Key points are not available for this paper at this time.
Traditional villages are human treasures left behind by the integration of material space and non-material culture in the process of agricultural civilization. Studying the spatial autocorrelation characteristics, heterogeneity, and quantitative attribution of the factors influencing traditional villages provides new ideas for the protection of traditional villages. This study took 75 traditional villages as the research object. From the perspective of spatial autocorrelation and spatial heterogeneity, the study used nuclear density estimation and Moran’s I index to analyze the spatial distribution patterns and selected 12 factors to construct the GWR modeling and geodetector to analyze the main driving forces and the interaction mechanism. The results showed that, firstly, the overall spatial layout of traditional villages in the Awa Mountain area had two cores, two sides, and a scattered distribution; the global Moran’s I was 0.774, and 55.6% of traditional villages showed a clustering phenomenon. Second, the spatial layout of traditional villages in the Awa Mountain area has been jointly promoted and mutually constrained by multiple factors in a dynamic and complex mechanism with obvious spatial heterogeneity. The natural factor is the basic factor, which determines the location and scale of development of villages; the spatial factor is the auxiliary factor; the social factor is the decisive factor, with a negative global correlation and a positive local correlation; the regional cultural factor is the key factor, and the regional factor and the social factor complement each other; and factors such as a backward economic level, restricted transportation, less external communication, and low population density play a protective role. Third, the main driving factor is the proportion of ethnic minorities (X10), and the explanatory power of q-value reaches 0.54; the proportion of ethnic minorities (X10) ∩ average annual precipitation (X4) has the strongest interactive driving force, which belongs to nonlinear enhancement, and the q-value is 0.93, which proves that the explanatory power of the two-factor model is much greater than the single-factor model from the system perspective.
Li et al. (Mon,) studied this question.