From Conflict to Coexistence: Integrated Landscape Optimization for Sustainable Tourism in Urban Tourism Areas

Abstract

Urban Tourism Areas (UTAs) face growing challenges in balancing tourism development with ecological preservation, particularly under the pressures of rapid urbanization and intensified land use. However, systematic approaches to optimizing landscape patterns in urban tourism contexts remain limited. The aim of this study is to develop and apply an integrated framework that combines ecological sensitivity evaluation and landscape eco-ethics to guide sustainable landscape optimization. Using Shihe District in Xinyang City, China—a region marked by diverse natural landscapes and intensive human–environment interactions—as a case study, we applied a multi-indicator ecological sensitivity assessment together with landscape pattern analysis, supported by Geographic Information Systems (GIS) and FRAGSTATS software. The results revealed significant spatial heterogeneity in ecological sensitivity across the district. High- and very-high-sensitivity zones accounted for 23.2% of the total area, primarily located in southwestern mountainous regions, while low-sensitivity zones covered 53.8%, concentrated in urban plains and lowlands. The landscape exhibited a Shannon’s Diversity Index (SHDI) of 0.8617 and an Edge Density (ED) of 17.05, reflecting a moderately fragmented spatial structure. Based on these findings, a hierarchical optimization strategy was proposed, delineating three protection zones: primary conservation zones (23.2%), secondary buffer zones (22.9%), and development-prioritized zones (53.8%). This framework promotes ecological integrity, supports balanced tourism development, and accommodates the needs of both tourists and local communities. The model has potential applicability to other global UTAs facing similar conflicts between ecological protection and tourism expansion.

1. Introduction

Urban Tourism Areas (UTAs), located at the intersection of natural ecosystems and urban infrastructure, are increasingly becoming focal spaces for sustainable tourism development [1]. These areas integrate diverse land-use functions—ecological conservation, cultural experience, recreational service, and urban support—while simultaneously facing growing ecological pressures due to high development intensity, landscape fragmentation, and environmental degradation [2,3,4]. Similar challenges have been observed in urban tourism areas across Europe, Latin America, and Southeast Asia, where spatial conflicts between conservation and tourism intensification are becoming increasingly salient. In such complex spatial systems, achieving a balance between ecological protection and tourism development has emerged as a central challenge in contemporary landscape planning [5]. Recent international research has increasingly emphasized the concepts of tourism carrying capacity [6] and nature-based solutions (NbS) as guiding frameworks for sustainable tourism planning [7]. Tourism carrying capacity provides a quantitative benchmark for limiting visitor pressure, while NbS highlight ecosystem-based strategies to enhance resilience in urban and peri-urban destinations. However, both approaches face challenges in UTAs, where high anthropogenic intensity and fragmented landscapes complicate the translation of thresholds and NbS principles into actionable zoning strategies.

To address this challenge, ecological sensitivity evaluation has become an important tool for identifying ecologically vulnerable areas and supporting spatial zoning decisions [8]. Ecological sensitivity models have been widely applied in rural contexts [9], such as mountain parks in Austria [10] and coastal heritage areas in Spain, where land-use conflicts are less intense and ecological baselines more stable. However, such rural-based models often neglect the micro-fragmentation and tourism-driven land pressures found in UTAs, limiting their transferability. UTAs present distinct spatial characteristics: high anthropogenic intensity, fragmented land mosaics, and strong functional overlap between built and natural environments [11]. These features reduce the applicability of existing models and call for tailored methods that reflect the spatial complexity and functional hybridity of urban tourism systems [12]. In particular, rural-oriented sensitivity frameworks often adopt indicator systems that overlook tourism-driven pressures such as visitor density, facility distribution, and accessibility, while their weighting schemes rarely reflect the ecological value of cultural landscapes. These shortcomings limit their explanatory power in highly disturbed and multifunctional UTAs.

Meanwhile, landscape pattern analysis, grounded in landscape ecology and supported by tools such as FRAGSTATS and GIS, enables quantitative evaluation of spatial heterogeneity, patch configuration, and ecological connectivity [13,14]. This approach offers valuable insights into how land-use dynamics affect ecological function and aesthetic value [15]. Yet, in most cases, ecological sensitivity assessment and landscape pattern analysis are applied in isolation, without forming a unified framework that can support practical landscape zoning and optimization [16]. Moreover, few studies integrate landscape eco-ethics—which emphasize the normative values of ecological protection, intergenerational responsibility, and harmonious coexistence—into empirical planning models, particularly in urban tourism contexts where land-use conflict is acute, typically manifested in overlapping ecological and infrastructural zones, competition for land allocation, and aesthetic degradation. To address these tensions, eco-ethical principles such as spatial equity, biodiversity rights, and intergenerational justice are embedded in our zoning logic. Thus, the central aim of this research is to develop an integrated framework that explicitly addresses these hybrid pressures by combining ecological sensitivity assessment, landscape pattern metrics, and eco-ethical principles.

In light of these limitations, this study proposes an integrated framework that combines ecological sensitivity evaluation, landscape pattern analysis, and eco-ethical zoning principles to support sustainable landscape planning in UTAs [17]. Taking Shihe District in Xinyang, China as a case study—an area marked by rich natural and cultural resources but also facing significant land-use pressures—this research aims to answer the following core scientific question: To what extent do conventional ecological sensitivity evaluation models fail to reflect the hybrid pressures of UTAs, and how can landscape eco-ethics reshape spatial optimization to overcome the protection–development dichotomy? Shihe District exemplifies the multifunctional land-use challenges typical of second-tier cities in China, where rapid urbanization intersects with ecologically sensitive zones and tourism-led development. Its fragmented landscape, historical heritage, and growing visitor flows offer a representative microcosm of UTA dilemmas faced in broader Asian contexts [18].

To this end, the study constructs a multi-indicator ecological sensitivity model, incorporating both natural and anthropogenic factors, and evaluates landscape structure through spatial metrics at multiple scales [19]. By integrating these results with the principles of landscape eco-ethics, a zoning and classification strategy is developed to guide differentiated landscape management [20]. The methodological innovation lies in: (1) Extending ecological sensitivity evaluation to high-disturbance, multifunctional UTAs; (2) Constructing an integrated spatial analysis framework that bridges ecological modeling and landscape pattern metrics to inform ethical and sustainable land-use planning [21,22].

The remainder of this paper is structured as follows: Section 2 outlines the study area, data sources, and methodological framework. Section 3 presents the evaluation results and zoning strategies. Section 4 discusses the implications for sustainable tourism landscape management. Section 5 concludes with key findings and directions for future research.

2. Materials and Methods

2.1. Research Area and Data Sources

Shihe District (Figure 1) is located in southern Henan Province, characterized by a mix of plains in the northeast and mountains in the southwest [23]. This ecological diversity creates both opportunities and constraints for tourism development. Key ecological reserves, including Nanwan Reservoir and Siwangshan Provincial Nature Reserve, are subject to strict protection, while expanding tourism and urban land use place increasing pressure on the environment. Thus, Shihe exemplifies the tension between ecological conservation and tourism expansion, making it a representative case for studying landscape optimization in urban tourism areas [24].

From an ecological standpoint, Shihe encompasses several critical conservation zones, such as the Nanwan Reservoir Drinking Water Source Protection Area, Chushankudian Reservoir, and Siwangshan Provincial Nature Reserve, in addition to nationally designated key ecological function areas [25]. The se zones are subject to strict ecological management and impose significant constraints on land use and tourism development. As such, the district exemplifies the tension between ecological conservation and tourism expansion—making it a representative and meaningful case for exploring landscape optimization strategies in urban tourism areas.

This study utilized a range of geospatial and textual data sources to support the ecological sensitivity evaluation and landscape pattern analysis (

Table 1. Data source.

Data Type	Year	Data Content	Data Source
Remote Sensing Data	2020	Landsat 8/9 OLI_TIRS satellite imagery	Geospatial Data Cloud (https://www.gscloud.cn/, accessed on 27 July 2025)
Basic geographic information	2021	Road vector data	1:1 million public version of the basic geographic information data
Land use data	2020	Shihe District land use raster data	Chinese Academy of Sciences Resources and Environmental Sciences and Data Center
Administrative division	2020	Henan Province, Xinyang City, Shihe District administrative division	Standard map service system (https://www.mnr.gov.cn/ accessed on 11 September 2025)
DEM	2020	Standard map service system (https://www.mnr.gov.cn/)	Geospatial Data Cloud (https://www.gscloud.cn/)
Tourism Resources	2020	Natural Attractions, Cultural Attractions, Historical Attractions	“Xinyang City” 13th Five-Year Plan “Tourism Industry Development Plan” “Xinyang City Master Plan (2015–2030)” “Xinyang Siwang Mountain core area red tourism master plan (2015–2025)” “Shihe District overall tourism master plan manual

). Remote sensing data were obtained from Landsat 8/9 OLI_TIRS satellite imagery (April 2020) with a spatial resolution of 30 m × 30 m, accessed via the Geospatial Data Cloud. Digital Elevation Model (DEM) data were sourced from the ASTER GDEM V2 dataset at 30 m resolution. Land use data were acquired from raster datasets (30 m resolution) provided by the Resource and Environmental Science and Data Center, Chinese Academy of Sciences. Basic geographic information data were obtained from the National Catalogue Service for Geographic Information. Administrative boundary data were sourced from the Standard Map Service System of the Ministry of Natural Resources of China. In addition, tourism resource data and related textual information were collected from several official planning documents. These data collectively provided the necessary inputs for constructing the ecological sensitivity evaluation system and conducting spatial analyses of landscape patterns within Shihe District.

2.2. Research Framework

This study adapts established ecological sensitivity and landscape pattern analysis methods but introduces innovation by incorporating tourism-specific indicators, a dual AHP–entropy weighting approach, and the integration of eco-ethical zoning principles. This study develops an integrated research framework that combines landscape eco-ethics with tourism planning to guide sustainable landscape optimization (Figure 2). The analysis follows a multi-step process: First, a comprehensive dataset was established through literature review, field surveys, and remote sensing data processing. Core data layers included land use, topography, hydrology, and tourism resources. Second, ArcGIS 10.8 was used to perform both single-factor and composite ecological sensitivity evaluations, incorporating multiple ecological and socio-environmental indicators [26,27]. Each factor was assigned a weight and spatially analyzed to generate a composite sensitivity map.

Third, land use data were reclassified into five categories and processed in FRAGSTATS 4.2 to calculate key landscape metrics such as Class Area (CA), Number of Patches (NP), Patch Density (PD), Percentage of Landscape (PLAND), Largest Patch Index (LPI), Edge Density (ED), and Shannon’s Diversity Index (SHDI). These indicators revealed spatial characteristics and fragmentation levels across the landscape. Finally, by integrating ecological sensitivity results with landscape pattern metrics, the study proposed a zoning-based and classification-based landscape optimization strategy, offering targeted guidance for sustainable management of UTAs [28,29].

2.3. Ecological Sensitivity Evaluation Model

The selection of evaluation indicators is a critical step in accurately assessing ecological sensitivity, as the complexity of ecosystems results in varying degrees of influence among different factors [30]. Drawing on comprehensive baseline data, relevant literature, and established research methodologies [31], this study sought to minimize multicollinearity among factors and to ensure that the selected indicators reflected the specific ecological characteristics of the Shihe River tourism region, which is dominated by forested areas and water bodies, particularly the extensive ecological environment surrounding Nanwan Lake.

Based on these considerations, four categories of ecological sensitivity factors were selected: Landform factors: include Elevation, Slope, and Aspect. Elevation and slope influence soil moisture distribution and plant growth, while aspect affects vegetation patterns, wildlife habitat, and microclimatic conditions [32,33]. Natural conditions: factors include Vegetation Coverage, River Buffer Zone, and Lake Buffer Zone, which reflect the ecological health and sensitivity of natural habitats [34]. Tourism resource factors: include Historical Attractions, Natural Attractions, and Man-made Attractions. Kernel density analysis was performed on these features, and the results were normalized to reflect their spatial influence on ecological sensitivity [35]. In addition, tourism density and landscape accessibility were incorporated as supplementary indicators to better capture tourism-related pressures in urban tourism areas. Tourism density represents the clustering of attractions and associated visitor flows, which directly influence ecological disturbance. Landscape accessibility, derived from distance to major scenic roads, reflects the dual role of transportation in improving tourism convenience while amplifying ecological sensitivity in adjacent habitats. Human activity factors: include Land Use Type and Road Buffer Zone, which capture the intensity of human exploitation and disturbance [36].

By incorporating both conventional ecological indicators and tourism-specific factors such as accessibility and density, the indicator system ensures that ecological sensitivity evaluation reflects not only environmental processes but also the unique pressures of urban tourism activities. The complete classification of factors is presented in

Table 2. The ecological sensitivity classification standard of evaluation factors.

	Sensitivity Level	Insensitive	Low Sensitive	Medium Sensitive	High Sensitive	Extremely Sensitive
	Rank assigned value	1	2	3	4	5
Landform	Elevation/m	45~100	100~200	200~300	300~400	400~895
	Slope/(°)	0~2	2~1	5~15	15~25	25~57
	Aspect	South, Flat	Southwest, Southeast	East, West	Northwest, Northeast	North
Natural conditions	Vegetation coverage	<0.1	0.1~0.15	0.15~0.25	0.25~0.35	>0.35
	River buffer zone	>800	500~800	200~500	50~200	<50
	Lake buffer zone	>800	500~800	200~500	50~200	<50
Tourism resource	Historical attractions	>800	500~800	200~500	100~200	<100
	Natural attractions	>500	200~500	100~200	50~100	<50
	Man-made attractions	>200	100~200	50~100	20~50	<20
	Density of attractions	0~0.2	0.2~0.4	0.4~0.6	0.6~0.68	0.8~1
Human activities	Land use type	Artificial Surfaces, Bare land	Grassland	Forest	Water Bodies	Cultivated Land
	Road buffer zone	>200	150~200	100~150	50~100	<50

. The thresholds for classification were derived from multiple sources; vegetation coverage and NDVI from ecological function benchmarks in recent studies; river, lake, and road buffers from ecological redline policies and planning guidelines; and tourism density from local planning documents and empirical distribution data of Shihe District. This combination of standards, literature, and local adaptation ensures that the thresholds are both scientifically robust and contextually relevant.

Given the complex interplay of natural and anthropogenic factors influencing ecological sensitivity, determining appropriate factor weights is essential for ensuring objectivity and analytical rigor [37]. To address variability and reduce subjectivity, this study employed a combined weighting approach integrating subjective and objective methods. First, the Analytic Hierarchy Process (AHP) was used to derive subjective weights based on pairwise comparisons by a panel of five domain experts, including two ecologists, one tourism planner, one landscape architect, and one urban/regional planner. This interdisciplinary composition ensured that both ecological and tourism-related perspectives were considered. To minimize bias, each expert provided independent judgments, and the consistency of the judgment matrices was tested (CR < 0.1). Second, the entropy weight method was used to compute objective weights based on the inherent variability of the dataset, thereby minimizing subjective bias by giving higher weights to factors with greater discriminatory power. Finally, the two sets of weights were integrated using an entropy-based minimization formula to obtain composite weights, which were subsequently applied in constructing the ecological sensitivity index. This combined method ensures a balanced representation of expert knowledge and data-driven evidence.

Scale method to construct factor judgment matrix.

By comparing the factors with each other and carrying out basic processing of the data to express the importance of the factors in the evaluation system [38].

Conduct consistency test on judgment matrix using Formula (1) to calculate the maximum characteristic root of judgment matrix, and then determine the consistency index C.I. (C.I. = λmax − n/n − 1, where n is the number of matrix factors), and finally calculate the stochastic consistency ratio coefficient C.R. (C.R. = C.I./R.I.), if C.R. < 0.10, the judgment matrix is considered to have acceptable consistency; otherwise (C.R. ≥ 0.10), the matrix must be reconstructed. In this study, CI refers to the Consistency Index, RI is the Random Index based on matrix order, and CR is the ratio coefficient defined as CI/RI [39].

$${\lambda }_{max}={\displaystyle \frac{1}{n}}{\sum }_{j=1}^n{\displaystyle \frac{{\left(XW\right)}_i}{W_i}}$$

(1)

where n is the number of matrix factors, XW is the judgment matrix X multiplied with W.

Calculation of factor weights.

Factor weights are calculated by using the product method, firstly, applying Formula (2) to multiply the factors of each row, (x is the judgment matrix, n is the order of the matrix); secondly, open the nth square root of $w_i$, i.e., $\overline{w_i}=\sqrt[n]{w_i}$, and finally, applying Formula (3) to vectorize the square root of each row in the judgment matrix x [40].

$$w_i={\sum }_{j=1}^n{x}_{ij}$$

(2)

$$X_i={\displaystyle \frac{\overline{w_i}}{\sum _{j=1}^n\overline{w_j}}}\left(\overline{w_i}=\overline{w_1},\overline{w_2},\dots \overline{w_n}\right)$$

(3)

where x is the judgment matrix and n is the order of the judgment matrix.

In this study, Yaahp 10.3 software was also utilized for the auxiliary calculation of hierarchical analysis weight values [41], which also facilitates the calculation of factor weights and related judgment matrix consistency test. The consistency CR values of the target layer and comprehensive evaluation layer are 0.0936, 0.0372, 0.0723, 0.0000, 0.0372, respectively, which are in the range of 0 to 0.1, proving that the factor judgment matrices have feasibility; and the weights of the comprehensive factor layer are W = [0.1563, 0.4906, 0.3089, 0.0443].

The basic principle of entropy value method to calculate the weights is to assign weights to the factors according to the relative effect value that the factors themselves have [42]; however, for the factors with a small degree of change, the weight value is about 0. The specific steps are as follows:

Standardize the judgment matrix x:

$$x_j={\displaystyle \frac{1}{xj}}, x_{ij}={\displaystyle \frac{ x_{ij}-\overline{x_j}}{{\delta }_j}}$$

(4)

where xij is the standardized matrix data, xj is the mean value of the jth evaluation factor, δj is the standard variance of the jth evaluation factor. For the negative data, use the translation method to eliminate negative values, z is the translation amplitude of the evaluation factor, the formula is as follows:

$$X_{ij}=x_{ij}+z$$

(5)

The weight of the jth evaluation factor in the ith sample is calculated and the standardized judgment matrix is Y.

$$Y_{ij}={\displaystyle \frac{x_{ij}}{\sum _{i=1}^m{x}_{ij}}}, Y={\left\{Y_{ij}\right\}}_{m,n}$$

(6)

where n is the number of factors in the judgment matrix, m is the number of factor data that the factor has, according to the relevant calculation method, the entropy value of the jth evaluation factor and the coefficient of variation is obtained as:

$$e_j={\displaystyle \frac{-1}{\mathit{ln}m}}{\sum }_{i=1}^m{y}_{ij}\mathit{ln}{y}_{ij},0<e_j<1$$

(7)

$$g_{ij}=1-e_j$$

(8)

Calculation of evaluation factor indicator weights and indices:

$$w_j={\displaystyle \frac{g_j}{{\sum }_{i=1}^m{g}_j}}, R_i=\sum _{i=1}^m{w}_{ij}{y}_{ij}$$

(9)

Assuming that the comprehensive weight is wj, theoretically it should be as close as possible to the weight Wi calculated by hierarchical analysis and the weight wj calculated by entropy value method. Based on the relative information entropy minimization for the calculation and combined with the Lagrange multiplier method to simplify the formula [43], so as to obtain the factor comprehensive weights (

Table 3. The comprehensive weight table of ecological sensitivity evaluation factors in Shihe tourist area.

Target Layer	Comprehensive Evaluation Layer	Evaluation Factor Layer
		Indicator	AHP Empowerment	Entropy Method Weighting	Comprehensive Empowerment	Order
Ecological sensitivity evaluation of Shihe tourist area	Landform	Elevation	0.1397	0.1121	0.1372	3
		Slope	0.0367	0.0556	0.0404	8
		Aspect	0.1229	0.1034	0.1189	4
	Natural conditions	Vegetation coverage	0.2314	0.1698	0.1819	2
		River buffer zone	0.0545	0.0867	0.0881	5
		Lake buffer zone	0.2191	0.2389	0.2187	1
	Tourism resource	historical attractions	0.0292	0.0301	0.0284	10
		natural attractions	0.0392	0.0692	0.0481	7
		Man-made attractions	0.0497	0.0173	0.0204	11
		Density of scenic attractions	0.0141	0.0589	0.0396	9
	Human activities	Land use type	0.0437	0.0498	0.0598	6
		Road buffer zone	0.0239	0.0156	0.0186	12

). The factor composite weighting formula is shown in Equation (10).

$$\overline{w_j}={\displaystyle \frac{\sqrt{W_j{w}_j}}{\sum _{j=1}^n\sqrt{W_j{w}_j}}}$$

(10)

In Shihe District factor comprehensive assignment table (Table 3), by the weight value ranking can be seen, vegetation type, water body area, soil type, elevation, watershed buffer, vegetation cover is more forward, the degree of influence on ecological sensitivity is higher, and the study area of the protection of the focus of the higher degree of fit.

2.4. Landscape Pattern Analysis

Landscape pattern analysis was conducted to quantify the spatial characteristics of land use types in Shihe District and to assess their relationship with ecological sensitivity. The comprehensive ecological sensitivity evaluation maps generated in ArcGIS 10.8 were converted from raster to vector format and spatially overlaid with land use maps. Subsequently, Fragstats 4.2 software was employed to calculate key landscape metrics, including Patch Area (CA), Number of Patches (NP), Patch Density (PD), Edge Density (ED), Largest Patch Index (LPI), Percentage of Landscape (PLAND), and Shannon’s Diversity Index (SHDI), among others.

Based on the Shihe District Land Use Master Plan, regional tourism development characteristics, and field survey results, the landscape was classified into five primary categories: First, Arable Land, which includes cultivated fields, paddy fields with irrigation infrastructure, and dryland fields dependent on natural precipitation [44]. Second, Woodland, encompassing both natural and planted forests with >30% canopy cover, short woodlands and shrublands with >40% canopy cover and height < 2 m, and sparse forests with 10–30% canopy cover [45]. Third, Grassland, comprising high-, medium-, and low-cover grasslands. Unused land with minimal spatial extent was merged into this category for calculation consistency [46]. Fourth, Water Bodies, including natural and artificial rivers, canals, lakes, reservoirs, ponds, and mudflats [47]. Fifth, Construction Land, consisting of urban and rural settlements, tourist facilities, squares, roads, and other areas of intensive human activity [48]. This classification scheme provided the basis for subsequent quantitative analysis of landscape patterns.

Based on the results of ecological sensitivity evaluation and spatial pattern analysis, a set of landscape metrics was selected to assess the spatial structure and ecological stability of the landscape in relation to tourism development [49]. These metrics provide valuable insights into landscape configuration and composition, which are critical for supporting sustainable tourism development. CA measures the total area of a given landscape class (ha). Larger CA values usually support greater species diversity and ecological stability, making this indicator useful for identifying key habitats and guiding conservation. NP represents the number of patches within the landscape. Higher NP indicates stronger fragmentation and spatial heterogeneity, while lower NP reflects more consolidated and homogeneous patterns. NP directly affects habitat connectivity, species distribution, and ecological disturbance dynamics.

PLAND is used to measure the various plant community composition components in a landscape and to calculate the proportion of different types of patches in the overall landscape area, thus providing a basis for elucidating their dominance. As its value approaches 0, it indicates that the variety of patches in the landscape is progressively decreasing, and at 100 points, it indicates that there is only one patch in the landscape [50].

$$PLAND=p_i={\displaystyle \frac{\sum _{i=1}^n{a}_{ij}}{A}}\left(100\right)$$

(11)

where a_ij represents the area of patch ij and A represents the total area of all landscapes.

Landscape patch density is the level of differentiation and fragmentation of patches in the landscape. High density signifies a higher number of patches of different types of landscape elements in the region, a lower scale, and a higher diversity of landscapes.

$$PD={\displaystyle \frac{n_i}{A}}$$

(12)

$${PD}_i={\displaystyle \frac{1}{A}}{\sum }_{j=1}^M{N}_i$$

(13)

$$A={\sum }_{i=1}^M{A}_i$$

(14)

where PD stands for the total number of different types of landscape elements in the landscape, A for the total area of the study area, Ni for the number of patches of category I landscape elements, and A_i for the total area of category I landscape elements. PD_i also represents the density of landscape element patches, which is the number of patches per unit area of a particular type of element in the landscape [51].

Landscape edge density is determined by the pattern of landscape elements and the density of patches, which effectively expresses the interaction of different types of matter, energy, species and other information [52]. Anthropogenic activity patterns determine the boundary distribution of man-made patches, while the boundary density of natural patches reflects the dynamics of their changes. The dynamic characteristics of each element can be revealed from the distribution of their boundaries.

$$ED={\displaystyle \frac{1}{A}}{\sum }_{i=1}^M\times {\sum }_{j=1}^M{P}_{ij}$$

(15)

$${ED}_i={\displaystyle \frac{1}{A}}{\sum }_{j=1}^M{P}_{ij}$$

(16)

where ED_i is the edge length between a particular kind of landscape element patch and its surrounding heterogeneous patches per unit area, and ED is the edge length between heterogeneous landscape element patches per unit area of the entire landscape [53]. M is the total number of different types of landscape elements; A is the total area of the landscape within the study area; N_i is the number of different patches of category i elements; A_i is the total area of category i elements; and P_ij is the length of the boundary in the landscape between the patches of category i elements and the patches of category j elements.

SHDI is an important indicator of the biodiversity of biomes, which can reflect the heterogeneity of the landscape, especially sensitive to the uneven distribution of landscape plants in different patches [54]. Meanwhile, SHDI, as a more sensitive indicator, can reflect landscape diversity and heterogeneity in different periods. When calculating landscape system indicators, if there are more land resources with higher fragmentation, more qualitative information is needed and the SHDI value will be higher.

$$SHDI=-{\sum }_{i=1}^m\left(p_{i}ln{p}_i\right)$$

(17)

where P_i represents the landscape patch type and i represents the occupied ratio.

3. Results and Analysis

3.1. Evaluation of Ecological Sensitivity

3.1.1. Terrain Conditions

Elevation (Figure 3a). Elevation significantly influences climate, vegetation types, and ecological stability [55]. As altitude increases, temperature drops, biodiversity decreases, and ecological sensitivity rises. In Shihe District, elevation gradually increases from northeast to southwest, with areas above 400 m—mainly in southwestern mountainous zones—identified as highly sensitive. These zones are ecologically fragile and critical for conservation. In contrast, lower elevations in the central and northeastern areas correspond to the main urban construction zone. Overall, high and very high sensitivity zones occupy about 30% of the district, underscoring the importance of targeted protection and optimization in upland regions.

Slope (Figure 3b). Slope affects vegetation distribution, erosion risk, and water runoff [56]. Steeper slopes are more ecologically unstable and less suitable for intensive development. GIS-based classification identified five slope categories, with flatter areas (<6°) concentrated in the north and northeast, where development intensity is higher. In contrast, steeper zones (>15°) in the west and south are ecologically sensitive and less disturbed. Notably, 60% of the area has moderate slopes (5°–15°), forming valleys and ridges particularly suitable for eco-tourism development, especially near Nanwan Lake.

Slope Orientation (Figure 3c). Aspect influences microclimate, solar exposure, and vegetation growth. In the Northern Hemisphere, north-facing slopes are typically cooler and more humid, supporting more diverse and stable plant communities. GIS classification divided aspect into five sensitivity levels. Medium sensitivity zones account for 46.5% of the area, while high sensitivity areas—primarily north-facing slopes—make up 16.67%. The relatively balanced spatial distribution of slope orientation suggests broad potential for sustainable tourism planning across different terrain conditions.

3.1.2. Natural Conditions

Water Resource Abundance (Figure 3d). Water bodies are fundamental to both ecological stability and tourism in Shihe District. High and very high sensitivity zones—mainly in the northeastern mountains and around Nanwan Reservoir—cover approximately 410 km² (23% of the district;

Table 4. Shihe District factors of ecological sensitivity grading area statistics.

Sensitivity Level	Insensitive Area		Low Sensitive Area		Medium Sensitive Area		Highly Sensitive Area		Extremely Sensitive Area
	Area/km2	Percentage/%	Area/km2	Percentage/%	Area/km2	Percentage/%	Area/km2	Percentage/%	Area/km2	Percentage/%
Elevation/m	284.7707	15.95%	898.2992	50.31%	226.8177	12.70%	164.1203	9.19%	211.6478	11.85%
Slope/(°)	123.9226	6.97%	362.7397	20.39%	826.1432	46.44%	342.5821	19.26%	123.6534	6.95%
Aspect	231.6348	13.02%	426.8599	23.99%	453.8903	25.51%	450.0195	25.30%	216.6509	12.18%
Vegetation coverage	104.7003	5.92%	94.12752	5.32%	205.5952	11.62%	741.0042	41.88%	623.8642	35.26%
River buffer zone	753.4799	42.20%	342.4928	19.18%	391.5181	21.93%	219.7936	12.31%	78.35949	4.39%
Lake buffer zone	71.9266	4.03%	30.08212	1.68%	48.43801	2.71%	36.57074	2.05%	1598.694	89.53%
historical attractions	1758.36	98.47%	16.65	0.93%	8.97	0.50%	1.28	0.07%	0.42	0.02%
natural attractions	1771.57	99.21%	11.87	0.67%	1.69	0.10%	0.42	0.02%	0.14	0.01%
Man-made attractions	1783.952	99.90%	1.319469	0.07%	0.329867	0.02%	0.092363	0.01%	0.017593	0.00%
Density of attractions	28.31651	1.59%	1.803752	0.10%	1064.68	59.62%	81.3591	4.56%	609.4814	34.13%
Land use type	1628.54	91.20%	38.51	2.16%	39.00979	2.18%	39.54839	2.21%	40.09173	2.25%

). These areas are critical for water regulation and ecosystem services. To preserve ecological integrity, strict protection of buffer zones is essential, including delineation of ecological red lines, pollution control, and restricted development [57]. A systematic network of protected areas and conservation corridors should be established to enhance waterfront ecosystem resilience and support sustainable tourism.

Vegetation Cover (Figure 3e). Vegetation plays a key role in air purification, soil stabilization, microclimate regulation, and biodiversity—factors that contribute to tourism attractiveness and ecological resilience. Using NDVI, vegetation was classified into five sensitivity levels. Very high sensitivity zones (>0.35) account for 30.63% of the area and are concentrated around South Bay Lake. Moderate sensitivity zones span over 50%, forming an extensive ecological buffer. Low-sensitive areas are primarily located in urban centers. Management should prioritize native vegetation conservation and ecological restoration, especially in high-sensitivity zones.

Water Source Buffer Zones (Figure 3i). Buffers around major water bodies are ecologically sensitive due to their role in protecting water quality. Low sensitivity zones cover 56.99% of the area, while high and very high sensitivity zones account for 22.15% and 3.61%, respectively. These zones are concentrated around South Bay Reservoir and Out-of-the-Mountain Reservoir, the district’s main water sources. Effective protection requires land-use controls, riparian vegetation restoration, and active monitoring to maintain hydrological functions and mitigate tourism-related pressures.

3.1.3. Tourism Resource and Human Activities

Tourism Resources (Figure 3f–h). Tourism resources in Shihe District exhibit strong spatial heterogeneity in ecological sensitivity. Historical landscapes are mainly clustered around the northwest and southeast peripheries of South Bay Reservoir, while natural and man-made attractions are more dispersed. However, the vast majority of tourism-related areas are classified as very low sensitivity, accounting for 98.47%, 99.21%, and 99.9% of the total area for historical, natural, and artificial attractions, respectively. This indicates a landscape with numerous but fragmented tourism nodes, requiring differentiated protection strategies. Sensitivity-based zoning should guide the intensity, type, and location of tourism development to balance utilization and conservation.

Human Activities (Figure 3j,k). Land use patterns reveal that human disturbance is a major driver of ecological sensitivity. Extremely sensitive zones—primarily in the north, east, and around South Bay Lake—occupy 34.13% of the district, while medium sensitivity areas represent the largest share (59.62%), spread mainly across the south and west. Low and insensitive zones are minimal (1.69% combined), indicating that most of the district experiences moderate to high human impact. These results highlight the urgent need for land-use regulation and ecological planning to mitigate tourism-induced degradation.

In terms of transportation infrastructure (Figure 3k), road networks are generally classified as insensitive, covering 91.2% of the area and aligning with highways and mountain roads. While crucial for tourism access, roads may fragment habitats. Therefore, strategic routing, ecological buffers, and connectivity corridors should be incorporated into transportation planning to reduce ecological disruption.

3.1.4. Comprehensive Evaluation

Based on the weighted integration of all sensitivity factors, a composite ecological sensitivity map of Shihe District was generated through GIS overlay analysis (Figure 4), with area statistics summarized in

Table 5. Results of comprehensive ecological sensitivity analysis.

Sensitivity Level	Insensitive	Low Sensitive	Medium Sensitive	High Sensitive	Extremely Sensitive
Area/km2	410.05	534.93	392.24	267.47	124.82
Ratio/%	23.11	30.71	22.96	15.36	7.86

. The district is divided into three broad sensitivity levels, each requiring differentiated management strategies.

High and Very High Sensitivity Zones. Covering 392.29 km² (23.22% of the district), these zones are concentrated in the southwestern mountainous areas, featuring steep terrain, dense vegetation, and ecological fragility. They serve as critical ecological buffers supporting landscape stability. Development in these areas should be strictly restricted, with only essential ecological monitoring infrastructure permitted. Priorities include enforcing ecological red lines, controlling land use, and preventing habitat loss and pollution. These proportions (23.2%, 22.9%, and 53.8%) were determined by integrating national ecological red-line guidelines with Shihe’s local land-use planning documents.

Moderate Sensitivity Zones. These areas span 392.24 km² (22.96%) and are mostly located around Nanwan Reservoir and its buffer zones. While less fragile, they remain ecologically vulnerable. Tourism in these zones should follow protection-oriented planning, with controlled visitor access and well-defined development boundaries. Eco-corridor construction and low-impact tourism—such as water-based recreation—can support conservation while enhancing landscape use.

Low and Less Sensitive Zones. Accounting for 944.98 km² (53.82%), these zones lie in urbanized and tourism development regions, where anthropogenic pressure is high. Key issues include water degradation, land fragmentation, and poor infrastructure layout. Here, emphasis should be placed on scientific land-use planning, improving spatial layout, enhancing tourism carrying capacity, and promoting synergies between urban growth and ecological sustainability.

3.2. Multi-Scale Landscape Pattern Analysis

Landscape diversity and fragmentation in Shihe District were evaluated using the Shannon Diversity Index (SHDI) and Shannon Evenness Index (SHEI). The SHDI value of 0.8617 (

Table 6. Calculation results of overall landscape pattern index in Shihe District.

TA (ha)	NP	PD	LPI	ED	SHDI	SHEI
178,693.02	566	0.3167	35.9951	17.0562	0.8617	0.7806

) reflects a moderately high level of land-use diversity, indicating a landscape composed of varied and spatially heterogeneous land-use types. The SHEI value of 0.7806 suggests that the overall distribution of land-use categories is relatively balanced, with limited dominance by any single type. However, certain subregions still exhibit localized anthropogenic disturbances, contributing to uneven spatial structures across the district.

Patch Density (PD). As shown in Figure 5a and

Table 7. Calculation results of landscape pattern patch type index in Shihe District.

TYPE	CA (ha)	PLAND	NP	PD	LPI	ED
Cropland	35,483.49	17.8572	161	0.0901	5.2224	11.5084
Woodland	126,944.1	71.0403	42	0.0235	35.9951	13.6037
Grassland	8531.73	4.7745	117	0.0655	3.5637	4.9506
Water	6739.74	3.7717	234	0.1310	1.3277	3.4554
Building land	993.96	0.5562	12	0.0067	0.2065	0.5906

, water bodies display the highest PD (0.1310), indicating fragmented spatial distribution and weak connectivity. Arable land (0.0901) and grassland (0.0655) also show moderate fragmentation, while construction land has the lowest PD (0.0067), reflecting its compact extent.

Edge Density (ED). Woodland exhibits the highest ED (13.60), followed by arable land (11.51), suggesting stronger edge effects and higher ecological vulnerability. In contrast, grassland (4.95) and water bodies (3.46) present lower edge density, indicating relatively cohesive boundaries.

Percentage of Landscape (PLAND). Forest land dominates the landscape (71.04%), with arable land (17.86%), grassland (4.77%), and water bodies (3.77%) forming the secondary classes. This structure highlights the ecological and tourism importance of forest and farmland.

Class Area (CA). Woodland covers the largest area (126,944.1 ha), followed by arable land (35,483.49 ha), grassland (8531.73 ha), and water bodies (6739.74 ha). Construction land is concentrated in the northeastern urban core and around South Bay Lake, emphasizing the need for integrated management of built-up and ecological spaces.

As shown in Table 7, woodland dominates the landscape with the largest class area (71.04%) and the highest Largest Patch Index (LPI = 35.99), indicating that forest patches are relatively continuous and ecologically stable. This provides a strong foundation for biodiversity conservation and forest-based tourism. In contrast, cropland, although covering nearly 18% of the landscape, shows a much lower LPI (5.22) and relatively high edge density (11.51), reflecting greater fragmentation and vulnerability to ecological disturbance. Grassland and water bodies, while smaller in proportion, exhibit high patch density (0.0655 and 0.1310, respectively), highlighting their scattered distribution and weaker connectivity. Such fragmentation may compromise ecosystem services and limit their recreational potential unless complemented by ecological corridors. Building land, though occupying less than 1% of the area, shows the lowest connectivity (PD = 0.0067) but exerts disproportionate pressure due to its concentration around key tourism hubs.

3.3. Landscape Pattern Optimization

3.3.1. Zoning Optimization

Based on the comprehensive ecological sensitivity assessment and landscape resource evaluation, Shihe District is divided into three protection levels to guide differentiated management and promote sustainable tourism (Figure 6).

Primary Protection Zone. Concentrated in the southwestern mountainous areas, this zone covers 23.2% of the district and is characterized by steep slopes, dense vegetation, and high biodiversity. These areas represent critical ecological buffers and are classified as extremely fragile. Development is strictly prohibited except for essential monitoring infrastructure. Conservation priorities include forest restoration, biodiversity protection, and slope stabilization. Limited low-impact tourism facilities such as observation decks may be considered to support environmental education without compromising ecological integrity.

Secondary Protection Zone. Covering 22.9% of the district, these areas are mainly distributed around Nanwan Reservoir and its surrounding buffer zones. With moderate ecological sensitivity and relatively higher carrying capacity, they can accommodate regulated ecotourism activities. Key management measures include establishing eco-corridors, limiting visitor flow, and enforcing strict buffer protection. Stakeholder participation is crucial, particularly involving local residents, cultural operators, and tourism enterprises in eco-friendly tourism projects such as eco-villages, cultural towns, and water-based recreation.

Tertiary Protection Zone. Spanning 53.8% of the district, this zone comprises the eastern and northeastern flatlands where human activities are most intensive, including cultivated land and construction areas. While less sensitive ecologically, they face risks such as land degradation and habitat fragmentation. These areas are designated for tourism service hubs and supporting infrastructure (e.g., accommodation, transport nodes, visitor centers). Planning strategies emphasize eco-friendly construction, integration of green infrastructure, and heritage conservation to enhance tourism capacity while mitigating ecological stress. These delineations not only provide a spatial management framework but also serve as practical references for local authorities in enforcing ecological red-line policies and regulating tourism intensity according to sensitivity levels.

3.3.2. Classification Optimization

To further promote ecological integrity and improve landscape functionality in line with sustainable tourism objectives, this study proposes a classification-based optimization strategy that organizes the district’s landscape into functional typologies—mountain, water, and woodland systems—each with tailored management priorities. This classification approach aims to reconcile ecological conservation and tourism development by regulating the spatial intensity, ecological roles, and tourism potentials of different landscape elements (Figure 7).

Mountain landscapes, such as the Dabie and Tongbai mountain ranges and other regional peaks, play a foundational role in soil and water conservation, biodiversity maintenance, and regional ecological stability. These areas also provide significant visual and recreational appeal. Management efforts in mountain regions should focus on ecological protection through the designation of upper watershed conservation forests, strict demarcation between natural forest and productive land uses (e.g., tea plantations), and the restoration of ecologically degraded slopes. Construction activities, if any, must be strictly controlled, with tourism infrastructure confined to low-impact formats that follow existing paths and preserve terrain morphology, thus aligning with the principles of landscape eco-ethics.

Water landscapes, including rivers, reservoirs, streams, ponds, and waterfalls, function as both critical ecological corridors and tourism resources. These areas should be prioritized for integrated water management by delineating comprehensive ecological protection zones and restoring riparian habitats through the implementation of buffer zones and wetland systems. To enhance flood resilience and ecosystem services, sponge city principles should be incorporated where appropriate. Recreational infrastructure, such as eco-sensitive trails or observation platforms, can be introduced in selected areas to foster environmental interpretation and human-nature interaction. Simultaneously, destructive activities—such as channel hardening, straightening, or illegal dredging—must be strictly prohibited to protect aquatic integrity.

Woodland systems—including native forests, plantations, and recreational green corridors—are essential for carbon sequestration, climate regulation, and biodiversity support, while offering opportunities for nature-based tourism. Management should emphasize the conservation of natural forests, reforestation with indigenous species, and improvement of forest connectivity across fragmented patches. In addition, preventative measures such as fire and pest control, alongside the establishment of welfare forests for community use, can enhance multifunctional landscape values. Particular attention should be given to the preservation of ancient and culturally significant trees through long-term monitoring and funding mechanisms, ensuring that both ecological and cultural functions are maintained for future generations.

4. Discussion

This study highlights the ecological fragility of UTAs by revealing how landform, water resources, vegetation, and tourism activities collectively shape spatial sensitivity patterns in Shihe District. The finding that high and very high sensitivity zones occupy 23.2% of the area indicates that ecological conflicts are concentrated in mountainous and reservoir-buffer regions, where ecological thresholds are easily exceeded under intensive tourism development. This underscores the need to recognize ecological sensitivity not merely as an environmental attribute, but as a spatial constraint guiding planning decisions [58].

The analysis also demonstrates how landscape metrics, such as patch density and edge density, reveal fragmentation processes that are particularly relevant in tourism-driven landscapes. For instance, high edge density in woodland areas reflects the pressure of tourism infrastructure encroaching on ecological boundaries, while the fragmentation of water bodies suggests potential conflicts between scenic utilization and hydrological stability [59]. These findings resonate with previous studies emphasizing that conventional tourism development often prioritizes accessibility and facility expansion at the expense of ecological thresholds, but extend the discussion by quantifying the spatial mechanisms through which such conflicts materialize.

A key implication of the results is that landscape optimization in UTAs requires reconciling ecological ethics with tourism demand. The delineation of conservation-prioritized zones operationalizes the principle of ecological ethics by embedding ecological thresholds into spatial zoning, thereby ensuring that development intensity respects the carrying capacity of sensitive ecosystems. Unlike traditional approaches that apply rural-based ecological models with limited attention to urban–tourism interactions, the framework here integrates tourism-specific indicators—such as attraction density and accessibility—making it more applicable to highly urbanized tourism contexts. Compared with conventional carrying capacity assessments, which often remain static and visitor-focused, our framework integrates ecological sensitivity with landscape pattern metrics, offering a more spatially explicit and adaptive tool for UTAs. Moreover, the incorporation of eco-ethical zoning parallels the NbS discourse, as both approaches emphasize the restoration of natural processes and the integration of ecological thresholds into planning decisions. This integrated approach not only addresses the static and visitor-centric limitations of traditional models but also fills a critical gap in applying spatially explicit ecological thresholds to highly fragmented UTAs, where land-use conflicts are most acute.

Beyond Shihe District, the proposed framework demonstrates potential transferability to other UTAs worldwide. In Southeast Asia, mountainous destinations such as northern Thailand face comparable challenges in reconciling hydrological protection with tourism expansion. In Europe, heritage-rich but densely visited regions, such as coastal UTAs in Italy, struggle with fragmented landscapes under intense visitor pressure [60]. Likewise, in Latin America, ecotourism areas on the urban fringes, including reserves in the Brazilian Atlantic Forest, must balance conservation goals with economic use [61]. While the framework requires contextual adaptation to local socio-political and data conditions, these parallels highlight its broader applicability and the relevance of integrating ecological sensitivity with tourism-driven pressures across diverse global contexts [62].

Finally, this study advances the discourse on sustainable tourism by linking quantitative ecological sensitivity results to practical zoning strategies. Yet, the discussion also highlights unresolved trade-offs: for example, restricting development in high-sensitivity zones may conflict with local economic expectations, while allowing controlled ecotourism in moderately sensitive areas requires strong governance capacity. Addressing these trade-offs calls for participatory planning, where residents, planners, and tourism operators co-design adaptive strategies to align ecological ethics with socio-economic realities. For policy makers, the quantitative findings translate into clear spatial thresholds for planning, such as prioritizing conservation in 23.2% of highly sensitive zones, while managers can use patch-level indicators to guide decisions on where eco-corridors, infrastructure, or restoration projects should be implemented.

5. Conclusions

This study provides a comprehensive assessment of ecological sensitivity and landscape pattern optimization for Shihe District, a tourism area characterized by diverse natural and cultural resources. The spatial distribution of tourism resources in the district exhibits a pattern of “widely dispersed with localized concentrations,” constrained by complex topography and limited land availability for large-scale centralized development. These conditions favor the promotion of decentralized, small-scale, and high-quality tourism initiatives that align with ecological conservation priorities.

The integrated analysis revealed distinct spatial patterns of ecological sensitivity, with low-sensitivity zones concentrated in the eastern and northern regions, moderately sensitive areas distributed in the northeast at elevations of 200–300 m, and highly sensitive zones located in the southwestern mountainous regions. Guided by the principles of ecological priority, green development, and landscape diversity, the district was delineated into a three-tier protection system: first-level protection zones prioritize landscape restoration, forest enhancement, and biodiversity conservation; second-level zones maintain landscape integrity and support controlled eco-tourism development; and third-level zones emphasize sustainable land-use planning and balanced development to enhance tourism carrying capacity while preserving cultural and natural heritage. Methodologically, this study demonstrates the utility of integrating ecological sensitivity evaluation with landscape pattern analysis, supported by GIS and multi-criteria decision-making tools. The approach offers a scientifically robust basis for informing spatial planning and targeted management strategies in complex tourism landscapes. The classification-based and zoning-based optimization strategies proposed herein provide actionable guidance for achieving a harmonious balance between conservation and development.

Nevertheless, several limitations should be acknowledged. The land-use classification system adopted here is relatively coarse, with only five broad categories, which may obscure fine-scale heterogeneity and lead to underestimation of localized ecological risks. Moreover, the model emphasizes natural and spatial factors, while socio-economic and policy drivers—key determinants of tourism-induced landscape change—are underrepresented, potentially reducing the explanatory power for human–environment interactions. In addition, the analysis relies on static land-use data from 2020, which limits its temporal generalization and may affect the robustness of results in dynamic tourism contexts. Future research should refine the framework by incorporating higher-resolution land-use datasets, longitudinal monitoring, and scenario simulations, while integrating socio-economic and governance variables to capture broader drivers of change. Cross-regional comparative studies will also be necessary to test the transferability of the framework. Addressing these limitations will improve the reliability of results and enhance the utility of ecological planning tools for adaptive, sustainability-oriented management in tourism-intensive regions.