Spatial structure characteristics of traditional villages
The study classify and organize 724 traditional villages from five batches, and create a radar chart to visually the quantity structure of traditional villages in Guizhou Province at different times, as shown in Fig. 3a–e.
From Fig. 3a–e, it can be seen that regardless of the period or batch, Qiandongnan has the highest number of traditional villages, ranking first among the nine cities, while Guiyang and Bijie have the lowest number of traditional villages. As shown in Fig. 3a–e, in the first, second, third, and fourth batches, the number of traditional villages in Tongren is second only to Qiandongnan. As can be seen from Fig. 3e. In the fifth batch, the number of traditional villages in Qiannan exceeded that of Tongren, ranking second.
To visualize the evolutionary characteristics of traditional villages in Guizhou Province, five batches of traditional villages were imported into ArcGIS 10.4, and the distribution of different batches of traditional villages in 88 counties was presented, as shown in Fig. 4a–e.
(a–e) Evolution characteristics of traditional villages in five batches of Guizhou Province.
As shown in Fig. 4a–e, overall, the five batches of traditional villages are mainly concentrated in the southeast of Guizhou, and the number is relatively concentrated in the central and northeastern parts of Guizhou. As shown in Fig. 4a, c, and d, it can be seen that in the first, third, and fourth batches of traditional villages, the number of traditional villages in Xixiu District of Anshun City is relatively concentrated.
Spatial imbalance of traditional villages
Coefficient of variation
To measure the degree of difference in traditional village indicators among 9 cities in Guizhou Province, the study used formula (1) to calculate the coefficient of variation values of traditional villages in 9 cities, as shown in Table 1.
From Table 1, it can be seen that the coefficient of variation in Qiandongnan is the highest, which indicates that the traditional villages in Qiandongnan have the highest volatility, it means the number of traditional villages in Qiandongnan deviates the most from the mean. And the coefficient of variation in Anshun City is the lowest, which indicates that the volatility of traditional villages in Anshun City is the smallest, it means the number of traditional villages in Anshun City is closest to the mean.
Herfindahl coefficient value
To measure the comprehensive index of concentration of traditional villages in 9 cities in Guizhou Province and reflect the degree of regional agglomeration of traditional villages in China, the Herfindahl coefficient was calculated using formula (2), as shown in Table 2.
From Table 2, Qiandongnan’s H > 0.18, it indicates that traditional villages have a high degree of agglomeration, and the agglomeration type belongs to the high oligopoly type. Except for Qiandongnan, the H of the other 8 cities is less than 0.1, it indicates that the clustering degree of traditional villages in these 8 cities is low, and the clustering type is competitive type.
Standard deviation ellipse
To reflect the centrality, spatial range, and evolutionary direction of the geographical elements of five batches of traditional villages, this study combined formulas (4), (5), and (6) to draw the standard deviation ellipse of traditional villages using ArcGIS 10.4, as shown in Fig. 5 and Table 3.
Standard deviation ellipse of five batches of traditional Chinese villages in Guizhou Province.
From Fig. 5, it can be seen that the first batch of traditional Chinese villages in Guizhou Province are mainly distributed along the southeast-northwest direction, and are mainly distributed in the southeast of Guizhou and Tongren. The second batch is mainly distributed along the southeast-northwest direction, and is mainly distributed in Qiandongnan and Tongren. The third batch is mainly distributed along the southwest- northeast direction, and is mainly distributed in Qiandongnan, Anshun, and Tongren. The fourth batch is mainly distributed along the southwest-northeast direction, and is mainly distributed in Qiandongnan, Tongren, and Anshun, with a small portion distributed in Zunyi. The fifth batch is mainly distributed along the southeast-northwest direction, and is mainly distributed in Qiandongnan and Anshun, with a small portion distributed in Zunyi and Qiannan.
The center of gravity of various batches of traditional villages in Guizhou from 2012 to 2019 in Fig. 5 first moved from southeast to northwest, and finally from northwest to southeast. And the center of gravity of the standard deviation ellipse all fell in the area of Qiandongnan, which indicates that the traditional villages in Qiandongnan ranked first in Guizhou Province.
In Table 3, the larger the area of the standard deviation ellipse, the wider the distribution of traditional villages in Guizhou Province, and vice versa. The short half axis of the standard deviation ellipse represents the range of data distribution. The shorter the short half axis is, the more pronounced the centripetal force presented by the data is. From Table 3, it can be seen that in the ratio of area, the fourth batch > the third batch > the fifth batch > the first batch > the second batch. Among the five batches, the fourth batch of traditional villages is the most widely distributed, followed by the third, fifth, and first batches, and the second batch has the narrowest distribution range. In the short half axis ratio, the fourth batch > the fifth batch > the third batch > the second batch > the first batch. The fourth batch of traditional villages had the highest degree of dispersion.
Overall, the five batches of traditional villages are mainly distributed in the eastern and southwestern regions of Guizhou Province, with the vast majority distributed in Qiandongnan Anshun, Tongren, and Zunyi, while a small number are distributed in other cities.
Spatial correlation
To explore the spatial correlation characteristics of traditional villages in Guizhou Province, according to formula (7), five batches of traditional villages were imported into ArcGIS 10.4 to obtain the global Moran index value, the Z-score and P-value are shown in Table 4.
From Table 4, it can be seen that the global Moran index of the five batches of traditional villages is greater than zero and shows a gradually increasing trend, it indicates that traditional villages in Guizhou Province are positively correlated in space. That is, the closer the global Moran index of traditional villages is to 1, the higher the degree of positive correlation with space; The Z-scores of each batch of traditional villages are all greater than 1.96, and the P-values are all less than 0.05, rejecting the null hypothesis, it indicates that there is a significant spatial autocorrelation in traditional villages in Guizhou Province.
The global Moran index can only reflect the spatial correlation characteristics of traditional villages in Guizhou Province as a whole, therefore, this study further applies the local spatial autocorrelation of formula (8) to analyze the local spatial characteristics of the batches of traditional villages in Guizhou Province.
According to formula (8), traditional villages in Guizhou are imported into ArcGIS 10.4 to obtain the spatial agglomeration distribution area of traditional villages in Guizhou Province, as shown in Fig. 6a–f and Table 5.
(a–f) Localised Moran index of five batch traditional villages in Guizhou Province.
According to Fig. 6a–f and Table 5, it is found that traditional villages exhibit an H-H, H-L, L-H, and L-L model. Overall, the spatial distribution of traditional villages in Guizhou Province is dominantly characterized by H-H, L-L clustering.Followed by L-H clustering, with a small amount of H-L local spatial clustering distribution observed in the analysis of traditional villages in the second, fourth, and fifth batches. Among them, The H-H aggregation area is mainly distributed in Qiandongnan, while the third and fourth batches are slightly distributed in Tongren and Anshun. Qiandongnan is a typical multi-ethnic region and also the most densely populated area of the Miao and Dong ethnic groups in China. The administrative region has preserved unique ethnic symbols such as residential architecture, food culture, folk festivals, song and dance art, and traditional skills, and has built a traditional village cluster with ethnic minority cultural characteristics. Tongren also promotes the gathering of traditional village elements due to the unique ethnic culture generated by the settlement of ethnic minorities such as Miao, Dong, and Tujia. Anshun, on the other hand, was influenced by the “military garrison system” of the Ming Dynasty in China and retained the customs constrained by historical policies in terms of village layout, residential architecture, and ethnic costumes, thus forming a traditional village group with historical and cultural characteristics. The L-L aggregation area is mainly distributed in Guiyang and Bijie, with a small amount distributed in Zunyi and Qiannan. As the capital of Guizhou Province, Guiyang has rapid economic development, strong cultural inclusiveness and high urbanization rate, while Bijie, Zunyi and Qiannan are spatially adjacent to Guiyang and driven by the radiation of the center of Guiyang, with good development of its urbanization, infrastructure, etc. Therefore, the gathering of traditional villages in these four regions is relatively scattered, thus forming a “low-low” gathering pattern.
Tourism attractiveness
To measure the tourism attractiveness of traditional villages, this study takes the total tourism income, total tourism visits, per capita tourism expenses, and travel distance between different regions in 2023, and using formulas (9) and (10), to calculate the tourism attractiveness values Rij of 9 cities in Guizhou, and then imported Rij into ArcGIS 10.4 for visualization, as shown in Fig. 7.
Tourism attractiveness of traditional villages.
As seen in Fig. 7, the tourism attractiveness of each city presents centering on Guiyang and radiating to the five cities under the Central Economic Circle. That is, Guiyang and Zunyi, Guiyang and Qiandongnan, Guiyang and Anshun, Guiyang and Qiannan, Guiyang and Bijie have larger values of tourism attractiveness. It indicates that Guiyang has the strongest degree of tourism economic ties with these five cities. Among them, the tourism attractiveness between Guiyang and Zunyi is especially the highest, indicating that the degree of tourism economic connection between Guiyang and Zunyi is the strongest. In addition, the tourism attractiveness values of Guiyang and Qiandongnan follow Guiyang and Zunyi are in the middle. This indicates that Guiyang and Qiandongnan have stronger tourism attraction and closer tourism economic ties between the two cities. It can be seen that the number of traditional villages affects the tourism economic links between regions and further affects the development of the tourism industry between regions.
Transportation accessibility
To analyze the road network accessibility of 724 traditional villages, the study used network analysis method combined with road network data from Guizhou Province to obtain the road network accessibility of 88 districts and counties of Guizhou Province, as shown in Fig. 8.
Transportation accessibility of traditional villages in Guizhou Province.
As shown in Fig. 8, traditional villages are mainly centered around Qiandongnan, forming six isochronous circles. Among them, traditional villages in Qiandongnan have the best accessibility, with a reach time between 1.5 and 2 h. Secondly, the accessibility of traditional villages in most areas of Qiannan, Guiyang, and Tongren is second only to Qiandongnan, with a reach time of 2–3 h. Thirdly, in some areas of Anshun, Zunyi, and Tongren, traditional villages can reach within 3–3.5 h. Fourthly, the accessibility of traditional villages in most areas of Qianxinan, Liupanshui, Bijie, and a small part of Zunyi is poor, with a reach time of over 4 h. It can be seen from Fig. 8 that, there is a positive relationship between the concentration of traditional villages and accessibility. That is, the more concentrated the traditional villages are, the better the accessibility.
Analysis of factors affecting relevance
The spatial distribution pattern of traditional villages can be affected by geographic location, elevation, slope direction and population factors. For example, in Guizhou, the Miao people tend to live on the hillsides and the Buyi people live on the riversides. At the same time, traditional villages, as an important rural tourism resources, will affect the distribution of tourists as well as influence the connection between tourist destinations. The generation of tourism attraction is closely related to the endowment of tourism resources, the degree of perfection of tourism service facilities, and the level of development of tourism economy. In addition, the tourism system is the result of the interaction between tourism sources, tourism destinations and tourism corridors. Transportation is seen as the medium between tourism sources and destinations and as a prerequisite for the development of tourism, it is both an important “engine” for tourism development and a key driver for enhancing the centrality of tourism destinations26. The magnitude of the interaction nodes in a transportation network is called transportation accessibility27. The convenience of transportation affects tourists’ willingness to travel28, and the accessibility of transportation affects tourists’ satisfaction29, in addition, the degree of integration of transportation and tourism affects the attractiveness of tourist destinations, and also affects the exchange and integration of passenger, capital, information, and material flows, and then affects the efficiency and level of inter-regional tourism cooperation30. Adeola and Evans argue, transportation accessibility of tourist attractions refers to the convenience of tourists to reach tourist attractions through various modes of transportation, covering transportation cost, transportation time, transfer efficiency31. Arinta et al. argued that congestion due to insufficient transportation capacity will significantly lengthen the time to reach tourist attractions and reduce accessibility32.
In order to accurately investigate the spatial pattern of traditional villages and the influencing factors of tourism attractiveness and transportation accessibility, the study selected 12 factors of DEM, elevation, acreage, demographic, tradional village, Intangible Cultural Heritage, A tourist attractions, hotels, total tourist arrivals, passenger terminal, transportation network density, miles of roads in operation as the measurement data. The study categorized the above 12 indicators into three dimensions, spatial differentiation system, tourism support system, and transportation facilitation environment system, as shown in Table 6.The study first used the ordinary least squares (OLS) method for the measurement, and the results found that the variance inflation factor (VIF) of the 12 indicators were less than 7.5, and there were no redundant indicators.
Then the geographically weighted regression was used to measure it, and the study listed “total tourism income” as the dependent variable and the above three dimensions as the explanatory variables, among which the R2 of the spatial differentiation system, the tourism support system, and the transportation facilitation system were all greater than 0.5, which showed that the model fit was better in general, and the specific spatial fitting differences are shown in Fig. 9a–c.
GWR results of spatial differences in the impact of systems on tourism development.
As can be seen from Fig. 9a–c, the standard error values of the three systems are less than 2.5, and the model indicators are relatively reliable. To further explore the relationship between the indicators, the study used the Spearman method to measure the correlation between them, and the results are shown in Fig. 10.
Heat map of Spearman square correlation analysis.
As can be seen from Fig. 10, firstly, there is a significant positive correlation between Demographic and elevation, Intangible culture heritage and traditional village, and total tourist arrivals and hotels. It can be seen that the degree of traditional village agglomeration will affect the advantages and disadvantages of rural cultural resources and rural tourism resource endowment. The advantages and disadvantages of resource endowment will make the evolution and development of tourist destinations produce polarization effect and convergence effect, which also affects the tourism vitality and tourism attraction of tourist destinations. The regional superposition effect produced by traffic accessibility will accelerate the formation of the central position of the tourist destination, the core advantage, and then contribute to the tourist destination to produce regional radiation effect, and promote the formation of the tourism corridor with traffic as the medium. Secondly, there is a relatively significant correlation between transportation network density and grade A tourist attractions, total tourist arrivals and passenger terminal, total tourist arrivals, total tourist arrivals and evelvation, total tourist arrivals and demographic. It can be seen that road transportation, as an important public facility infrastructure for promoting the high-quality development of tourism destinations, play a role in enhancing the attractiveness of tourism destinations and promoting tourism economy exchanges between cities. It is evident that the pioneering and strategic nature of transport, in conjunction with the integrated, interrelated and open nature of the tourism industry, facilitates the flow of material, information, financial and passenger flows to tourist destinations. In short, “traveling” and “touring” have become the new realities of tourism. Convenient, comfortable and cost-effective transportation accessibility will reduce the attraction decay due to the problem of long distance, promote the willingness to travel and improve tourist satisfaction, which is important to promote the transformation and upgrading of traditional villages as well as the high-quality development of tourist destinations. Nowadays, connecting scenic spots through roads, linking attractions through highways, and integrating scenery through roads have become the main trend to accelerate the development of tourist destinations. The series of derivatives developed around the integration of traffic and tourism, such as the integration of bridge tourism, cruise tourism, train tourism, self-driving tour and other new products will continue to inject new vitality into the development of tourist destinations.