Exploring the spatial patterns and influencing factors of rural tourism development in Hainan Province of China

Spatial distribution characteristics of the rural tourism spots in Hainan Province

Spatial distribution type characteristics

Results presented in Table 1; Fig. 4 summarize the findings of an average nearest-neighbor analysis on the spatial distribution of Rural Tourism Spots, revealing a clear pattern of clustering. The analysis compares the expected average distance between each tourism spot and its nearest neighbor—calculated as 8.52 km—with the observed average distance of 7.07 km. The observed distance is notably smaller than the expected distance, suggesting that these tourism spots are closer to one another than would be expected in a random distribution.

This difference is quantified through the nearest neighbor index, which is 0.83. Since the index is less than 1, it indicates a pattern of clustering rather than random or evenly spaced placement of the spots. The spatial structure type derived from this index confirms that the distribution of the tourism spots is “agglomerated” or clustered. This clustering pattern is further validated by the z-score of −4.03, which measures how much the observed pattern deviates from randomness. The negative z-score indicates that the observed clustering is statistically significant and unlikely to have occurred by chance. Additionally, the p-value of 0.000055 provides further confirmation that this clustering is statistically meaningful. The results indicate that these sites tend to be located near each other across Hainan Province, reflecting geography, economy, or infrastructure that shape the placement of tourism zones, often clustering in areas rich in resources or experiencing intense tourist interest.

Analysis results of the nearest neighbor index of the Rural Tourism Spots.

Spatial distribution density characteristics

Based on the nuclear density analysis method described earlier, all 154 Rural Tourism Spots in Hainan Province were analyzed as point data utilizing the ArcGIS 10.7 spatial analysis feature to determine the kernel density, as illustrated in Fig. 5. The analysis reveals a distinctive distribution pattern of these tourism spots across the province. The spots are primarily concentrated in several key areas: the southern core region encompassing Sanya, Baoting, and Lingshui, the central core area around Haikou and Chengmai, the Baisha Li and Miao Autonomous County, as well as Danzhou and Tunchang. These areas have more abundant natural landscapes and cultural heritage, and the richness of these resources provides a good foundation for rural tourism. In contrast, there are also low-density, scattered spots in the peripheral regions of the study area. These findings highlight a clear spatial differentiation, with clusters, bands, and dispersed areas coexisting across the province.

Specifically, the highest nuclear density of Wuye-level Rural Tourism Spots is found in the eastern part of Tianya District and the western part of Jiyang District in Sanya City, as well as in Boao Town in Qionghai City. Additionally, two secondary clusters are located in Meilan District and Cheng District in Haikou City. In the northern part of Mai County, there is a notable point-like agglomeration. These concentrated areas are all situated within the administrative boundaries of Hainan Province. The primary tourist hotspots, identified as four-coconut-level destinations, are concentrated in the northern part of Chengmai County, the northern areas of Haikou, the central region of Baoting County, and the southern part of Wanning City. The remaining areas along the central southern coastline exhibit a more dispersed distribution. Significantly, the distribution of rural tourism locations in these central zones exhibits considerable variation in density. For Sanye-level Rural Tourism Spots, the highest concentration is found in the Xiuying District of Haikou and the northern part of Chengmai County, while three additional subcore areas appear along the boundaries of Changjiang Li Autonomous County, Ledong Li Autonomous County, and Wuzhishan City, as well as in the central area of Wenchang City. These subcore areas form an L-shaped distribution pattern. Two- and one-coconut-level rural tourism spots are concentrated in Baoting Li and Miao Autonomous County and Wuzhishan City, respectively, with the one-coconut-level category having the fewest rural tourism spots among the five categories.

Overall, the rural tourism spots in Hainan are primarily concentrated in Baoting County, Haikou City, Wenchang City, Wuzhishan City, and Chengmai County. These regions have obvious advantages in natural resources, infrastructure, economic development, culture and history. Among them, Wuzhishan City is famous for its unique tropical rainforest resources, while Wenchang City attracts tourists with its beaches and cultural sites. The distribution in Wuzhishan and Baoting shows a T-shaped pattern, with the remaining spots scattered Spanning the central and eastern areas of the province. This analysis underscores the uneven geographical distribution of rural tourism spots in Hainan, with certain areas exhibiting higher concentrations while others remain more sparsely populated.

Core density analysis results.

Spatial distribution characteristics

For an exhaustive examination of rural tourism locations in Hainan Province, the standard deviation ellipse technique was applied, using the quantity of study subjects as a basis. This method offers a powerful means to visually represent the spatial distribution and development structure of the rural tourism spots. By utilizing this approach, the distinct features of the distribution space are diminished in complexity—both longitudinally and latitudinally—become more apparent, highlighting the patterns in the distribution of tourism sites.

As depicted in Fig. 6, the first-level standard deviation ellipse, which encompasses the centroids of the point elements, includes approximately 68% of the total rural tourism spots. The analysis reveals that the distribution direction of the Rural Tourism Spots follows a southwest-northeast trajectory. The directional angle is 45.09°, with a flatness value of 0.34, which suggests a clear directional tendency in the spatial arrangement of these sites. Notably, the ellipse covers approximately 48% of Hainan’s mainland area, indicating that the rural tourism spots are relatively concentrated in certain regions. However, the distribution along the outer edge of the ellipse becomes progressively more sparse, a phenomenon consistent with the findings from the nuclear density analysis. The central point of this ellipse is located in Shiling Town, Qiongzhong Li and Miao Autonomous County, which is notably distant from Haikou, the capital of Hainan Province.

Regarding the distinct types of rural tourism spots, the five-coconut-level spots exhibit a flatter distribution, with a flatness value of 0.44 and a directional angle of 41.97°. The center of this distribution is located in Fengmu Town, Tunchang County. Compared to the overall rural tourism spots, the directional distribution of the five-coconut-level spots is even more pronounced, demonstrating a clear spatial concentration. For the other levels, the flatness values are as follows: four-coconut-level spots (0.35), three-coconut-level spots (0.34), two-coconut-level spots (0.23), and one-coconut-level spots (0.86). The corresponding directional angles are 30.98°, 51.37°, 38.86°, and 62.21°. Among these, the three- and one-coconut-level spots display particularly strong directional tendencies, following a southwest-northeast distribution. The centers of these categories are located near Xiangyong Town, Baoting County, and Hongmao Town, Qiongzhong County, both situated in the southern region of the province.

Overall, the arrangement of the rural tourism spots at all levels indicates a significant concentration along a southwest-northeast axis. This directional pattern suggests that the rural tourism development in Hainan is not only spatially concentrated in certain core areas but also aligns with the geographic and cultural landscapes of the province. In particular, rural tourism spots with a rating of three-coconut level and above tend to cluster around specific central areas in Qiongzhong Li and Miao Autonomous County and Baoting Li and Miao Autonomous County. These areas, which are more advanced in tourism, are crucial in defining the rural tourism scene of the province.

Standard deviation ellipse analysis of the Rural Tourism Spots in Hainan Province.

Spatial distribution equilibrium analysis

The nearest neighbor index and nuclear density analysis offer significant understanding of the overall trend and distribution hotspots of rural tourism spots in Hainan Province. However, these methods alone do not offer a precise assessment of the degree of spatial dispersion across different cities. To delve deeper into the spatial arrangement and to assess the level of regional variations, the Gini coefficient and imbalance index are introduced. These statistical measures allow for a more refined examination of the dispersion of rural tourism spots across the geographical subregions of Hainan, offering a more distinct view of the spatial inequities and concentration of tourism development within the province.

(1) Gini coefficient.

According to the method to accurately portray the differences in the distribution of rural tourist sites in Hainan, and to eliminate the problem of imprecision in the contribution rate of the differences caused by the net differences. Substituting the appropriate values into the equation, we obtain G = 0.9738 and C = 0.0261. Therefore, the arrangement of the rural tourism spots in the cities and counties is uneven, and the gap is large.

(2) Imbalance index.

To better understand the uneven distribution of rural tourism spots across Hainan, this study employs the imbalance index for a more precise calculation. The rural tourism spots in each region are ranked in descending order based on their proportion to the total number of spots in the province. The data for each region are summarized in Table 2. By substituting n = 18 into the equation, an imbalance index (D) value of 0.22 is obtained. When combined with the Gini coefficient, the results indicate a significant degree of imbalance in the distribution of rural tourism spots across the province, with certain areas hosting a disproportionately large number of spots.

Table 2 illustrates the distribution of rural tourism spots across various regions of Hainan Province, providing a detailed breakdown of the number of spots, their proportional representation relative to the total, and the cumulative proportion. Chengmai and Wenchang stand out as the areas with the highest concentration, each hosting 14 rural tourism spots, which account for 9.09% of the total rural tourism spots in the province. These two regions together contribute to 18.18% of the total, showing a significant concentration of tourism resources. Following these, Haikou and Baoting each have 12 spots, contributing 7.79% to the total and together account for 33.77% of all rural tourism spots.

The cumulative proportion continues to increase, as smaller areas such as Wuzhishan and Danzhou are included, with the cumulative percentage reaching 53.90% after accounting for these regions. The distribution progressively spreads across the province, with areas such as Qiongzhong, Qionghai, Ding’an, and Lingao having between 5.00 and 5.19% of the total rural tourism spots, contributing to a cumulative 85.71%. At the bottom of the list, Tunchang, with just 6 rural tourism spots, accounts for 3.90%, pushing the cumulative proportion to 92.86%. This distribution pattern highlights the uneven concentration of rural tourism spots in a few core areas, such as Chengmai and Wenchang, while other regions are more sparsely populated with such spots, indicating an imbalanced spatial distribution across the province.

Factors influencing the spatial distribution of rural tourism spots in Hainan Province

Traffic factors

The primary sources of tourists for rural tourism in Hainan largely stem from nearby cities, with most urban residents traveling by private car. As a result, the layout of highways and other transportation routes has a significant impact on the spatial distribution of rural tourism spots. In recent years, Hainan Province has made substantial investments in infrastructure development, particularly in road and rail networks, to support and enhance transportation connectivity. Expressways, in particular, have played a crucial role in driving economic growth. According to statistics, as of September 2019, the total operational railway mileage in Hainan exceeded 1,170 km, while the road network spanned more than 1,160 km. The accessibility of rural roads has steadily improved, fostering greater connectivity to rural tourism sites. Plans for the province’s highway network aim to strengthen the existing infrastructure by enhancing the two-pole radiation phenomenon, expanding coverage to the central and western regions, and improving transport capacity in key areas. This strategy also focuses on improving the integration of various transportation modes, further boosting the overall transportation network and facilitating easier access to rural tourism destinations across Hainan.

The data in Table 3 above displays the outcomes of a Pearson correlation examination linking the evaluations of rural tourist attractions (ranging from one to five coconuts) and various factors associated with those spots. The table lists the Pearson correlation coefficients, the significance levels (two-tailed), and the sample size (N = 18) for each rating category. The Pearson correlation coefficient assesses the intensity and orientation of the association between the evaluations of rural tourist locations and the corresponding variables. A positive correlation denotes a mutual increase, whereas a negative correlation implies an opposing trend. The closer the correlation value is to 1 or −1, the stronger the relationship. A correlation value close to 0 indicates little or no relationship. The significance values assess whether the observed correlations are statistically significant. A significance value less than 0.05 typically suggests that the correlation is meaningful, i.e., it is unlikely to have occurred by chance.

In the table, the correlation values for the one to five-coconut categories are relatively low, with no values below 0.05 for significance, indicating that the relationships between the ratings and the associated factors are not statistically significant. For example, the correlation for one-coconut-rated spots is −0.141 with a significance of 0.577, which is not statistically significant. Similarly, the other categories show weak correlations and non-significant results (all above 0.05 for significance).

However, the category of four and five-coconut-rated spots stands out. The correlation value for this group is 0.477, which suggests a moderate positive relationship and the significance value is 0.045, which is statistically significant (below the 0.05 threshold). This indicates that there is a meaningful positive relationship between the ratings of these higher-level tourist spots and the associated factors, implying that factors influencing the ratings of four and five-coconut rural tourism spots are statistically significant and stronger compared to lower-rated spots. Thus, the analysis shows that the factors influencing the highest-rated rural tourism spots in Hainan Province are more closely linked and exhibit a stronger, significant correlation than those affecting the lower-rated spots.

Economic factors

The level of economic advancement within a region significantly influences the limitation or facilitation of local tourism growth, a phenomenon particularly evident in the following two dimensions. First, economic development stimulates demand for quality tourism. The primary origin of rural tourism demand is typically from neighboring regions. Residents in more economically prosperous areas tend to have greater incomes and a higher propensity and capacity for spending, which fosters the emergence and growth of tourist-generating markets. Second, in counties and cities with rapid regional economic development, the government has increased investment in infrastructure construction. The continuous improvement of the transportation system has promoted an increase in the willingness of tourists to travel, thus generating sustainable tourism behavior. This is one of the reasons why most rural tourism spots are located near economically well-developed areas.

Table 4 shows that the correlation between rural tourist spot ratings and associated factors is generally weak or insignificant for lower-rated spots (one to four coconuts). For instance, the correlation for one-coconut-rated spots is −0.241 (p = 0.336), indicating a weak negative relationship, which is not significant. Similarly, the correlations for two- and three-coconut-rated spots are also weak and not statistically meaningful. However, for five-coconut-rated spots, the correlation is 0.575 with a significance of 0.012, which is statistically significant. This indicates a moderate positive relationship between high-rated rural tourism spots and associated factors, suggesting that the factors influencing top-rated rural tourism areas have a more substantial and meaningful impact.

Table 5 reveals the correlation between the per capita disposable income of urban residents, the proportion of the tertiary industry, and the ratings of rural tourist spots. For lower-rated spots (one to four coconuts), the correlations with economic variables are weak and not statistically significant. However, for five-coconut-rated spots, there is a strong positive correlation (0.610) with a significance of 0.007, indicating a statistically significant relationship between higher income, a larger tertiary sector, and top-rated rural tourism areas. This suggests that higher-rated rural tourist spots are more closely associated with economic prosperity. The reason for this is that favourable economic benefits can further promote the transformation and upgrading of rural tourism, thus accelerating the construction of tourism resource sites and influencing the spatial pattern of rural tourism resource sites and the development of tourism.

The correlation analysis results in Table 6 show the relationship between the proportion of the tertiary industry and the ratings of rural tourist spots (from one to five coconuts). For rural tourism spots with ratings of one to four coconuts, there are weak or moderate correlations with the tertiary industry’s proportion, none of which are statistically significant. However, for five-coconut-rated spots, there is a strong positive correlation of 0.582 with a significance of 0.011, indicating a statistically significant relationship between these top-rated spots and the size of the tertiary industry.

Table 7 shows that for most rural tourism spot ratings, there is no significant correlation between the permanent population and the ratings. However, for five-coconut rated spots, the correlation is 0.530 with a significance of 0.024, indicating a moderate positive relationship. This suggests that areas with larger permanent populations tend to have higher-rated rural tourism spots, while the correlation is weak and insignificant for lower-rated spots. Population-rich areas can provide more diversified tourism products and services to meet the needs of different tourists, net attract more tourists, and promote the development of rural tourist attractions.

The correlation analysis presented in Table 8 shows no significant relationship between the rural population and rural tourism ratings for most categories, with weak correlations across the board. The five coconut-rated spots have a positive correlation of 0.415, but this is only marginally significant at the 0.086 level. Overall, the rural population appears to have minimal influence on the ratings of rural tourist spots.

The analysis of the permanent and rural population indicators, derived from population-scale elements, reveals a strong connection between the number of five coconut-rated tourism spots and these indicators. Specifically, the permanent population is significantly positively correlated with the ratings, indicating that areas with larger populations tend to have better-developed rural tourism spots. Since most rural tourists are urban residents, factors like disposable income, GDP, and population size (all from 2019) were selected as key standards. Using ArcGIS 10.7, relevant data for each city and county were mapped onto the vector map of Hainan Province, with different attributes represented by color classifications, as shown in Fig. 7.

The results indicate a positive correlation between the distribution of rural tourism spots and the disposable income and population of urban residents. This pattern is particularly evident in places like Haikou, Chengmai, and Wenchang, which aligns with the earlier findings from the nuclear density hotspot analysis. However, despite the relatively low per capita disposable income and lower population density in the central region, areas like Baoting County and Wuzhishan City still have a concentrated distribution of rural tourism spots. This is likely because these locations are in close proximity to Sanya, a popular tourist destination, where higher disposable incomes lead to greater demand for nearby leisure travel, especially for weekend trips.

Disposable income of urban residents, population, GDP and distribution map of the rural tourism spots.

Topographical factors

The growth of rural tourism is fundamentally connected to the regional natural environmental assets, and the enduring nature of rural tourism locations is intricately bound to the maintenance and conservation of the natural environment. Situated at the southernmost tip of China, Hainan is a key area for tropical climate conditions, accounting for roughly 40% of China’s tropical climate zone. This favorable environment provides abundant sunlight, warmth, and water resources, as well as rich animal, plant, and marine resources. Geomorphologically, Hainan features a relatively flat landscape on its peripheries, with a central highland region. This circular, stratified landform has a cascading distribution of elevations, all benefiting from ample sunlight and heat. Wuzhi Mountain, which rises to over 1,500 m, is Hainan’s highest peak, while the nearby Yinggeling Mountain is the second-highest.

The distribution of rural tourism spots in Hainan Province is closely linked to the region’s terrain. By using the 3D analysis tool in ArcGIS 10.7, a terrain elevation map of Hainan was overlaid with the three major rivers and administrative boundaries (Fig. 8). The results show that, with the exception of a few coconut-rated rural tourism spots near Wuzhi Mountain, most tourism sites are situated in flatter, lowland areas. Many rural tourism spots in the central region are located in the plains between mountain ranges, characterized by diverse terrain, a pleasant climate, and abundant natural resources. These areas, with their interspersed mountains and rivers, offer picturesque landscapes, which are a key attraction for tourists.

In contrast, most tourist spots are concentrated along the outskirts of urban centers. Many cities and counties in Hainan are autonomous regions inhabited by various ethnic minorities, preserving unique cultural characteristics in their natural living environments. From a tourism market perspective, the central mountainous region is conveniently located near Sanya, making it easier to integrate into the larger Sanya tourism circuit. Sanya, a popular destination for both domestic and international tourists, provides significant resources for nearby rural tourism spots. The western region, where rural tourism spots are more scattered, also predominantly features low-lying areas. This reflects the locality and terrain dependence of rural tourism resources development. Consequently, topography plays a crucial role in shaping the spatial distribution of rural tourism spots in Hainan.

Distribution map of the elevation and rural tourism spots in Hainan Province.

Tourism development level

The progression of rural tourism is intricately linked to the regional tourism setting, which has a substantial impact on the spatial layout of rural tourist spots. Several factors contribute to this relationship. First, different types of tourist attractions require varying levels of infrastructure. Tourist spots within a given area can share infrastructure, so the comprehensive advancement of regional tourism fosters the creation and expansion of rural tourist areas. To fully utilize existing infrastructure, these areas tend to remain clustered together.

Second, the existence of premium A-grade tourist sites in the area is pivotal in influencing the evolution of adjacent rural tourism locations. High-level attractions, despite the intense competition they face, often benefit from a steady flow of visitors, strong reputations, and well-developed infrastructure. This not only drives tourism in the immediate vicinity but also positively impacts nearby rural tourism locations by sharing resources such as facilities and transport networks.

Moreover, the existence of high-level attractions can stimulate regional economic growth, increase local employment opportunities, and shift public attitudes toward tourism development, encouraging greater community participation. The management practices and operational expertise gained from these high-profile sites can also be applied to nearby rural destinations, further fostering the creation of new tourism spots and influencing the broader spatial layout of rural tourism in the area.

The correlation analysis in Table 9 explores the relationship between the availability of hotel facilities and the ratings of rural tourist spots. For one-coconut-rated spots, the correlation is weakly negative (−0.221) with a significance level of 0.379, indicating no statistically significant relationship. Similarly, the correlations for two-coconut-rated spots (−0.309) and three-coconut-rated spots (−0.082) are both weak and not statistically significant, with p-values above the 0.05 threshold. For four coconut-rated spots, the correlation is slightly positive (0.135), but again, the p-value of 0.592 shows that the relationship is not significant. However, for five coconut-rated spots, the correlation is strongly positive (0.687) and statistically significant with a p-value of 0.002. This suggests a strong association between the quality of hotel facilities and the rating of the highest-rated rural tourist spots, indicating that areas with better hotel infrastructure tend to receive higher ratings. While there is little to no significant correlation between hotel facilities and the ratings for lower-rated rural tourist spots, there is a strong and significant positive correlation for five-coconut-rated spots. This implies that better hotel facilities are a key factor contributing to the higher ratings of these top-tier rural tourism destinations.

Table 10 shows the correlation between A-level tourist attractions and rural tourist spot ratings. For lower-rated spots (one to four coconuts), correlations are weak and not statistically significant. However, for five-coconut-rated spots, there is a strong positive correlation (0.690) with a significance of 0.002, indicating that better A-level attractions are strongly associated with higher-rated rural tourism spots. This suggests that A-level attractions significantly influence the quality of top-rated rural tourism destinations.