Spatial Configuration and Accessibility Assessment of Recreational Resources in Hainan Tropical Rainforest National Park

Abstract

Recreational resources, fundamental to ecological experiences, are critical in balancing conservation with development. Effective ecotourism planning is especially vital for newly established protected areas such as the Hainan Tropical Rainforest National Park in China’s developing system of natural conservation areas. Targeting Hainan Tropical Rainforest National Park, this study applies nearest neighbor index, kernel density analysis, and exploratory spatial data analysis (ESDA) to study the spatial pattern of 274 recreational resource points. Results indicate a clustered spatial pattern with significant differences in resource density among municipalities. Specifically, 98% of these resources can be reached in 3 h, with an average travel time of 91 min, and cultural resources exhibit greater accessibility than natural resources. Natural resource availability and ethnic culture are major factors of resource distribution and accessibility. This research offers a theoretical basis and practical guidance for optimizing recreational resource allocation and promoting ecotourism in the park, contributing to the ongoing discussion of sustainable tourism development.

1. Introduction

The establishment of the Hainan Tropical Rainforest National Park (HTRNP) is essential to strengthening Hainan’s ecological defenses. This park comprises one of the most concentrated, well-preserved, and extensive continental island-type tropical rainforests in China, representing an invaluable natural resource. The distribution and accessibility of recreational resources are critical factors of the quality of ecotourism in the park. To preserve ecological equilibrium, sustainable management of these resources holds significance.

Analyzing access to attractions in the Hainan Tropical Rainforest National Park is crucial for reconciling conservation and development in China’s newly established area system [1]. These attractions hold significant appeal for international tourists and constitute key ecotourism destinations. An accessibility assessment can measure the effectiveness and usability of existing tourism routes, identifying areas for improvement and opportunities for growth. Measuring accessibility can optimize the use of recreational resources, thereby promoting ecotourism and enhancing visitor experience.

1.1. Literature Review

Since Hansen’s conceptualization of accessibility [2], it has been the subject of extensive studies in the literature [3,4,5,6]. Accessibility, defined as the ease with which a destination can be reached from a given origin through a particular transport mode [7], represents a crucial socio-economic indicator for evaluating the locational advantages of geographic entities. The relationship between socio-economic factors and the accessibility of tourist destinations highlights its importance in measuring a region’s tourism capacity. For tourists, ease of access plays a critical role in destination choice and offers a key metric for evaluating the effectiveness of local tourism infrastructure [8,9].

A range of quantitative methods are currently adopted for assessing accessibility, including indicator statistics [10], nearest neighbor distance [11], buffer analysis [12], network analysis, space syntax, arrival time based on cost-weighted distance [13], and the two-step floating catchment area (2SFCA) method [14].

These methods have seen widespread application in urban green space accessibility research, but their applicability to contexts such as national parks and protected natural areas is highly variable. While indicator statistics offer simplicity, they fail to capture accessibility changes in areas. The nearest neighbor distance method can indicate relative shifts in accessibility, but may not accurately represent actual travel behavior. Network analysis, while incorporating the effect of travel costs, can nonetheless offer a limited perspective. Moreover, 2SFCA models simulate individual movement, yet struggle to accommodate the multimodal transportation choices of tourists (e.g., walking, cycling, boating, or sightseeing vehicles). Space syntax effectively integrates terrain, but its application to large areas such as national parks may necessitate extensive GIS data and complex modeling to ensure accurate accessibility assessments. The buffer method offers a simple accessibility evaluation by defining areas in set distances, proving particularly insightful for identifying critical recreational and conservation zones. However, its capacity to account for the effect of existing traffic routes and terrain obstacles on accessibility may be limited. Cost-weighted distance, factoring in road networks and terrain, offers a more accurate accessibility assessment by calculating arrival times [15].

Early accessibility research largely concentrated on national, regional, and local levels. For instance, Wright Wendel et al. utilized buffer analysis to evaluate preferences, perceptions, and obstacles concerning green spaces in rapidly developing Latin American cities [16]; Mavoa et al. leveraged network analysis to study access to public transit and pedestrian destinations in Auckland [17]. Wu and Chen analyzed the spatial distribution and accessibility of 277 major tourist attractions in Inner Mongolia, applying the shortest cost-weighted distance method [18]. Recent research has significantly advanced our understanding of urban park accessibility [19,20,21,22]. Nevertheless, the accessibility of national park recreational resources (NPRRs) remains understudied. The existing literature explores accessibility in tourist spots [23,24], national parks [25,26], and protected areas [27]; however, the specific question of NPRR accessibility in national parks has received insufficient research attention.

Access to urban green spaces serves the diverse needs of city residents, while access to national parks affects broader national interests, including ecological preservation, research, education, and tourism. While researchers have analyzed the spatial accessibility of conservation areas, comparative analyses of accessibility across different protected area designations, especially those employing factor analysis, are limited. Such research is vital for effective national park management and planning. The Great Smoky Mountains National Park has improved visitor access through enhanced services such as parking and transit options in gateway communities [28]. Likewise, Fiordland National Park in New Zealand has improved access through trail development designed to accommodate a range of visitor abilities and preferences [29]. These cases demonstrate that strategies specific for a national park’s specific environment and context can enhance recreational access while maintaining sustainability. For the HTRNP, creating access strategies that consider its particular ecological and cultural characteristics is essential to balancing conservation goals with the growth of tourism.

1.2. Research Aims

Understanding the distribution and availability of recreational resources in HTRNP is crucial for realizing its recreational capacity and cultivating ecotourism. National parks, in contrast to urban parks, vary significantly in their location, purpose, amenities, and services [30], significantly affecting the distribution and availability of recreational resources. The development of recreational activities in national parks is a complex process influenced by various factors [31], including natural features such as topography and geomorphology [32], as well as the distribution of recreational facilities [33]. These factors collectively influence the quality of recreational experiences and ease of access, both of which are critical factors of recreational function. This study addresses a gap in existing research by connecting natural factors to an accessibility assessment of NPRR, deriving more accurate and robust analytical results. Calculating accessibility utilizing GIS grid distances, while accounting for the impediments posed by water bodies and mountains, can effectively model accessibility to any location in the region.

In summary, accessibility, a key measure for evaluating the success of tourist destinations, represents a critical method and cornerstone of tourism research. Accurately understanding the spatial distribution and accessibility differences of resources is fundamental to both the advancement of ecotourism and the efficient allocation of recreational resources in the HTRNP.

The draft “Hainan Tropical Rainforest National Park Ecological Tourism Special Plan (2023–2030)” identifies the HTRNP, representative of tropical rainforest ecosystems, as having exceptional biological resources suitable for ecotourism. As one of the first batch of five national parks, the HTRNP is positioned to guide and act as a model for other regions in ecotourism and environmental preservation.

To advance the understanding of current recreational resources and expand the scope of the research perspective, this study addresses the following central questions:

• What are the characteristics of the spatial distribution of recreational resources within the HTRNP? Considering the uniqueness of the Hainan tropical rainforest, is there a significant spatial difference in the distribution of different types of recreational resources within the park?
• What is the current state of accessibility of recreational resources within the HTRNP? What factors significantly influence the spatial distribution and accessibility of recreational resources within the HTRNP?

To address these questions, this paper establishes three specific aims: (i) drawing on prior research, this study employs ArcMap 10.8 software, utilizing the average nearest neighbor index, kernel density method, and ESDA spatial association analysis to appraise the spatial pattern of recreational resources; (ii) combining multiple scales of road network data, this study applies the cost-weighted distance method to evaluate the accessibility of recreational resources in the HTRNP from the perspectives of both natural and cultural resources; (iii) by indicating the accessibility patterns and influential factors of recreational resources in the HTRNP, this study aims to offer a theoretical basis for the effective utilization of recreational resources and the development of potential recreational service facilities, thus furthering the growth of neighboring communities and regions.

2. Materials and Methods

2.1. Data Sources

Data for this study were sourced from the “Hainan Tropical Rainforest National Park Natural Education Spatial Pattern and Experience System Construction” survey. This survey utilized various methods, including data collection and semi-structured interviews. Following the “National Park Resource Survey and Evaluation Standards (LY/T3189-2020)” [34], recreational resources were documented and categorized. We identified a total of 761 distinct recreational resources, which were then divided into cultural and natural categories (

Table 1. Classification and types of recreational resources.

Category/Form	Cultural Resources	Natural Resources
Objects of recreational resources that can be independently appreciated or utilized, including individual and collective recreational resource entities	Local culture, historical sites, architectural structures, landscape gardens	Geomorphic features, hydrological features, biological features, astronomical and climatic phenomena

). The “Guiding Opinions on Establishing a Nature Reserve System with National Parks as the Main Body” (https://www.gov.cn/zhengce/2019-06/26/content_5403497.htm, accessed on 23 January 2019), stipulates that human activities are generally prohibited in core protection areas and restricted in general control areas. Accordingly, utilizing the boundary line of the national park’s general control area, a dataset comprising 274 recreational resource points located in this general control area was constructed. The administrative division spatial data employed in this study were sourced from the Ministry of Natural Resources of the People’s Republic of China (https://www.mnr.gov.cn/), while the requisite transportation network spatial information, including the road network connecting various towns and villages, was derived from OpenStreetMap (OSM) (https://www.openstreetmap.org/).

2.2. Study Area

The national park is functionally zoned into general control areas and core protection areas. The general control area, while maintaining its protective function, also incorporates public services such as scientific research, education, and recreational opportunities, as well as other legally permitted activities. This study evaluates the general control area of the Hainan Tropical Rainforest National Park and its associated cities (counties) in the planning scope. The national park comprises nine cities and counties, including Wuzhi Mountain, Qiongzhong, Baisha, Dongfang, Lingshui, Changjiang, Ledong, Baoting, and Wanning, covering 4269 square kilometers, which represents 26.2% of the total land area of these nine cities and counties. The general control area itself occupies 1654.3 square kilometers, comprising 37.57% of the national park’s total area (Figure 1). With the thorough implementation of the Hainan International Tourism Island and Free Trade Port policy, along with the strategic promotion of “mountain-sea linkage, blue-green mutual reflection”, the central region has witnessed rapid growth in forest ecological tourism, rural leisure tourism, and red cultural tourism since the full opening of the Hainan central expressway in 2018. Tourist numbers have climbed significantly, indicating a strong interest in recreational activities in the central mountainous area. The ongoing construction of the road around the tropical rainforest is projected to further attract tourists and cultivate the growth and prosperity of ecological tourism activities.

2.3. Methodology

The spatial arrangement of recreational resources illustrates their clustered distribution and interactions within geospatial environments. This analysis uncovers the geographic traits of recreational activities, highlighting their interdependencies. The spatial distribution of these resources is quantified through the application of nearest neighbor index, kernel density equation, and exploratory spatial data analysis (ESDA). Additionally, buffer zone analysis and raster-weighted distance methods were utilized to assess the accessibility of recreational activities across the HTRNP’s controlled areas, as depicted in Figure 2.

2.3.1. Measuring the Spatial Configuration of Recreational Resources

• Nearest neighbor index

The nearest neighbor index (NNI) evaluates the spatial distribution of point features in a region, specifically measuring the degree of proximity between these points. The index value can be categorized into three spatial patterns: uniform, random, and clustered. In the Hainan Tropical Rainforest National Park and its adjacent counties and cities, these patterns are instrumental in describing the spatial distribution of recreational resource points [35,36]. The formula for calculating the nearest neighbor index is as follows:

$$\mathrm{R}={\displaystyle \frac{\overline{{\mathrm{r}}_{\mathrm{e}}}}{\overline{{\mathrm{r}}_{\mathrm{i}}}}}, \mathrm{r}={\displaystyle \frac{1}{2}}\sqrt[2]{{\displaystyle \frac{\mathrm{n}}{\mathrm{d}}}}$$

(1)

where r_i refers to the actual average distance between recreational resources and their nearest neighboring points, r_e denotes the distance between the nearest points in a theoretical scenario, n represents the total number of recreational resources, and d signifies the total area of the study region.

2.

Kernel density equation

Kernel density estimation (KDE) is a method for calculating the density of point or line feature measurements in a specified neighborhood. It offers a visual representation of how discrete measurements are distributed across a continuous area, based on input point data features [37,38,39]. The formula for KDE is typically expressed as follows:

$$f\left(r\right)={\displaystyle \frac{1}{nd}}{\sum }_{i=1}^{n}k({\displaystyle \frac{r-r_i}{d}})$$

(2)

where f(r) represents the kernel density value of recreational resources in the study area, n denotes the number of recreational resources in the search radius, d expresses the search radius, k illustrates the weight value, and (r − r_i) refers to the distance from recreational resource r to sample recreational resource r_i.

3.

Exploratory spatial data analysis (ESDA)

Moran’s I index, a spatial autocorrelation statistical indicator expressing the covariance of the difference between the statistical value and the mean, is utilized to evaluate the overall clustering of the causal domain [40]. The global Moran’s I index, employed in the ESDA method, was introduced to quantitatively assess the overall characteristics of the spatial structure of regional recreational resources [41]. This index effectively measures the distribution of spatial elements, thus reflecting the average degree of spatial clustering of similar characteristics in a given area. Its formula is presented as follows:

$$I={\displaystyle \frac{\sum _{i=1}^n\sum _{j=1}^n\left(X_i-\overline{X}\right)\left(X_j-\overline{X}\right)}{S^2\sum _{i=1}^n\sum _{j=1}^n{W}_{ij}}}$$

(3)

where n represents the total number of spatial units, X_i and X_j denote the attribute values of the i spatial unit and the j spatial unit, respectively, which indicates the mean value of the attribute values of all spatial units, and W_ij expresses the spatial weight value.

4.

Spatial relevance analysis of accessibility distribution

The local correlation index, Getis-Ord Gi*, is utilized to detect spatial autocorrelation in the distribution of elements as a tool for analyzing the spatial distribution of hot- and coldspots of recreational resources in the Hainan Tropical Rainforest National Park [20]. The formula for this calculation is presented as follows:

$$\begin{array}{c}{{\mathrm{G}}_{\mathrm{i}}\left(\mathrm{d}\right)}^2={\displaystyle \frac{\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{W}}_{\mathrm{i}\mathrm{j}}\left(\mathrm{d}\right){\mathrm{X}}_{\mathrm{j}}}{\sum _{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{X}}_{\mathrm{j}}}}\end{array}$$

(4)

where ${\mathrm{X}}_{\mathrm{j}}$ is the observed value of attraction X in cell j, W_ij represents the distance weight matrix between attractions i and j, and n denotes the total number of recreational resource sites in the HTRNP.

2.3.2. Assessing Accessibility

• Buffer zone

Buffer analysis leverages geographic features (points, lines, or polygons) as a basis for creating buffer zones of a specified width concerning these features, thus generating new polygon layers [5,12,14]. Overlaying the buffer layer with the target analysis layer produces the desired results. This spatial analysis tool is commonly utilized to address proximity-related questions. Considering the distribution of recreational resources in national parks, this study adopts transportation features, including national highways, provincial highways, county highways, township roads, and urban roads, in the nine counties and cities containing the national park, as line features. A line-based buffer analysis is performed utilizing the spatial analysis capabilities of ArcGIS 10.8 to evaluate the accessibility of the transportation network in the study area. Tourist travel behavior is influenced by transportation mode availability, the road hierarchy, and the density of the road network. For instance, the availability of higher-level roads and convenient transportation options can increase tourists’ willingness to travel. Accordingly, buffer radii are set at 2 km for national and provincial highways, 1.5 km for county highways, and 1 km for township and urban roads.

2.

Arrival time based on cost-weighted distance

Accessibility is strongly influenced by several factors, including the condition of transportation infrastructure, the spatial distribution of destinations, and the level of regional development. Various methods exist for measuring accessibility; this paper employs time cost analysis to explore the time cost required for any location in the study area to reach recreational resources [14,42]. The formula is as follows:

$$A_i=\min \left(M_j{T}_{ij}\right)$$

(5)

where A_i represents the time cost for a point in the target area, i denotes any point in the region, M_j denotes the weight value for recreational resource j and the constant value of 1 is assigned, and T_ij expresses the travel time cost for the shortest route from point i in the target area to any recreational resource j.

In accordance with the “Technical Standard of Highway Engineering (JTG B01-2014)” [43], published by the Standardization Administration of China on 1 January 2014, and considering the natural environment of the HTRNP and prevailing traffic conditions, average travel times (in minutes) per kilometer on different road classes were established as the time cost values for accessibility measurement in a raster grid. These values were further optimized utilizing previous studies [35] (

Table 2. Velocity and cost of the cantonal land traffic network.

Road Type	National Highway	Provincial Highway	County Highway	Urban and Township Road	Waters	Other
Speed (km·h−1)	80	60	40	30	10	2
Time cost (min)	0.75	1	1.5	2	6	30

). Then, the road vector map in the study area was rasterized at a resolution of 1 km × 1 km and assigned corresponding values. When modeling water transportation, the tourism value and accessibility of water bodies must be considered. Therefore, a default speed of 10 km/h was assigned to water routes; whereas, for areas outside of water bodies and excluding established land routes, a speed of 2 km/h was applied to reflect the reduced pace of overland travel. In instances of overlapping transportation routes, route selection for time cost calculations was based on observed travel patterns. For areas beyond the established road network, the time cost was calculated based on a pedestrian travel speed of 30 min per kilometer.

3. Results

3.1. Spatial Distribution Characteristics of Recreational Resources

3.1.1. Spatial Agglomeration Features

The spatial distribution of point features is typically assessed utilizing the nearest neighbor index (NNI), which classifies distributions as random, clustered, or uniform. In the general control area of the HTRNP, 274 recreational resources are present, comprising 78 cultural and 196 natural resources (Figure 3). Analysis with the average nearest neighbor tool in ArcGIS 10.8 yielded an NNI (R) of 0.276535 for these resources. This value, significantly less than 1, coupled with a Z score of −22.909940, indicates a strongly clustered spatial distribution (

Table 3. Citywide recreational resources nearest neighbor indexes.

	Average Nearest Distance (km)	Expected Nearest Distance (km)	NNI	z Value	p Value	Distribution Type
Baisha	325.7944	1210.4638	0.269148	<0.001	11.185363	Clustered	64
Baoting	961.387	1947.4727	0.493659	<0.001	−5.564564	Clustered	33
Changjiang	454.3195	1138.2739	0.39913	<0.001	−6.702714	Clustered	34
Dongfang	1283.5349	1470.3087	0.87297	0.243827	−1.165473	Random	23
Ledong	6862.3489	3386.8455	2.026177	<0.001	3.926298	Dispersed	4
Lingshui	844.8666	1497.57	0.564157	<0.001	−3.998756	Clustered	23
Qiongzhong	824.2763	1628.2455	0.51	<0.001	−5.343495	Clustered	32
Wuzhishan	725.006	1832.5476	0.395627	<0.001	−9.030267	Clustered	61
Recreational Resources	677.7656	2450.925353	0.276535	<0.001	−22.90994	Clustered	274

NNI: nearest neighbor index; z-value: probability; p-value: multiple of standard deviation.

). Of the eight counties and cities, Ledong Li Autonomous County presents a uniform distribution, while Dongfang City exhibits a random distribution. The remaining five counties and cities, each with an NNI (R) below 1, demonstrate clustered distributions, likely attributable to differences in geography and resource availability.

3.1.2. Spatial Distribution Characteristics

To better visualize the spatial distribution typologies, ArcGIS 10.8’s Kernel density tool was employed to analyze the density of recreational resources (Figure 4). The resulting spatial distributions indicate significant difference in the distribution of these resources across jurisdictions in the general control area, exhibiting a broadly polycentric distribution with clear spatial clustering. The distribution of recreational resources in HTRNP is strongly associated with the regional environment and resource availability. Baisha Li Autonomous County has the highest concentration of recreational resources (64 sites), while Ledong Li Autonomous County contains the fewest (four sites). Spatially, these resources are concentrated in the Wuzhishan and Yinggeling mountain ranges. The Yajialing area of Changjiang County also represents a significant locus of recreational resources, benefiting from the abundant natural resources associated with these mountain ranges. In addition, Wuzhishan City, Baisha Li Autonomous County, and Changjiang Li Autonomous County, regions known for the preservation and promotion of Li and Miao cultures, also exhibit high kernel density values for recreational resources, suggesting a strong correlation between resource distribution and regional cultural heritage.

3.1.3. Recreational Resource Spatial Correlation Analysis

To analyze the spatial distribution pattern of recreational resources in the national park, this study employed exploratory spatial data analysis (ESDA) and utilized the spatial autocorrelation tool in ArcGIS 10.8 to calculate Moran’s I statistic (I) for all recreational resources. The results (

Table 4. Recreational resources global Moran’s I index.

Parameter	Value
Global Moran’s I index	0.360214
z value	13.201553
p value	<0.001
Variance	0.00076
Expectations index	−0.003663

p-value: determines the probability that the observed results are due to chance. Variance: quantifies the spread of data points. Expectations index: compares observed values to theoretical expectations. z-value: reflects the number of standard deviations separating the observed value from the mean, indicating statistical significance.

), as presented in the table, indicate a global Moran’s I estimate of 0.360214, a value slightly greater than zero, and a Z value of 13.201553, exceeding 1.96. This suggests a lack of significant spatial autocorrelation in the distribution of recreational resources. Further analysis of hotspot classified the 274 recreational resources into five categories, as illustrated in Figure 5. The spatial distribution of recreational resources demonstrates significant differences between hotspot and coldspot areas, with these differences being particularly significant at the township scale. In general, the hotspot areas for cultural resources exhibit a relatively concentrated pattern, especially in Shenling Town; whereas, the hotspot areas for natural resources are distributed across Donghe Town, Wangxia Town, and Tongshe Town, where the aggregation of these resources may be attributable to the local natural environment and cultural characteristics.

3.2. Accessibility Analysis

3.2.1. Spatial Characteristics of Recreational Resource Traffic Accessibility

This study analyzed the traffic accessibility of 274 recreational resource points in the HTRNP. The analysis indicated that 54.7% of these points are located in the park’s buffer zone (Figure 6). This distribution suggests a relationship between the transportation network and the location of recreational resources, indicating potential areas for improvement. Specifically, the concentration of over half of the tourism resource points in the buffer zone highlights the need for enhanced traffic connectivity and improved resource accessibility in the park’s core area.

The accessibility of recreational resources directly relates to visitors’ ability to enjoy the diverse recreational experiences offered in the park. A convenient and efficient transportation network can significantly improve visitor satisfaction and cultivate the sustainable development of the national park. Based on the findings of this study, it is recommended that recreational route planning in the national park prioritize enhancing the accessibility of these resources. For instance, the addition of road signs in areas requiring pedestrian access to recreational areas could enhance the accessibility of resource points, thereby promoting a balance between resource protection and visitor utilization.

3.2.2. Characteristics of Recreational Resource Time Accessibility

This study indicated spatial accessibility differences by classifying travel times to the tropical rainforest national park into five segments and mapping the temporal distribution of recreational resources in HTRNP (Figure 7). The proportional area of each segment was calculated, and the frequency distribution for each time segment is demonstrated in

Table 5. Distribution frequency and cumulative frequency of recreational resource accessibility.

Time (min)		0~60	60~120	121~180	181~300	300~516
Cultural Recreational Resources	Distribution Frequency	32.01%	39.72%	16.25%	10.49%	1.52%
	Cumulative Frequency	32.01%	71.74%	87.99%	98.48%	100.00%
Nature Recreational Resources	Distribution Frequency	31.47%	40.30%	16.97%	10.00%	1.26%
	Cumulative Frequency	31.47%	71.76%	88.74%	98.74%	100.00%
Recreational Resources	Distribution Frequency	35.38%	39.40%	14.92%	9.14%	1.16%
	Cumulative Frequency	35.38%	74.78%	89.70%	98.84%	100.00%

.

The data indicate that 98% of recreational resource points are readily accessible in 3 h, with a mean travel time of 91 min. Cultural recreational resources exhibit a slightly longer mean travel time of 98 min compared to natural recreational resources. This difference may be attributable to the clustered distribution of cultural resources and existing transportation infrastructure; whereas, natural recreational resources are situated in areas reflected by unique topography and geomorphology, often further from population centers and transportation networks, resulting in comparatively reduced accessibility.

A significant proportion of recreational resources can be reached in 1 to 2 h, suggesting the ease of short-trip recreational activities. Nevertheless, 1.16% of the area’s recreational resources require travel times greater than 6 h, potentially hindering the recreational value of these locations.

Among the nine counties and cities covered by the Hainan Tropical Rainforest National Park, Wuzhishan City offers the greatest ease of access to recreational resources, establishing a high-accessibility zone. This superior access results primarily from Wuzhishan City’s robust transportation infrastructure and central location in the park, its proximity to major transportation routes significantly shortening travel times for visitors; whereas, Changjiang Li Autonomous County, Dongfang City, Ledong Li Autonomous County, and Wanning City present comparatively limited access to recreational resources, constituting low-accessibility areas. Wanning City’s restricted access is due in part to the small area of the national park being in its administrative boundaries. The dispersed nature of recreational resources in the park also contributes significantly to the lower accessibility values observed in these areas. The quality of transportation infrastructure and the spatial pattern of recreational resources are crucial factors of recreational resource accessibility across these counties and cities. Improvements to transportation networks and the optimization of the layout of recreational resources can significantly enhance access to recreational resources in the HTRNP.

4. Discussion

This study covers nine cities and counties within the HTRNP planning area, providing an in-depth analysis of the spatial distribution and accessibility of recreational resources.

4.1. Investigating the Spatial Pattern of Recreational Assets in the HTRNP

Recreational resources in the Hainan Tropical Rainforest National Park exhibit a clustered spatial pattern, with a greater concentration of cultural resources compared to natural resources. Statistical analysis indicates a primarily mountainous distribution, resource richness increasing with elevation, consistent with prior research [44,45,46] demonstrating a positive association with topography.

However, the distribution of recreational resources demonstrates no strong correlation with water system characteristics, contrary to studies [47,48] emphasizing the effect of water systems on accessibility. This difference may arise as resource distribution in the park is more closely associated with rainforest characteristics such as forest cover, terrain, and biodiversity than to water systems. In addition, human development and planning likely contribute significantly to the distribution of these resources. Seasonal variations significantly influence the number of tourists engaging with aquatic recreational resources [49]. Future research could benefit from examining accessibility through the lens of seasonal fluctuations in the spatial distribution of these resources. Such an approach would likely yield a more nuanced and optimized understanding of tourist patterns and resource utilization.

4.2. Assessing the Accessibility of the HTRNP’s Recreational Features

Accessibility analysis demonstrates that recreational resources in the buffer zone are readily accessible, while those beyond may be less accessible due to transportation constraints. Consistent with Templeton et al. [50], who demonstrated that enhanced electric vehicle infrastructure improves accessibility in the U.S. National Park System, transportation infrastructure directly affects recreational resource accessibility.

Furthermore, 98% of recreational resource sites are accessible within a three-hour timeframe, averaging 91 min to reach. Cultural recreational resources have an average accessibility time of 98 min, marginally higher than that for natural resources. The integration of cultural resources with the road network appears more effective than for natural resources, suggesting that road grade could significantly affecting the spatial decay rate of accessibility. Our analysis drew upon diverse data sources to ensure accuracy and comprehensiveness. However, data acquisition limitations and spatial constraints confined our focus to road networks, excluding air traffic, which potentially restricted our accessibility assessment. Future studies could be improved by integrating further data, including air travel and informal transportation methods such as personal vehicles, taxis, and motorcycles, to achieve a more complete perspective.

The cultural heritage of ethnic minorities residing in the national park significantly enhances access to recreational resources, corroborating findings from Meng et al. [51] on traditional ethnic villages. Parallel benefits are evident in the Kinabalu UNESCO Global Geopark [52], where tourism and park development have improved social and economic conditions. Cultural features, including local customs, historical accounts, distinctive architecture, cuisine, and souvenirs, collectively and subtly shape the distribution of access in tourist areas, contributing to a more equitable experience.

Moreover, access to recreational resources is connected to factors such as national park branding and value enhancement, the effectiveness of management and service delivery, financial administration, collaborative development and resource sharing, and comprehensive oversight [53,54,55,56,57]. The distribution and accessibility of recreational resources are further affected by nature education programs [58], including the development and administration of nature education centers and facilities. Further research is necessitated to determine the accurate means in which these elements affect access to recreational resources.

4.3. Strategies for Enhancing Accessibility

While the external ring road encircling the HTRNP remains incomplete, and despite relatively good overall accessibility in the park’s constituent counties and cities, access to recreational resources in peripheral areas is poor, marked by significant internal differences. This long-term constraint will evidently hinder the full development of recreational activities. Both Wuzhishan City and Baisha Li Autonomous County have rich concentrations of recreational resources, complemented by advantageous natural environments and robust socio-economic conditions. Adhering to the “mountain-sea interaction, internal and external linkage, community participation, and resource utilization” development concept of the “Hainan Tropical Rainforest National Park Ecological Tourism Special Plan”, and building upon existing resource endowments and land use planning, simultaneous improvements in both resource utilization and transportation infrastructure are feasible. To improve resource utilization, existing infrastructure can be employed to identify and develop usable resource points in the general control area, supporting ecological tourism activities, expanding into adjacent areas surrounding the national park to enlarge the available touring area, and further enhancing tourism service facilities.

Research has consistently shown that high service quality positively influences tourist satisfaction and loyalty [59,60]. This evidence implies that by improving the accessibility and quality of recreational resources in the Hainan Tropical Rainforest National Park (HTRNP), we could expect a similar rise in visitor contentment and the likelihood of their return visits. Expanding recreational venues facilitates the alleviation of the imbalanced spatial distribution of recreational resources, promotes a cohesive relationship between the general management area and surrounding communities, and establishes a coordinated overall distribution. Regarding transportation infrastructure development, gateway communities are critical access points to the national park. Integrating these communities, increasing road network density, optimizing road network structure, comprehensively planning regional resources, and improving the safety of recreational traffic will create multiple connections between recreational resources and existing major transportation routes. In addition, strengthening internal transportation infrastructure in the national park, enhancing scenic spot wayfinding signage, developing resources with local characteristics, and enhancing high-quality recreational offerings are also essential.

5. Conclusions

Through GIS technology and raster data analysis, this study offers a new quantitative method for managing recreational resources in the HTRNP. It explores park accessibility, connecting theoretical foundations with practical application through empirical findings, thus establishing a basis for sustainable park management. The study’s significance is highlighted by its direct applicability to resource management and policy development, ensuring tangible impact. In addition, its interdisciplinary approach offers novel perspectives on resource evaluation and planning in national parks, especially concerning the ecologically sensitive HTRNP, emphasizing its crucial contribution to both conservation efforts and tourism growth.

Moreover, we recommend that future work should appraise the effect of tourists’ socio-economic backgrounds, cultural preferences, and policy and planning considerations to achieve a holistic understanding of the factors influencing accessibility. These improvements will create a more robust foundation for the sustainable management and planning of national parks.