas (accounting for 16.99%) are distributed as a transition zone around the periphery of the higher-tier areas, with prominent unbalanced development across internal dimensions. Taking Chuzhou as an example, although its ecological resources and resilience perform adequately, its ecological quality indicators lag significantly. This structural imbalance restricts the overall level of recovery capacity. Areas with low and the lowest recovery capacity (accounting for 27.82% in total) are mainly distributed in the northern fringe zones, typically exemplified by Fuyang and other places. The ecosystem fragility in these regions is notable, making them more susceptible to irreversible damage under anthropogenic disturbances such as rural tourism development. Targeted ecological restoration and protection measures are urgently needed.

To ensure the scientific rigor of our assessment, the ecosystem recovery capacity index was validated against independent remote sensing data. Specifically, we performed a spatial correlation analysis between our calculated recovery capacity scores and the 2023 MODIS Normalized Difference Vegetation Index (NDVI) for the Yangtze River Delta. The results revealed a significant positive correlation (Pearson’s r = 0.74, p < 0.01), confirming that areas identified as having high recovery capacity (such as the forested regions of southern Zhejiang and western Anhui) align closely with objective high-vegetation-density observations. This cross-validation with independent datasets provides empirical support for the robustness of our results.

3.2. Coupling Coordination Relationship of the Coupled System of Rural Tourism

Based on the calculation of the coupling coordination degree of the Coupled System of rural tourism, the values for the 41 regions range from 0.034 to 0.824, with an average value of 0.475, indicating an overall state on the verge of incoordination. Specifically, 13 regions have a coupling coordination degree of the coupled system greater than or equal to 0.6, positioning them in the coordination stage, which accounts for 35.3% of the total spatial area. Among them, four regions—Hangzhou, Suzhou, Shanghai, and Huangshan—have reached the intermediate coordination type or above, accounting for 12.1% of the total spatial area, and are spatially distributed mainly along the Yangtze River in the YRD region. There are 19 regions in the transition stage, accounting for 49.03%, which are relatively densely distributed in the central and northern parts of the YRD. Regions in the incoordination stage total 8, accounting for 13.1%, mainly located in the northwestern boundary area of the YRD, with only Taizhou and Tongling sporadically distributed in the central region (Figure 4).

Further dissecting the internal structural characteristics of the regions in the transition and incoordination stages, based on the relative lagging degree of the social and ecological systems, they can be identified as three typical types. The first is the society-lagging type, typically represented by Tongling and Suzhou (Anhui). Their extreme incoordination state primarily stems from the unidirectional constraint of insufficient development in the social subsystem. The second is the ecology-damaged type, such as Taizhou. Although its economic development is acceptable, the significant lag of the ecological subsystem has become the key bottleneck leading to its extreme incoordination. The third is the overall unbalanced type, taking Bozhou as an example. It faces a simultaneous collapse of both the social and ecological systems, presenting a serious incoordination state characterized by “double lows.”

3.3. Associated Factors of the Coupling Coordination of the Coupled System

3.3.1. Single-Factor Detection Results

Table 6. Results of Single-Factor Detection.

Dimension	Impact Factor	q-Statistic	p-Value
Material production foundation	Per capita disposable income of rural residents	0.2506	0.007
	Comprehensive grain production capacity	0.2371	0.004
	Per capita total power of agricultural machinery	0.3037	0.002
Human resources and investment support	Demographic structure	0.1290	0.021
	Local public budget expenditure	0.4414	0.000
	Rural greening rate	0.1828	0.012
	Average years of schooling for rural residents	0.2266	0.003
Innovative technology drive	Level of rural informatization	0.3129	0.004
	Cable TV coverage rate	0.1983	0.004
	Government investment in science and technology	0.4606	0.000
Surface background conditions	Area of soil erosion	0.2096	0.023
	Green space development rate	0.0540	0.037
	Soil organic matter content	0.3055	0.000

demonstrates the single-factor detection results, indicating that the coupling coordination degree of the Coupled System of rural tourism in the Yangtze River Delta region is subject to the compound influence of multiple dimensions: material production foundation, human resources and investment support, innovative technology drive, and surface background conditions. First, government investment in science and technology (0.4606) and local public budget expenditure (0.4414) possess the strongest explanatory power, indicating that the injection of capital and technology under administrative guidance is the dominant force initiating system coupling [65]. The level of rural informatization and soil organic matter content follow closely. The former breaks the geographical shielding effect of rural tourism by constructing digital networks, significantly reducing the search and transaction costs of the tourist source market [69]. The latter, as the core matrix of agricultural landscapes and ecosystems, determines the environmental carrying threshold for industrial development [23]. Respectively, they constitute the technological channel for market connection and the material basis for ecological carrying capacity, jointly establishing their core driving positions. Second, the per capita total power of agricultural machinery, the per capita disposable income of rural residents, and the comprehensive grain production capacity exhibit strong explanatory power, confirming that the level of agricultural modernization and the consumption capacity of residents serve as the endogenous support for maintaining system operation [70]. The average years of schooling are directly related to the human resource quality of rural service supply, ensuring the effective operation of the social subsystem [71]. Finally, the area of soil erosion and demographic structure are located in the third tier, primarily functioning as the bottom-line constraints of the system. The green space development rate (0.0540) has the minimum impact, demonstrating that natural resources possess passive attributes and are difficult to independently transform into development momentum without the intervention of capital and technology [72]. Therefore, it is not the key variable restricting regional differentiation at the current stage.

3.3.2. Interaction Detection Results

Based on the OPGD model, interaction detection of the driving factors was conducted to further analyze the impact of factor interactions on the coupling coordination degree of the Coupled System of rural tourism. The results indicate (Figure 5) that the interaction effects all manifest as either bivariate enhancement (EB) or non-linear enhancement (EN), with no independent or weakening relationships observed. This demonstrates that the development level of rural tourism in the Yangtze River Delta region exhibits complex coupling effects at both the social and ecological levels. Its spatial differentiation is not dominated by a single element but is the result of the synergistic action of multiple factors.

From the perspective of interaction intensity, local public budget expenditure and government investment in science and technology demonstrate the strongest interaction control, with their interaction values with other factors generally remaining at high levels (indicated by the dark-colored regions). The highest interaction explanatory power (0.772) occurs in the interaction between local public budget expenditure cap and green space development rate, indicating that fiscal support is strongly spatially associated with the utilization of rural green space resources and transforming them into advantages for tourism development [73]. This is followed by the interaction between government investment in science and technology cap green space development rate (q = 0.720) and government investment in science and technology cap area of soil erosion (q = 0.693). This reveals the significant amplification effect of technological empowerment on ecological governance and resource development [74]. Notably, the green space development rate, which exhibited the weakest explanatory power in the single-factor detection, demonstrates astonishing “explosive power” in the interaction detection. When it interacts with factors such as local public budget expenditure, government investment in science and technology, or soil organic matter content, its explanatory power multiplies non-linearly. This finding profoundly reveals the inherent mechanism of “passive resources and active capital”. Once green space resources are combined with elements such as capital and technology, their carrying capacity and landscape value as an ecological foundation are fully activated, becoming a core engine for promoting the high-quality development of rural tourism.

4. Discussion

4.1. Spatial Characteristics of Associated Factors

This study confirms that the spatial pattern of rural tourism in the Yangtze River Delta (YRD) region exhibits a strong association with capital and innovation rather than mere resource reliance. The dominant position of fiscal and technological factors has surpassed natural resource endowment, revealing the “non-natural determinism” characteristic of system evolution: high-quality ecology is merely a necessary condition for development, while “external proactive factors” represented by capital and technology are the sufficient conditions to break low-level coupling and propel systematic transition. On the one hand, the significance of fiscal input validates the public goods attribute of rural tourism. Regarding ecological governance and infrastructure with long return on investment (ROI) cycles, continuous government guidance effectively compensates for market failures, establishing a stable baseline for the coupling coordination of the system [75]. On the other hand, the rise in technological factors responds to the reshaping of traditional service industries by the digital economy. It indicates that the driving force for development has shifted from simple agricultural scale expansion to innovation-driven growth. Through digital marketing and efficient resource management, the industrial chain and value chain of rural tourism have been fundamentally reshaped [76].

4.2. Factor Interaction and Value Transformation of Natural Resources

The interaction detection results profoundly reveal the spatial synergistic association between ecological resources and tourism economic value. The leapfrog enhancement of the green space development rate—from weak explanatory power under single-factor action to a core driver after interacting with capital and technology—confirms that natural resources exhibit significant passive asset characteristics in system coupling. A high-quality natural foundation is only a static stock of development; without the intervention of external proactive factors, it is often difficult to independently transform it into a driving force for system development [77]. In this process, fiscal funds and technological input serve as critical value leverages: fiscal support builds the “hard connectivity” of resource development by improving transportation and service facilities, breaking down the physical barriers to tourist inflow [78]; meanwhile, technological input achieves “soft empowerment” through ecological restoration and digital experiences, enhancing the attractiveness and carrying capacity of static resources [79]. This synergistic mechanism indicates that excellent ecological environments do not automatically translate into economic wealth. The effective intervention of “capital + technology” is an indispensable prerequisite for crossing the resource transformation.

4.3. Differentiated Governance Strategies

Given the spatial differentiation and interactive characteristics of the driving factors, this study proposes a multi-dimensional governance pathway characterized by “categorized activation, bottom-line control, and region-wide synergy.” First, for society-lagging type regions, the core lies in implementing factor injection. Relying on the dominant driving forces of fiscal and technological factors, public fund guidance and digital infrastructure construction should be increased. Digital means must be utilized to break market barriers and activate dormant ecological assets. Second, for ecology-damaged type regions, the focus should shift to resilience restoration. The rigid constraint status of natural elements must be established, restricting the disorderly expansion of construction land. The fiscal focus should pivot towards soil amelioration and ecological compensation, thereby enhancing environmental carrying capacity by constructing composite ecological corridors. Finally, for overall unbalanced type regions, a cross-regional mechanism of factor symbiosis should be constructed. Administrative barriers must be broken to guide the flow of urban talent, technology, and capital toward underdeveloped ecological functional areas, building a closed value loop of “capital and technology flowing into rural areas coupled with ecological products flowing into urban areas.” This will narrow the regional development gap and achieve high-quality synergy of the Coupled System for rural tourism in the YRD region.

5. Conclusions

Based on the comprehensive analytical framework of “Evaluation-Coupling-Association”, this study evaluates and identifies the spatial coupling relationship and its associated factors between rural tourism development capacity and ecosystem recovery capacity in the YRD region. The main conclusions are as follows: (1) At the socio-economic level, rural tourism development in the YRD exhibits a spatial differentiation where the southeast is significantly superior to the northwest, with 60.46% of the regions reaching an intermediate level or above; (2) At the ecosystem level, areas with high ecosystem recovery capacity for rural tourism (accounting for 50.28%) are concentrated in the south; (3) The overall regional coordination level is relatively low, with 13 regions in the coordination stage (accounting for 35.3% of the total spatial area); (4) The spatial associations present the characteristics of “administrative guidance + technological empowerment,” reflecting a spatial pattern that is “innovation-driven” rather than “resource-dependent.” It is difficult to build core competitive advantages relying solely on the natural foundation; we hypothesize that external interventions such as capital injection and technological support may serve as potential pathways to improve the coordination level and facilitate ecological value realization in future practices.

The significance of these findings lies in the proposition of a novel evaluation perspective and attribution framework, providing a scientific basis for the industrial structure adjustment and eco-environmental optimization of different types of rural tourism. This methodology of measuring and attributing dual-system synergy can offer precise policy foundations and practical guidance for the sustainable development of rural tourism.

However, certain limitations remain in this study. The first is the non-exhaustiveness of the indicator dimensions. Constrained by macro-statistical data, the evaluation system focuses on objective indicators such as material capital, with insufficient depiction of subjective elements like tourist perception and community identity, thus simplifying the complexity of human-land interactions to some extent. The second is the constraint of spatial scale. The second is the constraint of spatial scale. The analysis relies on prefecture-level administrative units, which may mask micro-level spatial heterogeneity and carry the risk of the Modifiable Areal Unit Problem (MAUP). This implies that the observed coupling patterns at the city level might vary if examined at finer spatial hierarchies (e.g., county or village levels). Future research should prioritize multi-scale nested analysis or focus on typical case areas at the township scale to conduct robustness checks and verify if the regional patterns observed in this study persist across different spatial scales. The third is the static nature of the dynamic mechanism. While the Geographic Detector excels at attributing spatial differentiation, it inadequately reveals the time-lag effects of driving factors and the dynamic feedback mechanisms of system evolution. Fourth, potential endogeneity concerns exist; for instance, government investment might be endogenous to the inherent tourism potential of a region, making it difficult to isolate pure administrative effects. Fifth, measurement errors must be considered, as POI data from commercial platforms like Amap and Dianping may not capture all informal or small-scale rural tourism activities. Finally, the generalizability of our findings might be limited. As the Yangtze River Delta is an economically advanced coastal region, its development patterns and coupling mechanisms may not directly transfer to inland or less-developed areas. Future research should utilize multi-source heterogeneous data and spatiotemporal models to explore these issues at more granular scales and over longer time series.